EUROPEAN
COMMISSION
DG Competition
PUBLIC VERSION
Case M. 10262-META (FORMERLY FACEBOOK) /
KUSTOMER
(Only the English text is authentic)
REGULATION (EC) No 139/2004
MERGER PROCEDURE
Article 8(2) Regulation (EC) 139/2004
Date: 27/01/2022
This text is made available for information purposes only. A summary of this decision is
published in all EU languages in the Official Journal of the European Union.
Parts of this text have been edited to ensure that confidential information is not disclosed;
those parts are enclosed in square brackets.
EUROPEAN
COMMISSION
Brussels, 27.1.2022
C(2022) 409 final
COMMISSION DECISION
of 27.1.2022
declaring a concentration to be compatible with the internal market and the EEA
agreement
(Case M.10262 – META (FORMERLY FACEBOOK) / KUSTOMER)
(Only the English text is authentic)
1
TABLE OF CONTENTS
1. The Parties and the Operation ...................................................................................... 6
2. Jurisdiction ................................................................................................................... 7
3. The Procedure .............................................................................................................. 7
4. The Commission’s Investigation.................................................................................. 9
5. Industry Overview ........................................................................................................ 9
5.1. CRM software solutions ............................................................................................. 11
5.1.1. Types of CRM software solutions .............................................................................. 11
5.1.2. Data held by CRM software solutions ....................................................................... 11
5.2. B2C communications ................................................................................................. 12
5.2.1. B2C communication channels and CRM software .................................................... 13
5.2.2. B2C communication channels, CPaaS solutions and CRM software solutions ......... 13
5.3. Online advertising sector ............................................................................................ 14
6. Relevant Market Definition ........................................................................................ 16
6.1. CRM software ............................................................................................................ 16
6.1.1. Product market definition ........................................................................................... 17
6.1.1.1. Commission precedents ............................................................................................. 17
6.1.1.2. The Notifying Party’s view ........................................................................................ 17
6.1.1.3. Commission’s assessment .......................................................................................... 18
6.1.1.4. Conclusion.................................................................................................................. 23
6.1.2. Geographic market definition .................................................................................... 23
6.1.2.1. Commission precedents ............................................................................................. 23
6.1.2.2. The Notifying Party’s view ........................................................................................ 24
6.1.2.3. Commission’s assessment .......................................................................................... 24
6.1.2.4. Conclusion.................................................................................................................. 24
6.2. B2C communication services ..................................................................................... 25
6.2.1. Product market definition ........................................................................................... 25
6.2.1.1. Commission precedents ............................................................................................. 25
6.2.1.2. The Notifying Party’s view ........................................................................................ 26
6.2.1.3. Commission’s assessment .......................................................................................... 27
6.2.1.4. Conclusion.................................................................................................................. 34
6.2.2. Geographic market definition .................................................................................... 35
6.2.2.1. Commission precedents ............................................................................................. 35
6.2.2.2. The Notifying Party’s view ........................................................................................ 36
6.2.2.3. Commission’s assessment .......................................................................................... 36
6.2.2.4. Conclusion.................................................................................................................. 37
2
6.3. Online Display Advertising ........................................................................................ 37
6.3.1. Product market definition ........................................................................................... 37
6.3.1.1. Commission precedents ............................................................................................. 37
6.3.1.2. The Notifying Party’s view ........................................................................................ 37
6.3.1.3. Commission’s assessment .......................................................................................... 38
6.3.1.4. Conclusion.................................................................................................................. 41
6.3.2. Geographic market definition .................................................................................... 42
6.3.2.1. Commission precedents ............................................................................................. 42
6.3.2.2. The Notifying Party’s view ........................................................................................ 42
6.3.2.3. Commission’s assessment .......................................................................................... 42
6.3.2.4. Conclusion.................................................................................................................. 42
7. Competitive Assessment ............................................................................................ 43
7.1. Introduction ................................................................................................................ 43
7.1.1. Rationale for the Transaction ..................................................................................... 43
7.1.2. Affected markets ........................................................................................................ 43
7.2. Market Shares ............................................................................................................. 44
7.2.1. CRM software ............................................................................................................ 44
7.2.2. B2C communication services ..................................................................................... 47
7.2.3. Online Display Advertising ........................................................................................ 50
7.3. Assessment of vertical effects .................................................................................... 59
7.3.1. Legal framework ........................................................................................................ 59
7.3.2. Foreclosure of access to APIs to Meta (formerly Facebook)’s messaging channels to
the detriment of competing providers of CRM software (input foreclosure) ............ 60
7.3.2.1. Notifying Party’s views.............................................................................................. 61
(a) Ability to foreclose ..................................................................................................... 61
(b) Incentive to foreclose ................................................................................................. 66
(c) Effects on competition ............................................................................................... 68
7.3.2.2. Commission’s Assessment ......................................................................................... 70
(a) Ability to foreclose .................................................................................................... 70
(a.i) Important input for the downstream market ............................................................... 70
(a.ii) Meta (formerly Facebook) holds market power on the upstream market for OTT B2C
messaging ................................................................................................................... 77
(a.iii) Meta (formerly Facebook) has the ability to restrict or degrade access to its APIs ... 83
A. Meta (formerly) Facebook has the ability to restrict access to its APIs outright or on a
selective basis ............................................................................................................. 83
B. Meta (formerly Facebook) has the ability to degrade API access to its messaging
channels for all third parties or on a selective basis ................................................... 85
3
C. Meta (formerly Facebook) would be able to identify a specific group of CRM
providers to target ...................................................................................................... 89
D. Foreclosed CRM providers would not be able to evade or workaround such targeting
by other means .......................................................................................................... 90
E. Meta (formerly Facebook) is unlikely to have the ability to obtain data on CRM
competitors to provide a competitive advantage to Kustomer ................................... 92
(a.iv) Conclusion.................................................................................................................. 92
(b) Incentive to foreclose ................................................................................................. 93
(b.i) Benefits from foreclosure are numerous, diverse and significant .............................. 96
A. SaaS revenues ...................................................................................................................... 96
B. Additional data for online ads purposes .............................................................................. 98
C. CTM ads ............................................................................................................................ 107
D. Meta (formerly Facebook)’s ecosystem of products ......................................................... 110
(b.ii) Meta (formerly Facebook) can limit losses from foreclosure .................................. 111
A. Losses from foreclosure .................................................................................................... 112
B. Switching is likely to be sufficiently significant to limit losses ........................................ 116
C. Aggravating factors ........................................................................................................... 124
(b.iii) Conclusion................................................................................................................ 126
(c) Impact on effective competition............................................................................... 126
(c.i) Potentially foreclosed firms play a sufficiently important role in the competitive
process ...................................................................................................................... 127
A. Small CRM providers focus on SMB customers (while large CRM providers focus on
large enterprises) ...................................................................................................... 128
B. Small CRM players that target SMB customers play a disproportionally important role in
the competitive process (as these are particular drivers of innovation) ................... 130
(c.ii) The mere likelihood that the merged entity would carry out a foreclosure strategy
will likely already have anticompetitive effects ....................................................... 141
(c.iii) Competition from vertically integrated downstream competitors or alternative input
providers would not constrain the merged entity ..................................................... 141
(c.iv) Countervailing factors would not be sufficient to maintain effective competition .. 142
(c.v) Claimed efficiencies have not been substantiated .................................................... 142
(c.vi) Conclusion................................................................................................................ 143
7.3.2.3. Conclusion................................................................................................................ 144
7.4. Assessment of horizontal effects .............................................................................. 144
7.4.1. Legal Framework ..................................................................................................... 147
7.4.2. Data held by Kustomer ............................................................................................. 148
7.4.3. Raised barriers to entry and expansion as a result of data accumulation ................. 149
7.4.3.1. Notifying Party’s views............................................................................................ 149
7.4.3.2. Commission’s Assessment ....................................................................................... 150
4
(a) Meta (formerly Facebook)’s market position in online display advertising ............ 150
(a.i) Meta (formerly Facebook) holds significant market power in online display
advertising ................................................................................................................ 150
(a.ii) Meta (formerly Facebook) has significant data collection capabilities ................... 151
(a.iii) Conclusion................................................................................................................ 154
(b) Potential data accumulation through the acquisition of Kustomer .......................... 154
(b.1) Meta (formerly Facebook) could encourage business customers to agree to the use of
data ........................................................................................................................... 154
(b.ii) Data stored by Kustomer that is accessible to Meta (formerly Facebook) today .... 155
(b.iii) Conclusion................................................................................................................ 156
(c) Effects of data accumulation on the market for online display advertising services 156
(c.i) Barriers to entry and expansion ................................................................................ 157
(c.ii) Availability of similar datasets and counterstrategies .............................................. 159
(c.iii) Conclusion................................................................................................................ 160
7.4.3.3. Conclusion................................................................................................................ 160
7.5. Assessment of conglomerate effects ........................................................................ 160
7.5.1. Legal Framework ..................................................................................................... 160
7.5.2. Leveraging of Meta (formerly Facebook)’s position in the online display advertising
market into the CRM market .................................................................................... 161
7.5.2.1. Notifying Party’s views............................................................................................ 162
7.5.2.2. Commission’s Assessment ....................................................................................... 163
(a) Ability to foreclose ........................................................................................................... 163
(b) Incentive to foreclose ........................................................................................................ 166
(c) Impact on effective competition ....................................................................................... 166
7.5.2.3. Conclusion................................................................................................................ 166
8. Commitments ........................................................................................................... 166
8.1. Introduction .............................................................................................................. 166
8.2. Analytical Framework .............................................................................................. 167
8.3. The Initial Commitments ......................................................................................... 168
8.3.1. Description of the Initial Commitments ................................................................... 168
8.3.1.1. The public API access commitment ......................................................................... 168
8.3.1.2. The core API functionality-parity commitment ....................................................... 169
8.3.1.3. Anti-circumvention, carve-outs and other aspects ................................................... 169
8.3.2. Results of Markets Tests of the Initial Commitments .............................................. 170
8.3.3. Commission’s assessment of the Initial Commitments ........................................... 176
8.4. The Final Commitments ........................................................................................... 177
8.4.1. Description of the Final Commitments ................................................................... 177
8.4.2. Commission’s assessment of the Final Commitments ............................................. 179
5
9. Conclusion................................................................................................................ 181
10. Conditions and Obligations ...................................................................................... 181
6
COMMISSION DECISION
of 27.1.2022
declaring a concentration to be compatible with the internal market and the EEA
agreement
(Case M.10262 – META (FORMERLY FACEBOOK) / KUSTOMER)
(Only the English text is authentic)
THE EUROPEAN COMMISSION,
Having regard to the Treaty on the Functioning of the European Union
1
,
Having regard to the Agreement on the European Economic Area, and in particular Article 57
thereof,
Having regard to Council Regulation (EC) No 139/2004 of 20.1.2004 on the control of
concentrations between undertakings
2
, and in particular Article 8(2) thereof,
Having regard to the Commission's decision of 2 August 2021 to initiate proceedings in this
case,
Having given the undertakings concerned the opportunity to make known their views on the
objections raised by the Commission,
Having regard to the opinion of the Advisory Committee on Concentrations
3
,
Having regard to the final report of the Hearing Officer in this case
4
,
Whereas:
1. THE PARTIES AND THE OPERATION
(1) On 25 June 2021, the Commission received a notification of a proposed
concentration pursuant to Article 4 of the Merger Regulation by which Meta
Platforms, Inc., formerly Facebook, Inc. (“Meta” or “Facebook” or the “Notifying
Party”, U.S.) acquires within the meaning of Article 3(1)(b) of the Merger
Regulation sole control of Kustomer, Inc. (“Kustomer”, U.S.) (the “Transaction”).
Facebook and Kustomer are collectively referred to as “Parties”.
(2) Meta (formerly Facebook) was founded in 2004 and is headquartered in the U.S. It
provides various websites and applications for mobile devices offering social
networking, consumer communications and photo and video-sharing functionalities.
These include Facebook, Messenger, Instagram and WhatsApp. Meta (formerly
Facebook) offers most of its services at no cost to users. Meta (formerly Facebook)’s
1
OJ C 115, 9.8.2008, page 47 (“TFEU”).
2
OJ L 24, 29.1.2004, page 1 (the “Merger Regulation”). With effect from 1 December 2009, the TFEU
has introduced certain changes, such as the replacement of ‘Community’ by ’Union’ and ‘common
market’ by ‘internal market’. The terminology of the TFEU will be used throughout this decision.
3
Opinion of the Advisory Committee on Concentrations of 14 January 2022.
4
Final report of the Hearing Officer.
7
primary revenue-generating activity is offering ads space and related services (such
as tools for designing, and measuring the effectiveness of, ads campaigns) to third
parties, which currently accounts for the vast majority of its annual revenues
globally. Meta (formerly Facebook)’s FY 2020 worldwide revenues were EUR
[Revenue Information], of which EUR [Revenue Information] or [Revenue
Information] were generated in the EEA (excluding the UK).
(3) Kustomer was founded in 2015 and is headquartered in the U.S. It offers a Software
as a Service (“SaaS”) customer relationship management (“CRM”) software
specialising in customer service. It is designed to help agents in a company’s
customer service function manage communications with consumers, e.g., when a
consumer contacts the business with a product enquiry, in connection with a late
delivery or a faulty product or return. Kustomer displays all interactions that a
consumer has had with the business in a single timeline view, including the
consumer’s purchase history, refunds, complaints and all other communications.
Kustomer’s CRM software integrates with a wide range of business-to-consumer
(“B2C”) communication channels and supports agent communications with
consumers through communication channels including phone, email, webchat, SMS,
Messenger, WhatsApp, Instagram and Twitter. Kustomer’s FY 2020 worldwide
revenues were EUR [Revenue Information], of which EUR [Revenue Information] or
[Revenue Information] were generated in the EEA (excluding the UK). Kustomer’s
worldwide and EEA revenues are expected to be EUR [Revenue Information] and
EUR [Revenue Information] respectively in FY 2021.
(4) The Transaction consists in the acquisition of sole control by Meta (formerly
Facebook) over Kustomer, pursuant to a merger agreement dated [Transaction
Structure]. Therefore, the Transaction constitutes a concentration within the meaning
of Article 3(1)(b) of the Merger Regulation.
2. JURISDICTION
(5) The Transaction does not have a Union dimension within the meaning of Article 1 of
the Merger Regulation as the Parties' turnover does not meet the thresholds of Article
1(2) or 1(3) of the Merger Regulation.
5
(6) On 2 April 2021, the Commission received a referral request from the Federal
Competition Authority of Austria (the “Austrian NCA”) pursuant to Article 22(1) of
the Merger Regulation.
6
The national competition authorities of Belgium, Bulgaria,
France, Iceland, Ireland, Italy, the Netherlands, Portugal and Romania subsequently
joined the request made by the Austrian NCA. On 12 May 2021, the Commission
accepted the request and therefore decided to examine the Transaction pursuant to
Article 22(3) of the Merger Regulation.
3. THE PROCEDURE
(7) The Transaction was notified to the Commission on 25 June 2021.
(8) After a preliminary examination of the notification and based on the first phase
market investigation, the Commission raised serious doubts as to the compatibility of
the Transaction with the internal market and adopted a decision to initiate
5
Kustomer achieves revenues of less than EUR [Revenue Information] in the EEA.
6
The Transaction was notified to the Austrian NCA on 31 March 2021 on the basis that it met the
applicable turnover thresholds under Austrian competition rules.
8
proceedings pursuant to Article 6(1)(c) of the Merger Regulation on 2 August 2021
(the "Article 6(1)(c) Decision").
(9) On 2 August 2021, the Commission provided a number of key documents to the
Notifying Party.
(10) On 6 August 2021, the second phase investigation period was extended by 5 working
days at the request of the Notifying Party pursuant to the first sentence of the second
subparagraph of Article 10(3) of the Merger Regulation.
(11) On 23 August 2021, the Notifying Party submitted its written comments to the
Article 6(1)(c) Decision (the "Article 6(1)(c) Response").
(12) On 24 August 2021, the second phase investigation period was extended by 10
working days in agreement with the Notifying Party pursuant to the third sentence of
the second subparagraph of Article 10(3) of the Merger Regulation.
(13) On 27 August 2021, a State of Play meeting between the Notifying Party and the
Commission took place.
(14) The Commission held technical meeting with the Parties on 21, 22, 23 and 24
September 2021.
(15) On 29 September 2021, the Commission provided preliminary feedback to the
Notifying Party on the results of its second phase market investigation.
(16) On 10 October 2021, the Notifying Party submitted a paper setting out their reactions
to the feedback provided by the Commission on 29 September 2021. The underlying
data and calculations were shared with the Commission two days later on 12 October
2021.
(17) A State of Play meeting between the Notifying Party and the Commission took place
on 14 October 2021 during which the Commission explained its preliminary findings
based on its second phase investigation.
(18) After reviewing the available evidence collected in Phase I and Phase II, the
Commission came to the preliminary view that it is likely, and there is even a strong
probability, that the Transaction would significantly impede effective competition in
a substantial part of the internal market within the meaning of Article 2(3) of the
Merger Regulation, and issued a Statement of Objections to the Notifying Party on
18 October 2021 (the “SO”).
(19) On 28 October 2021, the Notifying Party confirmed that it would not request an Oral
Hearing.
(20) On 3 November 2021, the Notifying Party submitted its written comments on the SO
(the “SO Response”).
(21) A State of Play meeting between the Notifying Party and the Commission took place
on 18 November 2021.
(22) On 24 November 2021, the Notifying Party submitted commitments pursuant to
Article 8(2) second subparagraph of the Merger Regulation in order to address the
competition concerns identified by the Commission. Accordingly, the legal deadline
for the Commission decision was automatically extended by 15 working days,
pursuant to Article 10(3) first subparagraph of the Merger Regulation.
(23) On 26 November 2021, the Commission launched a market test of the commitments
submitted by the Notifying Party.
9
(24) On 3 December, the Commission sent a Letter of Facts to inform the Notifying Party
about additional factual elements which were not yet expressly relied on in the SO
and that would be potentially relevant for its final decision (the “LoF”). The
Notifying Party submitted a response to the LoF (the “LoF Response”) on
13 December 2021.
(25) The Commission gave the Notifying Party detailed feedback on the outcome of the
market test during a State of Play meeting on 10 December 2021.
(26) On 17 December 2021, the second phase investigation period was extended by five
working days in agreement with the Notifying Party pursuant to the third sentence of
the second subparagraph of Article 10(3) of the Merger Regulation.
(27) On 20 December 2021, the Notifying Party submitted revised and final commitments
pursuant to Article 8(2) of the Merger Regulation.
(28) On 22 December 2021, the Commission sent a draft of the decision taken pursuant to
Article 8(2) of the Merger Regulation (the “Decision”) to the Advisory Committee
with the view of seeking the Committee’s opinion. The meeting of the Advisory
Committee took place on 14 January 2022 and the Committee issued its positive
opinion.
4. THE COMMISSION’S INVESTIGATION
(29) This Decision contains the Commission's findings on the basis of the market
investigation it carried out prior to the notification of the Transaction, in the first
phase and in the second phase of the investigation until the adoption of the Decision.
(30) Prior to the notification of the Transaction, the Commission sent eleven requests for
information (“RFI”) to the Parties, responses to which were included in the
notification. The Commission also exchanged draft data requests with Meta
(formerly Facebook) and its competitors in the market for online advertising. Finally,
the Commission conducted several interviews with the Parties' customers and
competitors.
(31) During the first phase investigation, the Commission sent two requests for
information to the Parties pursuant to Article 11 of the Merger Regulation. The
Commission also sent requests for information in the form of four different
questionnaires to competitors and customers of the Parties. The Commission also
finalised the draft data requests it had exchanged with Meta (formerly Facebook) and
its competitors in the market for online advertising, and sent these final data requests.
(32) Over the course of its second phase investigation, the Commission sent twelve
requests for information to the Parties pursuant to Article 11 of the Merger
Regulation, including two detailed internal documents requests, resulting in the
submission of over [1-2] million internal documents (approximately [1-1.5] million
internal documents from Meta (formerly Facebook) and approximately [300,000-
500,000] from Kustomer). Further, the Commission held several calls with market
participants and sent several requests for information to competitors and customers
of the Parties, including four different questionnaires.
5. INDUSTRY OVERVIEW
(33) In this Section, the Commission provides an overview of all sectors that are relevant
for the purpose of assessing the Transaction in order to provide context for the
competitive assessment undertaken in Section 7.
10
(34) Businesses and consumers communicate with each other using a range of different
communication channels including traditional channels like phone, email and more
modern channels such as messaging channels (e.g., Meta (formerly Facebook)’s
WhatsApp or Messenger, Twitter, or webchat
7
). Businesses can procure B2C
communication services either directly from the relevant provider (e.g., from a fixed
telephone provider, an enterprise email provider, or Meta (formerly Facebook)) or
indirectly from intermediaries that provide connectivity to multiple communication
channels. These include CRM software tools, such as the SaaS CRM tool offered by
Kustomer, or Communication-Platform-as-a-Service (“CPaaS”) solutions, which are
cloud-based middleware solutions that integrate communication channels (such as
voice, email, SMS, webchat and over-the-top (“OTT”)
8
messaging) into their
applications via application programming interfaces (“APIs”)
9
and then provide
businesses (including CRM software providers) with access to the relevant channels
without needing to build backend infrastructure and interfaces.
(35) The figure below sets out the manner in which such communications can take place.
Source: The Notifying Party (Form CO, Figure 6.2)
(36) In general, the means and channels through which a business will typically
communicate with its customers can depend on various factors, including its size, the
sector it operates in, and the types of customers it serves. At one end of the spectrum,
7
A webchat is accessible via a web browser and does not require a specific chat software.
8
OTT messaging channels are messaging channels that are provided via an internet connection, as
opposed to messaging channels that are provided via a cellular (i.e., mobile phone) network, such as
SMS. OTT messaging channels include Meta (formerly Facebook)’s messaging channels (Messenger,
Instagram messaging and WhatsApp) as well as third party messaging channels such as Twitter DM,
Telegram, Line and WeChat.
9
An API is a particular set of rules and specifications that a software programme follows in order to
access and make use of the services and resources provided by another software programme or
hardware that also implements that API. In essence, APIs allow software programmes and hardware, or
different software programmes, to communicate with each other.
11
it is typical for small businesses
10
to procure B2C messaging channels directly (i.e.,
to communicate with customers via a phone number, email address, Facebook or
WhatsApp account or any other means), not to have dedicated customer service staff,
and not to use a CRM software tool. Medium to large businesses (i.e., with a
reasonably high volume of customer interactions and which may have a number of
customer service staff), may need to opt for a CRM software tool that integrates a
range of B2C communication channels. Large to very large enterprises, depending on
their size and specific needs, may use a highly customisable CRM tool or
alternatively build their own in-house solution which may integrate different B2C
messaging channels both directly and indirectly (e.g., through CPaaS providers) and
employ (or outsource to) large teams of customer service staff.
5.1. CRM software solutions
(37) CRM software solutions help companies of various industry sectors manage their
customer interactions by organising, automating and synchronising data from various
sources, such as sales, marketing, customer database, customer service and technical
functions. CRM software solutions collate sets of data and display them in a user
friendly manner. This enables companies to improve customer relationships, to better
manage accounts, to enhance sales effectiveness, to optimise data quality and to
mitigate regulatory compliance risks.
11
5.1.1. Types of CRM software solutions
(38) CRM software solutions can offer different functionalities, such as (i) Customer
Service and Support; (ii) Sales, (iii) Marketing; and (iv) Digital Commerce.
12
CRM
software solutions can be on-premises or cloud-based (i.e., delivered as SaaS), and
can be customisable to the industry sector of the business customer. Kustomer, for
example, offers a SaaS CRM software tool specialised in customer service and
support.
5.1.2. Data held by CRM software solutions
(39) CRM software solutions specialised in customer service and support may hold a
variety of different types of data on behalf of their business customers, including for
example, the end customer’s identity, location, contact information and purchasing
history (conversions data) as well as pre-sale customer data (e.g., product queries,
website visits, preferences), post-sale customer data (e.g., returns, delivery queries)
and unstructured conversations data (i.e., that take place over one or more B2C
communication channels integrated into the CRM software tool) between a business
and their customers.
13
(40) The data that is stored in a CRM software solution (whether on-site or on a server of
the CRM software provider) is owned / under the control of the business using the
CRM software tool, who retains the decisions on the means and purposes of the
processing of the data. The CRM software provider will therefore not be able,
without prior agreement of / instruction from the business and if relevant the end-
customer, to use or sell any of this data. All CRM providers expressing an opinion
indeed confirmed that they do not sell data that is stored in their CRM software
10
The Notifying Party segments customers based on the following criteria: [Business Plans] – e.g., Form
CO, paragraph 2.2.9.
11
Commission decision of 6 December 2016 in case M.8124 – Microsoft/LinkedIn, paragraph 29.
12
These functionalities are the most relevant ones identified under the industry classification by Gartner.
13
Replies to questionnaire 1 to CRM software providers, question 60.
12
tool.
14
This does however not mean that this data cannot be shared with third parties
such as online ads service providers (through partnerships between CRM software
providers and online ads service providers or otherwise) with the agreement of the
business and, if relevant, the consent of the end-customer. The majority of CRM
providers expressing an opinion indicated that their CRM tool currently allows
business customers to share their data with a provider of online display advertising
services through their CRM software.
15
Moreover, all CRM providers expressing an
opinion indicated that business customers indeed share data from their CRM
software with online ads service providers
16
, although only a minority of customers
of CRM software responding to the market investigation indicated that they currently
do so.
17
(41) From the perspective of a business that makes use of CRM software, there is a
commercial interest to share data with an online ads service provider. The sharing of
data will allow that business to measure and optimise the performance of their ad
campaigns. Businesses may however decide to share their data through other
channels than their CRM software, for example through a point of sale (such as
Amazon or Shopify) or to share it themselves directly.
(42) The data that is stored in a CRM software tool (e.g., whether a particular customer
has already purchased, or expressed interest in purchasing, from a company) is also
valuable to online ads service providers. In particular, this data could be used to
improve ad targeting, and to measure “conversions” (i.e., the effectiveness of
particular ads campaigns in leading to more sales).
18
Moreover, with regard to Meta
(formerly Facebook) in particular, data is also used to [Information on how Meta
uses its data]. This involves the [Information on how Meta uses its data].
19
5.2. B2C communications
(43) Businesses and consumers communicate using a wide range of different
communication channels. Traditionally, these B2C communication channels included
telephone/voice, letters, fax, email and SMS. However, many alternative
communication channels have since emerged. These more modern channels include,
for example, form messages that consumers may fill in on the website of a business
(also called webchat) that can be included on a website or in a business’ app.
(44) Most notably, more modern B2C communication channels include different types of
messaging. Messaging channels can be cellular-based (i.e., making use of a mobile
telephone connection as is the case with SMS), OTT (as is the case with Meta
(formerly Facebook)’s channels such as WhatsApp), or a combination of both
cellular-based and OTT (such as Google Rich Communication Services (“RCS”) or
Apple Business Chat
20
). A wide range of B2C messaging channels exist today,
14
Replies to questionnaire 1 to CRM software providers, question 62.
15
Replies to questionnaire 7 to CRM software providers, question 51.
16
Replies to questionnaire 7 to CRM software providers, question 52
17
Replies to questionnaire 5 to business customers, question 16.2.
18
Form CO, paragraph 6.120.
19
Parties’ reply to RFI 21, question 11.
20
Both Google RCS and Apple Business Chat are directly available on mobile phones running
respectively on Android or iOS. These services are intended to replace SMS and add functionalities of
OTT messaging channels (i.e., to see when another user types or was last online, sending larger texts or
files etc.). Both Google RCS and Apple Business Chat make use of an internet connection. However,
unlike OTT messaging channels such as WhatsApp, no account or registration with the messaging
channel is required (other than the general account with which the user has signed up for the mobile
13
including: WhatsApp, Messenger, Instagram Messaging, Twitter, iMessage, Apple
Business Chat, Google RCS, WeChat, Telegram, Line, Signal, Talk, Viber, Kakao
Talk, IMO, etc.
(45) Many of these B2C messaging channels are available globally. However, the degree
to which consumers use these channels, i.e., the penetration rate of a given
messaging channel, differs per country.
5.2.1. B2C communication channels and CRM software
(46) A business’ chosen CRM software solution and the B2C communication channels it
uses are not automatically interoperable. A B2C communication channel can
however be integrated in all of a business’ systems, including its CRM software,
through an API.
(47) The API allows for the integration into the CRM software solution and enables
access to the communication channel by multiple agents/employees. The API is
generally provided by the provider of the B2C communication channel and APIs are
either sold or available as open source. A B2C communication service provider is not
required to make an API available as open source.
(48) Kustomer’s CRM software integrates with a range of B2C communication channels,
including telephone/voice, email, SMS, webchat, Messenger, Facebook WhatsApp,
Instagram (comments and direct messages), and Giphy, Twitter (tweets and direct
messages) and form messages. Other B2C communication channels that certain
CRM software providers integrate with include OTT messaging channels like Viber,
WeChat, Line and Telegram.
5.2.2. B2C communication channels, CPaaS solutions and CRM software solutions
(49) CPaaS providers are intermediaries that offer cloud-based middleware which allows
businesses and CRM software providers to integrate communication channels
(including voice, email, SMS, webchat and OTT messaging) into their applications
without needing to build backend infrastructure and interfaces. The CPaaS integrates
with the relevant B2C communication channel provider via APIs and provides its
customers with access to the relevant channel (or series of channels), for example via
a further API. CPaaS providers may for example integrate voice and SMS channels
from various network operators and package that into a single product, which can be
integrated into a CRM software tool. By way of example, a CPaaS provider may
provide access to voice, SMS, as well as various OTT messaging channels such as
Meta (formerly Facebook)’s WhatsApp. As outlined above, Kustomer integrates
Messenger, Instagram and WhatsApp messaging channels. [Information on
Kustomer’s means of integration of Messenger, Instagram Messaging and
Whatsapp].
21
(50) Business customers may procure a CRM software tool that already integrates certain
communication channels (e.g., via API access directly from B2C communication
service providers for certain channels (such as many OTT messaging channels,
which have open APIs) and via API access to CPaaS solutions for other channels
phone to run on the respective OS) and messages can be sent using a cellular connection in case there is
no internet connection. The additional functionalities of Google RCS and Apple Business Chat will, as
of today, not work for messages that are sent between iOS and Android devices. Rather, messages sent
between the two operating systems will be regular SMS messages without the additional functionalities.
21
Meta (formerly Facebook)’s WhatsApp for Business API is currently not open access [Business Plans].
See Form CO, paragraph 2.2.8.
14
(typically SMS and voice or OTT messaging channels with no open APIs)). It is also
possible for businesses to contract separately with a CPaaS provider (or even with a
B2C communication channel provider) and with a CRM software provider, and in a
second step arrange for the API to the B2C channels that it acquired either directly or
through CPaaS software to be integrated into their chosen CRM solution.
(51) Ultimately, as acknowledged by the Notifying Party, whether a business contracts
only with a CRM software provider, or separately with a CPaaS provider (or a B2C
communication channel provider) as well as with a CRM software provider, the B2C
channels are generally integrated in the chosen CRM software solution.
22
5.3. Online advertising sector
(52) Demand for ads services in the online advertising sector comes from businesses (i.e.,
advertisers) who wish to present ads to (potential) customers to drive a particular
audience response (such as the purchase of a certain product or service). The goal for
advertisers is to generate incremental sales and profits.
(53) Advertisers seeking to launch an ads campaign can choose between two types of
online ads:
(a) Online search ads: these are displayed on the basis of search queries entered by
users into internet search engines (i.e., advertisers can specify the keywords for
which they want their ads to be triggered or the queries for which they are most
likely to be relevant). Search ads are typically presented next to the search
results on the search engine’s own pages or other search results pages.
(b) Online display ads: these can be either contextual ads, displayed according to
the content of the page on which they appear, or non-contextual ads. Since no
query-keyword is usually available to trigger display ads, the data collected
about the user accessing the pages is relatively more important for display ads
than for search ads.
(54) On the supply-side of the online advertising sector, ads venues aim to attract and
retain advertisers by giving the opportunity to convey the advertiser’s message in an
appealing and engaging way to drive a customer response that generates incremental
sales and profits.
(55) Digital advertising space (or “inventory”) can be sold at a fixed price through direct
deals between the publishers (the “sell side”) and individual advertisers or media
agencies (the “buy side”). The matching between the two sides can also be made by
intermediaries. The advertising supply chain involves a diversified network of
intermediaries that provide technologies and/or data (“ad tech”) to facilitate the
programmatic sale and purchase of digital advertising inventory.
(56) Meta (formerly Facebook) provides online display advertising services at the various
levels of the ad tech value chain. It sells ads on its own properties (i.e., Facebook,
Instagram and Messenger) and also through Meta (formerly Facebook) Audience
Network (“MAN”). As regards the ads Meta (formerly Facebook) sells on its own
properties, it offers several types of advertising formats both on desktops and on
mobile devices. Advertising formats can include video advertising, non-video
advertising and click-to-message (“CTM”) advertising. CTM ads will send a
(potential) customer who clicks on a particular ad to a conversation with the
advertiser on Messenger, Instagram or WhatsApp.
22
Meta (formerly Facebook)’s Comments on State of Play Meeting, 22 July 2021, paragraph 5.1(B).
15
(57) Through MAN, Meta (formerly Facebook) matches advertisers seeking to deliver
their ads off Facebook to businesses looking to fill their unsold advertising space on
their mobile apps. On the demand side, Meta (formerly Facebook) enables
advertisers to buy advertising space through its Ads Manager. Meta (formerly
Facebook)’s Ads Manager allows a business among others to build ads, choose an
audience, set up ad budgets and following the ads campaign to monitor its
effectiveness.
23
(58) Advertisers who wish to run an ads campaign will go through the following
processes. First, a marketing objective will be defined. While ultimately all
advertising is designed to increase sales, objectives can also include the increase of
conversion (e.g., user action), consideration (e.g., user traffic) or awareness (e.g.,
user awareness of a brand). Second, the advertiser will have to select a target
audience. Through Meta (formerly Facebook)’s Ads Manager, advertisers can choose
an audience in three ways, through Core Audiences, Custom Audiences, and
Lookalike Audiences. “Core Audiences” allows businesses to define their target
audience on the basis of for example location, demographics, interests, consumer
behaviour and connections that a business may have on its Facebook page. “Custom
Audiences” allows businesses to connect with people that have already shown an
interest in their company (e.g., customers, people using a company app, or people
that have visited the company’s website). It allows businesses among others to
connect with “customers and contacts on Facebook using information from [their]
CRM system or email lists”.
24
Finally, “Lookalike Audiences” allows business to
create a source audience of people they know, and Meta (formerly Facebook) will
enable that the ads campaign reaches similar people with common interests and
traits. The third step for running an ads campaign is for the advertiser to select the
surfaces, format and content of the ads. Finally, the fourth step is for advertisers to
measure the performance of their ads campaign against the marketing objectives
pursued and the audiences that are targeted. Meta (formerly Facebook)’s Ads
Manager does allow businesses to do so.
25
(59) A key aspect of an advertising campaign is targeting. For an advertiser, better
targeting will help to achieve its marketing objective. For an online ads service
provider, the ability to provide better targeting will result in a more valuable ads
service. In this regard, the range of signals (i.e., data) that an online ads service
provider receives are important. Meta (formerly Facebook) receives a range of those
signals from both Facebook users and from third parties (such as advertisers) that
choose to share data with Meta (formerly Facebook). The data that Meta (formerly
Facebook) receives can be categorised as follows:
(a) On-site signals: these are signals gathered as a result of user behaviour on Meta
(formerly Facebook)’s services. These signals can include data on: (i) users’
connections, preferences, interests and activities; (ii) how users interact with
Meta (formerly Facebook)’s services (including interactions with ads delivered
on Meta (formerly Facebook)’s services); (iii) information relating to the
devices used by users to access Meta (formerly Facebook)’s services; and (iv)
the people, places or things users are connected to and interested in.
23
See https://en-gb.facebook.com/business/tools/ads-manager, last accessed on 7 October 2021 (ID:
2814).
24
See https://en-gb facebook.com/business/ads/ad-targeting?ref=fbb_adsmanager_products, last accessed
on 7 October 2021 (ID: 2817).
25
Parties’ reply to RFI 6, question 14(a).
16
(b) Off-site signals: these are signals that Meta (formerly Facebook) collects or
receives from third parties and relating to users’ behaviour on those third
parties’ products and services. Off-site signals include things such as: (i) how
users interact with third-party websites and apps (e.g., visits, purchases and ads
seen/interacted with); (ii) information relating to the devices used by users to
access third-party websites / apps; and (iii) data relating to the performance of
Meta (formerly Facebook)’s products (e.g., Meta (formerly Facebook)
Business Tools) that third parties have chosen to integrate on their websites /
apps.
26
(60) Specifically in relation to data stored on a business’ CRM software, this can be used
for the purposes of ads services in several ways. First, data may be used to help
businesses improve ad targeting efforts as set out above. Second, data may be used to
measure “conversions” and optimise the performance of a business’ ad campaign.
27
Third, data is also used to improve Meta (formerly Facebook)’s ad delivery systems
generally. General system improvement involves the improvement to predictions
around which users are likely to take the actions that are important to the relevant
advertiser.
28
Meta (formerly Facebook)’s improved ad delivery systems can then be
relied upon by any advertiser through Meta (formerly Facebook)’s Ads Manager. In
the Ads Manager, advertisers can choose how their ad campaign will be optimised to
address their marketing objective. For example, if an advertiser seeks to engage with
people, Meta (formerly Facebook) can optimise the ads campaign through the
following option: “Conversations – We’ll deliver your ads to people most likely to
have a conversation with you through messaging”.
29
6. RELEVANT MARKET DEFINITION
(61) For the assessment of the Transaction in this Decision, the following business
activities of the Parties are relevant: (1) Kustomer is active in the supply of CRM
software; (2) Meta (formerly Facebook) is active in the supply of B2C
communication channels, and (3) Meta (formerly Facebook) offers online advertising
services.
6.1. CRM software
(62) Kustomer is active in the market for the supply of CRM software. It provides a SaaS
CRM tool specialised in customer service and support. Meta (formerly Facebook) is
not active in this market.
30
26
Parties’ reply to RFI 6, question 14(b).
27
Form CO, paragraph 6.120.
28
Parties’ reply to RFI 21, question 11.
29
See https://en-gb facebook.com/business/help/416997652473726, last accessed on 8 October 2021 (ID:
2811).
30
For completeness, the Commission notes that Meta (formerly Facebook) offers Unified Inbox to
businesses with a Facebook business page or an Instagram business profile. It consolidates
communications between a business and consumers that occur via Facebook, Messenger and Instagram
(provided the business has linked its Facebook and Instagram accounts). It is designed to help
(generally small) businesses manage incoming customer enquiries on a small scale, and is typically
free. In the Form CO, Meta (formerly Facebook) submits that Unified Inbox is not a CRM software tool
and [Business Plans]. See Form CO, paragraphs 6.57 and following.
17
6.1.1. Product market definition
6.1.1.1. Commission precedents
(63) The Commission previously considered the market for the supply of CRM software
solutions in Oracle/Siebel, where it identified CRM as a distinct product market
within the overall category of enterprise application software (“EAS”).
31
In that case,
the Commission found that CRM software solutions are a separate product market as
opposed to other software categories (or “pillars”) within EAS. The Commission left
open the question whether the market for CRM software solutions could be further
segmented on the basis of (i) functionality of the software, such as Customer Service
and Support, Sales, Marketing, or Digital Commerce; (ii) industry sector in which the
customer is active; (iii) mode of deployment, that is to say installation on the
premises or host-based deployment (i.e., SaaS); or (iv) customisation, i.e., custom-
built solutions as opposed to standardised software.
32
(64) In IMS/Cegedim Business, the Commission found that the market investigation did
not give clear indications as to whether a separate CRM software solutions market
should have been identified for the industry relevant in that case, that is to say the
pharmaceutical sector, but ultimately left open the question whether the market
should be further segmented by industry, functionality, mode of deployment or
customisation.
33
(65) In Microsoft/LinkedIn, the Commission considered that CRM software solutions may
likely be further distinguished on the basis of functionality, in particular as customers
viewed the different functionalities as complementary, rather than substitutable, and
there appeared to be also limited scope for supply-side substitution by CRM
providers. However, for the purposes of that decision the exact product market
definition for CRM software solutions was left open. Similarly, the Commission
considered but left open whether the market may be segmented by mode of
deployment (on-premises v. cloud based), customer size or based on the type of
industry of the customers.
34
6.1.1.2. The Notifying Party’s view
(66) The Notifying Party submits that there is a single market for the supply of CRM
software.
35
(67) The Notifying Party considers that there is no reason to segment the market for the
supply of CRM software on the basis of functionality. In particular, the Notifying
Party submits that there is no distinct market for the provision of customer service
and support CRM software. According to the Notifying Party, customers of CRM
software procure varying functionality demands and most suppliers of CRM software
provide most or all of the relevant CRM functionalities or offer the ability for
customers to mix and match functionalities with different CRM providers.
36
31
Commission decision of 22 December 2005 in case M.3978 – Oracle/Siebel, paragraphs 11-16.
32
Commission decision of 22 December 2005 in case M.3978 – Oracle/Siebel, paragraph 16.
33
Commission decision of 19 December 2014 in case M.7337 – IMS Health/Cegedim Business,
paragraphs 91-92. See also Commission decision of 27 November 2007 in case M.4944 – SAP/Business
Objects, paragraph 7; Commission decision of 4 January 2008 in case M.4987 – IBM/Cognos,
paragraphs 7-9; Commission decision of 20 July 2010 in case M.5904 – SAP/Sybase, paragraph 21;
Commission decision of 29 August 2014 in case M.7334 – Oracle/Micros, paragraph 9.
34
Commission decision of 6 December 2016 in case M.8124 – Microsoft/LinkedIn, paragraphs 49-50.
35
Form CO, paragraph 6.14.
36
Form CO, paragraph 6.15.
18
(68) The Notifying Party also considers that a segmentation between large enterprises and
small and medium enterprises is not required as the different needs of businesses can
be considered as part of the competitive assessment. Nevertheless, the Notifying
Party submits that very small businesses which do not have the scale or resources to
justify the investment in dedicated CRM software fall outside of the scope of the
market.
37
(69) In the Article 6(1)(c) Response, the Notifying Party argues that the Commission
arrives at narrow segmentations of the market for the supply of CRM software based
on a functional characteristics approach and without applying the substitutability test
set out in the Commission Notice on the definition of relevant markets for the
purposes of Community competition law (the “Market Definition Notice”).
38
In this
regard, the Notifying Party submits that products may differ in their characteristics,
price or use, but nevertheless belong to the same relevant market if they are regarded
as substitutable for customers and the Commission does not properly consider
demand-side substitutability.
39
However, the Notifying Party considers that, since the
Transaction does not raise competitive concerns under any plausible market
definition, the exact scope of the product markets for the supply of CRM software
can be left open.
40
6.1.1.3. Commission’s assessment
(70) With respect of a possible distinction of CRM software solutions based on
functionality, the market investigation indicated that there may be a separate market
for the CRM software functionality of customer service and support, which also
corresponds to a distinction used by Kustomer itself
41
, as well as in third party
reports, such as Gartner.
42
From a demand-side perspective, a strong majority of
responding market participants (both CRM software providers and business
customers) considered that CRM software with different functionalities are
complementary rather than substitutable
43
and that customer service and support
CRM software is the only type of CRM software functionality that is substitutable
for Kustomer’s CRM software product.
44
The majority of CRM providers that
expressed a view considered it would not be technically possible (or at best only to a
limited extent), nor would it be likely from a commercial perspective, for business
customers to switch to other CRM software functionalities (i.e., other than customer
service and support CRM software) in a scenario where the price of customer service
and support CRM software increased.
45
From a supply-side perspective, while larger
CRM software providers may offer a range of different functionalities, smaller
players (such as Kustomer) may only offer one or a limited number of
37
Form CO, paragraph 6.14.
38
OJ C 372, 9.12.1997.
39
Article 6(1)(c) Response, paragraphs 2.2 and 2.3.
40
Form CO, paragraph 6.13 and Article 6(1)(c) Response, paragraph 2.1.
41
E.g., Form CO, paragraph 2.2.8 and PowerPoint presentation session 2 “Facebook’s proposed
acquisition of Kustomer - Kustomer and the CRM Market” of the site visit held on 22 September 2021,
slide 7.
42
Gartner data is used to calculate market shares in the Form CO for both the broader CRM software
market and the narrower market for customer service and support CRM software.
43
Replies to questionnaire 1 to CRM software providers, question 10; replies to questionnaire 4 to
business customers, question 10.
44
Replies to questionnaire 1 to CRM software providers, question 11; replies to questionnaire 4 to
business customers, question 9.
45
Replies to questionnaire 7 to CRM software providers, question 11.
19
functionalities. There was no clear consensus among CRM providers on the ability
for a company specialised in other functionalities to start offering customer service
and support CRM software in the short term and without incurring significant
investment costs. While in the Phase I market investigation, a majority of
respondents that expressed a view said that this would not be possible in the short
term and without incurring significant investment costs
46
, in the Phase II market
investigation, the results were mixed, with several respondents indicating that it
would depend on the CRM software provider’s existing capabilities (e.g., existing
technology and know-how).
47
In light of the above, the Commission concludes that,
from a demand side perspective, there is a separate market for customer service and
support CRM software. To the extent the market would be wider than customer
service and support CRM software, this would thus only be due to supply-side
substitutability arguments (i.e., a CRM software provider offering another
functionality starting to offer customer service and support CRM software).
However, the Commission considers that any CRM software provider active in
another segment wanting to start offering customer service and support CRM
software would require B2C communication channels as an input, as customer
service and support CRM by definition requires interaction with customers, and that
the mere fact that some CRM software providers offer multiple functionalities does
not establish full supply-side substitutability if starting to offer a different
functionality cannot necessarily be done in the short term and without incurring
significant investment costs.
(71) For the purposes of this Decision, the market definition can be left open in this
regard as the Commission considers that it is likely, and there is even a strong
probability, that the Transaction would significantly impede effective competition in
both the overall market for CRM software and the market for customer service and
support CRM software.
(72) As regards a possible distinction depending on the mode of deployment, namely
cloud-based/SaaS or on-premises, the results of the market investigation were mixed
from a demand side perspective. There was no consensus between business
customers that expressed a view on whether on-premises and cloud-based CRM
software solutions are substitutable for one another.
48
From a supply-side
perspective, the majority of respondents that expressed a view considered that it
would not be possible for a CRM software provider specialised in on-premises
solutions to start offering cloud-based CRM solutions in the short term and without
incurring significant investment costs.
49
However, the market investigation also
indicated that a majority of CRM software customers use cloud-based solutions, and
most CRM software providers only or primarily offer cloud-based solutions.
50
One
market participant indicated that “companies that specialized in on-premises
solutions have or are moving towards offering cloud-based solutions”
51
and a
leading CRM software provider confirmed that it “offers only cloud-based
solutions.”
52
The same is true of Kustomer. Indeed, the majority of respondents that
46
Replies to questionnaire 1 to CRM software providers, question 9.
47
Replies to questionnaire 7 to CRM software providers, question 9.
48
Replies to questionnaire 4 to business customers, question 14.
49
Replies to questionnaire 1 to CRM software providers, question 15.
50
Replies to questionnaire 1 to CRM software providers, question 14; replies to questionnaire 4 to
business customers, question 12.
51
Reply to questionnaire 1 to CRM software providers, question 14.1 (ID: 759).
52
Reply to questionnaire 1 to CRM software providers, question 14.1 (ID: 846).
20
expressed a view also indicated that cloud-based CRM software solutions are either
progressively replacing on-premises solutions, or have almost entirely done so.
53
In
light of the mixed market investigation results and the fact that the majority of the
market already appears to be cloud-based/SaaS, the Commission considers that it is
not necessary to segment the broader CRM software market or the narrower
customer service and support CRM software market between SaaS/cloud based
solutions on the one hand and on-premises solutions on the other hand. However, the
precise segmentation can be left open as the Commission considers that it is likely,
and there is even a strong probability, that the Transaction would significantly
impede effective competition in both an overall market for customer service and
support CRM software and a potential segment for cloud-based customer service and
support CRM software solutions (as well as on the broader CRM software solutions
market and a potential segment for cloud-based CRM software solutions).
(73) As regards a possible segmentation based on business customer size, the vast
majority of business customers that responded to the market investigation indicated
that such a distinction was justified based on the needs and resources of a particular
business (i.e., between small, medium and large businesses)
54
55
, with one such
respondent noting that “a small company would rather take an easy-to-use cloud-
based solution with a per license cost model [whereas] on-premises or customized
solutions require substantial investments and their benefits would only outweigh
costs in bigger companies that can afford the necessary manpower and investment
for customization.”
56
A majority of CRM software providers that expressed a view
also supported a segmentation based on business size, on similar grounds.
57
In this
regard, the majority of CRM software providers that expressed a view not only
indicated that there are material differences in the types of CRM software solutions
or the manner in which they are sold depending on whether the CRM software
solution is being sold to small, medium-size or large business customers, but also
confirmed that their own offering is specifically designed for, or primarily targets, a
particular size of business customer (either primarily large enterprises or small and
medium-sized business customers).
58
One such respondent explained that “small
business customers require CRM solutions that that are easy to set up and use,
require little administrative overhead, and are able to cover a lot of CRM
functionality (…). They expect to buy through a fixed-fee model (...). Medium-sized
business customers require CRM solution that they can implement quickly while
allowing them to scale usage as they might need due to growth or seasonal business
conditions. (…) Large business customer prioritize solutions that they are able to
customize to match their existing business workflow and processes. (…) Large
business customers expect to purchase through tenders or individual negotiations,
and similar to medium-sized business customers, typically require a longer sales and
implementation process.”
59
53
Replies to questionnaire 1 to CRM software providers, question 15; replies to questionnaire 4 to
business customers, question 17.
54
Small business (less than 8 employees), medium-size business (between 8 and 500 employees) and
large business (more than 500 employees).
55
Replies to questionnaire 4 to business customers, questions 16 and 16.1.
56
Reply to questionnaire 4 to business customers, questions 16 and 16.1 (ID: 776).
57
Replies to questionnaire 1 to CRM software providers, questions 18 and 19.
58
Replies to questionnaire 7 to CRM software providers, questions 12 and 13.
59
Reply to questionnaire 7 to CRM software providers, question 12.1 (ID: 1713).
21
(74) Moreover, the majority of CRM providers that expressed a view indicated that a
company specialised in CRM software solutions for small and medium-sized
business (“SMB”) customers would not be able to start providing CRM software
services to large enterprises in the short term without incurring significant
investments, while several others noted that this would depend on the company’s
existing capacities (e.g., human and financial resources or relevant integrations) and
the customer’s use case.
60
Respondents that expressed a view confirmed that such a
segmentation by business size is equally applicable to customer service and support
CRM software.
61
Although the majority of CRM providers that expressed a view
considered it would be technically possible for SMBs to switch to customer service
and support CRM software designed for large enterprises in the scenario where the
price of customer service and support CRM software designed for SMBs increased,
they also considered that such a switch would be unlikely, from a commercial
perspective.
62
Internal documents and submissions of the Parties indicate that
Kustomer is primarily focused on [Type of customers that Kustomer focuses on].
63
Accordingly, and in particular based on the consistent responses to the market
investigation and Kustomer’s own primary focus on[Type of customers that
Kustomer focuses on], the Commission considers that there may be grounds to
further segment the relevant broader market for CRM software and the narrower
market for customer service and support CRM software based on business size.
(75) However, there was no clear consensus expressed on the precise segmentation to be
applied as between small medium and large businesses. For the purposes of the
present Decision, the market definition can be left open in this regard as the
Commission considers that it is likely and there is even a strong probability that the
Transaction would significantly impede effective competition in an overall market
for customer service and support CRM software and potential segments for customer
service and support CRM software based on business size (as well as on the broader
CRM software solutions market and potential segments of such a market based on
business size). Nevertheless, the segmentation based on business size is taken into
consideration as part of the Commission’s competitive assessment in Section 7 in
what regards Kustomer’s close competitors.
(76) With respect to a potential segmentation according to the industry sector in which the
business customer is active, again, the majority of market participants that expressed
a view indicated that such a segmentation was warranted from a demand-side
perspective, including in the case of customer service and support CRM software.
64
The reasons include differences in regulation and policy (e.g., financial, health or
retail services), different needs (e.g., public vs private sector) and different levels of
data sensitivity (e.g., healthcare sector) according to the sector. However, there was
no clear consensus between CRM providers that expressed a view on whether it
would be technically possible for business customers in the e-commerce sector
(where many of Kustomer’s business customers are active) to switch to customer
service and support CRM software designed for business customers in other
60
Replies to questionnaire 7 to CRM software providers, question 14.
61
Replies to questionnaire 4 to business customers, question 17.
62
Replies to questionnaire 7 to CRM software providers, question 15.
63
These customers are referred to as [Customer Information]. See Form CO, paragraph 2.2.9 Table 2.4,
which shows that the [Customer Information] of Kustomer’s customers belong to the [Customer
Information] segment.
64
Replies to questionnaire 1 to CRM software providers, question 20; replies to questionnaire 4 to
business customers, questions 18 and 19.
22
industries in the scenario where the price of the former increased, while the majority
considered the likelihood of such switch, from a commercial perspective, would
depend on the customer.
65
Nevertheless, one respondent indicated that “(…) typically
not eCommerce customers, (Traditional Retail), as well as other service centric
organizations will have radically different requirements and workflows. As a result
customer service functionality that is designed to support e-Commerce, will not likely
meet the needs of other businesses, such as Travel, Government, Utilities, High-Tech
Manufacturing, etc.”
66
From a supply side, the majority of CRM software providers
that responded to the market investigation indicated that they are capable of
providing CRM software to customers active in different sectors, with one
respondent pointing out that “the core CRM functionalities are generally the same
across industries”
67
and another indicating that “by and large, CRM providers don't
tend to focus on specific industries to the exclusion of others, and beyond [creating
industry-specific veneers], most underlying CRM products remain fairly similar.”
68
There was also no clear consensus expressed on the precise segmentation to be
applied as between different industry sectors of CRM software customers, including
whether there was a distinct market for CRM software for customers active in the e-
commerce sector.
69
As a result, the Commission considers that there may be grounds
to further segment the relevant broader market for CRM software and narrower
market for customer service and support CRM software based on the industry sector
of the business customer.
(77) For the purposes of this Decision, the market definition can be left open in this
regard as the Commission considers that it is likely, and there is even a strong
probability, that the Transaction would significantly impede effective competition in
an overall market for customer service and support CRM software and potential
segments for customer service and support CRM software based on the industry
sector of the business customer (as well as on the broader CRM software solutions
market and potential segments based on the industry sector of the business
customer).
(78) Finally, the Commission sought to understand whether CPaaS solutions should be
included within the same market as CRM software, or as customer service and
support CRM software. The majority of market participants that expressed a view
considered that CPaaS solutions are not in the same market as CRM software (or
customer service and support CRM software) solutions, as they are different in terms
of use, product characteristics and price.
70
Several market participants highlighted
that they are used for different purposes, and that CRM solutions tend to use CPaaS
solutions to complement their service offering.
71
One CRM software provider
indicated that “while CPaaS typically focuses on facilitating the communication
between customers and an organization, managing the customers’ records and
composing the customers’ experience typically lies within the CRM. CPaaS and
65
Replies to questionnaire 7 to CRM software providers, questions 19 and 20.1.
66
Reply to questionnaire 7 to CRM software providers, question 19.1.1 (ID: 1655).
67
Reply to questionnaire 1 to CRM software providers, question 20.1 (ID: 846).
68
Reply to questionnaire 1 to CRM software providers, questions 20 and 20.1 (ID: 726).
69
Replies to questionnaire 1 to CRM software providers, question 20.1.
70
Replies to questionnaire 1 to CRM software providers, questions 23 and 24; replies to questionnaire 2 to
CPaaS providers, questions 9 and 10; replies to questionnaire 4 to business customers, questions 20
and 21.
71
Replies to questionnaire 1 to CRM software providers, question 23.1; replies to questionnaire 2 to
CPaaS providers, question 10.1; replies to questionnaire 4 to business customers, question 20.1.
23
CRM are more complementary than comparable.”
72
A CPaaS provider indicated that
“Kustomer, being a CRM, covers a broader scope than a CPaaS platform [and] it
consumes [via API access] CpaaS platforms in the communication space (i.e. SMS,
Voice over MNO connections, other OTT communication channels like Apple
Business Chat, Google Business Messages, etc.).
73
The Notifying Party also does not
argue that CPaaS should be included within a broader market for CRM software or a
narrower market for customer service and support CRM software. Accordingly, in
particular because of the consistent responses of market participants active in the
supply of CRM software and CPaaS solutions, the Commission concludes that
CPaaS solutions do not form part of the narrower market for customer service and
support CRM software, or the broader CRM software market.
6.1.1.4. Conclusion
(79) In light of the above, for the purpose of this Decision, the Commission concludes
that the question whether there is a separate market for the supply of customer
service and support CRM software or whether it is part of the overall market for the
supply of CRM software can be left open, and that CPaaS solutions do not form part
of either of these markets. The question whether the market for the supply of
customer service and support CRM software (or the broader market for CRM
software overall) should be further segmented based on (i) mode of deployment, (ii)
business customer size, and (iii) industry sector in which business customers are
active can be left open in this Decision, as the Commission considers that it is likely,
and that there is even a strong probability, that the Transaction would significantly
impede effective competition in the market for customer service and support CRM
software (or the broader market for CRM software overall) irrespective of such
potential segmentation. The Commission will carry out its competitive assessment on
that basis.
6.1.2. Geographic market definition
6.1.2.1. Commission precedents
(80) In Oracle/Siebel, the Commission noted that “trade patterns of CRM solutions do not
vary to any significant extent across different geographic regions”, that there were
“no indications that CRM Solutions would be made either specific for the EEA
region or specific to any other region”, and that local customisation is made solely
for “language reason, or through customised add-ons to meet specific local
needs.”
74
However, the Commission ultimately left open whether the market could
be defined as worldwide or EEA-wide, since the transaction did not give rise to
competitive concerns.
75
The same conclusion was reached in IBM/Cognos.
76
(81) In IMS/Cegedim Business, the Commission noted that the market was likely EEA-
wide in scope, if not broader: the market investigation in that case did not clearly
indicate that there were marked regulatory, technical or linguistic differences in the
supply of CRM software among Member States. The Commission ultimately left the
72
Reply to questionnaire 1 to CRM software providers, question 23.1 (ID: 759).
73
Reply to questionnaire 2 to CPaaS providers, questions 10.1 (ID: 738).
74
Commission decision of 22 December 2005 in case M.3978 – Oracle/Siebel, paragraph 18.
75
Commission decision of 22 December 2005 in case M.3978 – Oracle/Siebel, paragraph 19.
76
Commission decision of 4 January 2008 in case M.4987 – IBM/Cognos, paragraph 16.
24
exact geographic market definition open, as the transaction did not raise concerns
irrespective of the precise geographic scope of the market.
77
(82) In Microsoft/LinkedIn, the Commission likewise considered that the market for CRM
software solutions overall and possible narrower markets or segments of that market
to be EEA-wide, if not worldwide, in scope but ultimately left the precise scope of
the geographic market open.
78
6.1.2.2. The Notifying Party’s view
(83) In line with the Commission’s decisional practice, the Notifying Party submits that
the geographic scope of the market for CRM software is at least EEA-wide, if not
worldwide. The Notifying Party submits that (i) CRM software does not vary
materially across different regions, and (ii) that there are no material regulatory or
technical differences or local software customisation when CRM software is
provided in different regions (other than data requirements in certain regions that the
software be hosted locally). As regards linguistic differences, the Notifying Party
points out that most CRM software offers support for multiple languages and
embedded translation technology, and that there are no significant regional or local
CRM providers.
79
6.1.2.3. Commission’s assessment
(84) In the present case and based on the results of the market investigation, the
Commission considers that the geographic scope of the market for CRM software
solutions overall, or any segmentation thereof, including the narrower market for
customer service and support CRM software solutions, is at least EEA-wide, if not
worldwide, in scope. Indeed, market participants considered that the geographic
scope is at least EEA wide, if not global. In this regard, respondents to the market
investigation among CRM providers unanimously agreed that the geographic market
for CRM software solutions is worldwide in scope
80
and the vast majority of
business customers that responded to the market investigation likewise indicated that
the geographic scope of the broader CRM software market is EEA-wide or
worldwide in scope.
81
One CRM software provider indicated that they “do not see
any significant differences geographically since it is a global market and companies
are looking to solve similar challenges. Companies that serve customers worldwide
design their software/solutions to work universally regardless of the geography. Any
geographical differences are configured within the solution (i.e. selecting the
language to use, date/time format to use, workflows for a region/area, etc.)”.
82
6.1.2.4. Conclusion
(85) In light of the above, for the purpose of this Decision, the Commission considers that
the geographic scope of the market for CRM software solutions overall, or any
segmentation thereof, including the narrower market for customer service and
support CRM software solutions, is at least EEA-wide, if not worldwide.
77
Commission decision of 19 December 2014 in case M.7337 – IMS Health/Cegedim Business,
paragraph 96.
78
Commission decision of 6 December 2016 in case M.8124 – Microsoft/LinkedIn, paragraph 56.
79
Form CO, paragraph 6.17.
80
Replies to questionnaire 1 to CRM software providers, question 36; replies to questionnaire 7 to CRM
software providers, question 30.
81
Replies to questionnaire 4 to business customers, question 49.
82
Reply to questionnaire 1 to CRM software providers, question 36.2 (ID: 759).
25
6.2. B2C communication services
(86) Meta (formerly Facebook) provides a number of messaging services that can be used
by businesses to communicate with consumers (and vice versa), namely Messenger,
Instagram and WhatsApp. Kustomer offers a live webchat functionality to businesses
using its software, which allows those businesses to communicate with consumers
via webchat (e.g., on the businesses’ website).
6.2.1. Product market definition
6.2.1.1. Commission precedents
(87) In its decisional practice, the Commission never specifically considered the market
definition for B2C communication services. However, the Commission did
previously assess a number of related markets.
(88) In particular, the Commission previously considered that there are distinct markets
for (i) consumer communication services and (ii) enterprise communication services
and that, within a market for consumer communication services, a market for
consumer communication apps for smartphones can be distinguished.
(89) In Microsoft/Skype, the Commission concluded that consumer communication
services and enterprise communication services differ in terms of prices and features,
while there are only a limited number of providers of communications applications
designed for both consumers and enterprises.
83
As a consequence, the Commission
concluded that consumer communication services and enterprise communication
services form two distinct markets.
84
(90) As regards the market for consumer communication services, the Commission
assessed whether the market could be further segmented by functionality (such as
instant messaging, voice and video calls), by platform (including, among others,
personal computers, smartphones and tablets) or by operating system. However, the
Commission ultimately left open the exact product market definition.
85
(91) As regards the market for enterprise communication services, the Commission
assessed whether any of the segmentations mentioned above should be taken into
account, as well as a possible segmentation by size of customers. However, also as
regards the market for enterprise communication services, the Commission
ultimately left open the exact product market definition.
86
(92) In Microsoft/Nokia, the Commission further assessed the market for customer
communication services and considered that there are indications that there may be a
specific product market for consumer communications apps further divided by
platform.
87
However, the Commission ultimately left the exact scope of the relevant
market for consumer communication apps open.
(93) In Facebook/WhatsApp, the Commission considered it appropriate to assess a market
for consumer communications apps for smartphones.
88
Furthermore, the Commission
considered whether this market should be broadened to include traditional electronic
communication services, such as voice calls, SMS, MMS and email. While the
83
Commission decision of 7 October 2011 in case M.6281 – Microsoft/Skype, paragraphs 13-16.
84
Commission decision of 7 October 2011 in case M.6281 – Microsoft/Skype, paragraph 17.
85
Commission decision of 7 October 2011 in case M.6281 – Microsoft/Skype, paragraphs 20-43.
86
Commission decision of 7 October 2011 in case M.6281 – Microsoft/Skype, paragraphs 49-63.
87
Commission decision of 4 December 2013 in case M.7047 – Microsoft/Nokia, paragraphs 37-45.
88
Commission decision of 3 October 2014 in case M.7217 – Facebook/WhatsApp, paragraph 21.
26
market investigation indicated that these traditional channels and consumer
communications apps are substitutable, it also pointed to the following elements
which suggest imperfect substitutability or complementarity. First, while both type of
services are used for the same purpose, consumer communication apps are richer in
terms of functionalities (e.g., seeing when another user was last online, whether
another user is typing etc.).
89
Second, there is a difference in pricing conditions as
consumer communications apps are mainly offered for free.
90
Third, telecoms
operators responding to the market investigation indicated that the competitive
interaction between traditional services and consumer communications apps is only
one-way (i.e., consumer communications apps constrain electronic communication
services but not the other way around).
91
However, the Commission ultimately left
the exact scope of the relevant market for consumer communication apps on
smartphones open.
92
(94) In addition to consumer communication services, the Commission previously
considered whether a distinct relevant market for Application to Person (“A2P”)
SMS services exists. A2P SMS is an enterprise to customer messaging service for the
types of messages that companies, such as airlines and banks, exchange with their
customers.
(95) In Syniverse/MACH, the Commission considered that the ultimate customers of A2P
SMS services are companies while, for Person to Person SMS services the ultimate
customers are mobile network operators. For this reason, and because of the fact that
the specific services are materially different, the Commission considered in that case
that those two activities constitute separate product markets but ultimately decided
there was no need to take a conclusive view on the exact product market definition.
93
6.2.1.2. The Notifying Party’s view
(96) The Notifying Party considers that the relevant product market includes all B2C
messaging channels and should not be further segmented by either (i) B2C written
communication channels, (ii) B2C messaging channels, (iii) OTT B2C messaging
channels or (iv) device.
94
(97) The Notifying Party submits that from the perspective of businesses, the primary
factor in choosing which B2C channels to use is their ability to reach consumers and
that many B2C channels offer a set of communication channels (almost always
including phone and/or e-mail, and increasingly webchat). From the perspective of
consumers, the Notifying Party considers they have access to and use multiple
different channels as well as devices to contact businesses. According to the
Notifying Party, the relative convenience or functionality of different communication
channels is not sufficient to consider particular channels as distinct relevant
markets.
95
(98) The Notifying Party also does not consider it appropriate to segment the market
based on the B2C procurement channel (e.g., direct vs. indirect via B2C providers,
via CPaaS providers, or via CRM software providers). According to the Notifying
89
Commission decision of 3 October 2014 in case M.7217 – Facebook/WhatsApp, paragraph 30.
90
Commission decision of 3 October 2014 in case M.7217 – Facebook/WhatsApp, paragraph 31.
91
Commission decision of 3 October 2014 in case M.7217 – Facebook/WhatsApp, paragraph 32.
92
Commission decision of 3 October 2014 in case M.7217 – Facebook/WhatsApp, paragraph 34.
93
Commission decision of 29 May 2013 in case M.6690 – Syniverse/MACH, paragraphs 152-154.
94
Form CO, paragraphs 6.23 and 6.24.
95
Form CO, paragraph 6.24.
27
Party, the communication channels offer the same functionality to businesses
whether they are procured directly or indirectly and many suppliers enable direct and
indirect access to other channels.
96
(99) Finally, the Notifying Party does not consider it appropriate to distinguish between
A2P and non-application to person communications. The Notifying Party considers
there is no consistent definition of A2P communication and submits that while its
messaging channels can be used by businesses for both automated and non-
automated messaging, the Notifying Party does not supply A2P solutions.
97
(100) In the Article 6(1)(c) Response, the Notifying Party argues that in the Article 6(1)(c)
Decision, the Commission arrived at narrow segmentations of the market for B2C
communications based on a functional characteristics approach and without applying
the substitutability test set out in the Market Definition Notice.
98
In this regard, the
Notifying Party submits that products may differ in their characteristics, price or use
but nevertheless belong to the same relevant market if they are regarded as
substitutable for customers and the Commission does not properly consider demand-
side substitutability.
99
(101) In particular, in relation to the substitutability between cellular-based and OTT B2C
messaging channels, the Notifying Party considers the Commission does not address
whether businesses and consumers consider that cellular-based and OTT B2C
messaging services are substitutes.
100
Furthermore, the Notifying Party considers that
the Commission does not take into consideration that: (i) there are CRM providers
that have chosen to support a limited number (or only one) of B2C channels
101
; (ii)
different B2C channels score comparably in terms of user preference
102
; and (iii)
most consumers will have access to many, if not all, of the range of different B2C
channels and it is common for consumers to switch to different channels on a regular
basis, including to contact the same business.
103
(102) However, the Notifying Party considers that, since the Transaction does not raise
competitive concerns under any plausible market definition, the exact scope of the
product markets for B2C communication services can be left open.
104
6.2.1.3. Commission’s assessment
(103) The Commission did not previously consider the market for B2C communication
services. In the market investigation, the Commission therefore first assessed
whether respondents considered different B2C communication channels to be
substitutable.
(104) A majority of CRM software providers expressing an opinion indicated that APIs for
different B2C communication channels are not substitutable.
105
This is not
contradicted by a majority of CPaaS providers expressing an opinion, who indicated
96
Form CO, paragraph 6.24.
97
Form CO, paragraphs 6.25 and 6.26.
98
Article 6(1)(c) Response, paragraph 2.2.
99
Article 6(1)(c) Response, paragraphs 2.2 and 2.3.
100
Article 6(1)(c) Response, paragraph 2.3.
101
The Parties cite Helpshift as appearing to integrate only its in-app/web chat.
102
The Parties cite a 2018 Salesforce survey (“State of the Connected Customer, Insights from 6,700
consumers and business buyers”, second edition, 2018, page 47) and [Business Plans].
103
Article 6(1)(c) Response, paragraph 2.4.
104
Form CO, paragraph 6.22 and Article 6(1)(c) Response, paragraph 2.1.
105
Replies to questionnaire 1 to CRM software providers, question 31.
28
that substitutability depends on the situation while no respondent directly stated that
APIs for different B2C communication channels are substitutable.
106
From the
business customer’s perspective, the market investigation results support that B2C
communication channels are generally complementary as the vast majority of
responding customers indicate that B2C communication channels are complementary
to each other.
107
(105) More specifically with regard to the substitutability of individual B2C
communication channels, a majority of B2C communication service providers
indicated that Meta (formerly Facebook)’s channels (WhatsApp, Messenger and
Instagram) could be considered as substitutable with other OTT messaging channels,
rich cellular-based B2C messaging channels (such as Google RCS and Apple
Business Chat), SMS and webchat.
108
However, only a minority of responding CRM
software providers, CPaaS providers or customers of Meta (formerly Facebook) and
Kustomer indicated that there was a single overall category of B2C communication
channels in which Meta (formerly Facebook)’s channels were substitutable with
other such channels.
109
(106) Moreover, a vast majority of respondents confirm that the use of B2C
communication channels is generational.
110
In this regard, more traditional B2C
communication channels, including in particular telephone/voice and email, are
regarded as complementary from this perspective.
111
One respondent clarified that
“[t]hese services are old time services, and as we understand their use is dropping
and younger generations prefer new methods of communication.”
112
The
Commission notes that the Salesforce survey results from 2018 that were submitted
by the Notifying Party are in line with these findings. [Survey data on customers’
preferred channels].
113
(107) Considering that only B2C communication service providers indicate that, from a
supply side perspective, other types of B2C messaging channels are substitutable
with Meta (formerly Facebook)’s messaging channels while from a demand side
perspective other respondents indicate that different B2C communication channels
are complementary to each other, the Commission considers that a further
segmentation of the overall market for B2C communication services is appropriate.
(108) The Commission investigated five potential segmentations of an overall market for
B2C communication services: (i) synchronous versus asynchronous B2C
communication services; (ii) cellular-based B2C messaging services versus OTT
B2C messaging services; (iii) rich cellular based B2C messaging services versus
OTT B2C messaging services; (iv) webchat versus OTT B2C messaging services;
106
Replies to questionnaire 2 to CPaaS providers, question 18.
107
Replies to questionnaire 4 to business customers, question 25.
108
Replies to questionnaire 3 to B2C communication service providers and online ads service providers,
question 13.
109
Replies to questionnaire 1 to CRM software providers, question 34; replies to questionnaire 2 to CPaaS
providers, question 21; replies to questionnaire 4 to business customers, question 26.
110
Replies to questionnaire 3 to B2C communication service providers and online ads service providers,
question 15; replies to questionnaire 4 to business customers, question 27.
111
Replies to questionnaire 3 to B2C communication service providers and online ads service providers,
question 14; replies to questionnaire 4 to business customers, question 25.
112
Reply to questionnaire 3 to B2C communication service providers and online ads service providers,
question 14.1 (ID: 741).
113
Kustomer’s internal document, […]. The Commission has not verified the underlying data or the
methodology of this survey and does not use this information to draw any conclusions.
29
and (v) B2C messaging services available on desktops versus services available on
mobile devices.
(109) As regards a potential segmentation between synchronous and asynchronous B2C
communication services, that is to say a segmentation between phone/voice and other
B2C communication channels, the vast majority of respondents indicated that both
from the business’ perspective (i.e., the business using CRM software) and the end
customer’s perspective (i.e., the customers of businesses using CRM software) these
differ in terms of use, price or characteristics.
114
One respondent indicates that
“[t]here are different use cases for synchronous and asynchronous B2C
communication based on the urgency of this communication and internal company
structures.”
115
Respondents moreover indicate that both from the business’
perspective as well as from the end-customer’s perspective there are price differences
between asynchronous and synchronous B2C communication channels. Specifically,
respondents indicated that there are significant cost differences for a business
customer between offering synchronous and asynchronous B2C channels, among
others because employees can handle several asynchronous conversations at one time
as a result of which less customer service employees are required.
116
Finally, within a
potential market for asynchronous B2C communication services a small majority of
respondents indicated that there might be a separate market segmentation for quasi-
synchronous B2C communication services (e.g., B2C messaging channels or
webchat), where a relatively quick response is expected by a customer, versus
letter/fax/email, where a slower response would be expected.
117
Overall, a majority
of the respondents indicate that separate markets for synchronous and asynchronous
B2C communication services can be distinguished within the overall market of B2C
communication services.
118
(110) The Commission considers that, both for businesses and end customers, there are
significant differences between synchronous and asynchronous B2C communication
services. For businesses, there are significant costs involved as customer service
agents can only handle one call at a time. For customers, there are often costs
involved in calling, while they might also need to queue before speaking to a
representative of a business. The results of the market investigation do not indicate as
clearly whether a separate segmentation for quasi-synchronous B2C communication
services (including both cellular based and OTT B2C messaging services as well as
webchat) exists within all asynchronous B2C communication services. As mentioned
before however, other asynchronous B2C communication services such as email are
considered to be complementary and not substitutable to quasi synchronous B2C
communication services.
119
The Commission considers that a distinction could
potentially be made between synchronous and asynchronous B2C communication
114
Replies to questionnaire 3 to B2C communication service providers and online ads service providers,
questions 16 and 18; replies to questionnaire 4 to business customers, question 28.
115
Reply to questionnaire 4 to business customers, question 28.1 (ID: 747).
116
Replies to questionnaire 3 to B2C communication service providers and online ads service providers,
question 20; replies to questionnaire 4 to business customers, question 30.
117
Replies to questionnaire 3 to B2C communication service providers and online ads service providers,
question 17.
118
Replies to questionnaire 3 to B2C communication service providers and online ads service providers,
question 22; replies to questionnaire 4 to business customers, question 32.
119
Replies to questionnaire 1 to CRM software providers, question 34; replies to questionnaire 2 to CPaaS
providers, question 21; replies to questionnaire 4 to business customers, question 26.
30
services, without it being necessary to take a view on the existence of a narrower
segment for quasi-synchronous B2C communication services.
(111) Within a potential market for asynchronous B2C communication services, the
Commission investigated a potential segmentation between cellular-based B2C
messaging services and OTT B2C messaging services. Cellular-based B2C
messaging services concern in particular SMS, while OTT B2C messaging services
could include internet-based messaging services such as Meta (formerly Facebook)’s
messaging channels but also webchat. While the results of the market investigation
are mixed as regards the classification of webchat, at least some respondents
considered webchat to be substitutable with Meta (formerly Facebook)’s messaging
channels.
120
Moreover, webchat appears to be offering at least some of the same
functionalities of B2C messaging services and it is provided OTT. In considering a
potential segmentation between cellular-based and OTT B2C messaging services, the
Commission therefore considers that webchat is included in the latter segmentation.
(112) As to the differences between cellular-based and OTT B2C messaging services, a
majority of responding B2C communication service providers indicate that both from
the business’ and the end customer’s perspective these differ in terms of use, price or
characteristics.
121
All responding customers of Meta (formerly Facebook) and
Kustomer expressing an opinion agree that there are differences from their
perspective.
122
One respondent states that “[u]sers are increasingly using OTT
messaging for their day to day conversations, and SMS is used more for utilities.
SMS is sometimes associated [sic] with costs, especially in the international context,
while OTT is usually free for the user.”
123
From a business’ perspective, a respondent
observed that “Business customers are able to leverage a much deeper, and richer
set of features with OTT messaging services that are not available in the cellular-
based B2C messaging services. […]”.
124
(113) The Notifying Party submitted that some CRM software providers have chosen to
support a limited number of B2C messaging channels, which would suggest
substitutability. In this regard, the Notifying Party only provided an example of one
company, Helpshift, which exclusively integrates in-app / web chat.
125
The
Commission notes that Helpshift states itself that it is the only mobile-first customer
service platform and that it specifically targets businesses that make use of their own
app.
126
As such, this example is not sufficient to support the claim that the entire
cellular-based and OTT B2C messaging segments would be substitutable, contrary to
what the results of the market investigation reveal. Another argument the Notifying
Party brings forward is that a Salesforce survey would show that B2C channels score
comparably in terms of user preference.
127
However, the Commission notes that a
more recent version of this survey states that [Survey data on customers’ preferred
120
Replies to questionnaire 4 to business customers, question 26.
121
Replies to questionnaire 3 to B2C communication service providers and online ads service providers,
questions 23 and 25.
122
Replies to questionnaire 4 to business customers, question 33.
123
Reply to questionnaire 3 to B2C communication service providers and online ads service providers,
questions 23.1 (ID: 741).
124
Reply to questionnaire 3 to B2C communication service providers and online ads service providers,
questions 25.1 (ID: 723).
125
Article 6(1)(c) Response, paragraph 2.4.
126
See: https//www.helpshift.com, last accessed on 13 October 2021 (ID: 2799).
127
Article 6(1)(c) Response, paragraph 2.4 and footnote 36: Salesforce, “State of the Connected Customer,
Insights from 6,700 consumers and business buyers”, second edition, 2018, page 47.
31
channels] and, more specifically on substitutability between cellular-based and OTT
B2C messaging segments, [Survey data on customers’ preferred channels].
128
Finally, the Notifying Party submitted that most consumers will have access to many
different B2C channels and switch between these different channels even within the
context of a single transaction.
129
In this context, the Commission notes that while
consumers might indeed use multiple channels to start and complete a transaction,
[Survey data on customers’ preferred channels].
130
Overall, the Commission
considers therefore that the arguments brought forward by the Notifying Party do not
contradict the results from the market investigation, wherein only a minority of CRM
and CPaaS providers as well as business customers indicate that a single category of
communication channels (including SMS) is substitutable with Meta (formerly
Facebook)’s channels.
131
(114) The Commission further investigated the likelihood for business customers to switch.
In the scenario where the cost of using OTT B2C messaging services would increase
permanently by 5%, most of responding customers indicate that while it would be
technically possible to switch to using SMS they could only do so in part.
132
One
respondent indicates this is the case for three reasons: (i) because of Wi-Fi and cross-
country use of an OTT B2C messaging channel it would be difficult to switch to
SMS only; (ii) because horizontal communications through for example a help desk
could not be supported by SMS as this requires more than one person to access ticket
information; and (iii) because of data transfer and compliance concerns.
133
From a
commercial perspective, a majority of the responding customers indicate moreover
they would not or only to a limited extent be able to switch to SMS.
134
Several
respondents indicate that this has to do with a lack of capabilities for SMS, stating
for example that OTT B2C messaging channels have “the ability to support multi-
media communications (videos and not just text) and the ability to be viewed on
multiple devices at any point in time”.
135
Other respondents point to the character
limitation on SMS and the fact that they might not maintain a record of phone
numbers of all customers that would be required for communications via SMS.
136
The results of the market investigation are inconclusive as to the question whether
there are significant costs involved for a business to offer cellular-based B2C
messaging channels.
137
However, responding B2C communication service providers
indicate that, if there were no significant costs involved, businesses would be able to
switch relatively quickly from offering cellular-based B2C messaging services to
OTT services.
138
128
Kustomer’s internal document, […]. The Commission has not verified the underlying data or the
methodology of this survey and does not use this information to draw any conclusions.
129
Article 6(1)(c) Response, paragraph 2.4.
130
Kustomer’s internal document, […].
131
Replies to questionnaire 1 to CRM software providers, question 34; replies to questionnaire 2 to CPaaS
providers, question 21; replies to questionnaire 4 to business customers, question 26.
132
Replies to questionnaire 5 to business customers, question 3.1.
133
Reply to questionnaire 5 to business customers, question 3.1.1 (ID: 1577).
134
Replies to questionnaire 5 to business customers, question 3.2.
135
Reply to questionnaire 5 to business customers, question 3.2.1 (ID: 1496).
136
Replies to questionnaire 5 to business customers, question 3.2.1.
137
Replies to questionnaire 3 to B2C communication service providers and online ads service providers,
questions 27; replies to questionnaire 4 to business customers, question 35.
138
Replies to questionnaire 3 to B2C communication service providers and online ads service providers,
questions 28.
32
(115) Overall, however, a majority of respondents to the market investigation indicated that
separate markets for cellular-based B2C messaging and OTT B2C messaging
services can be distinguished within a potential market for asynchronous B2C
communication services. Therefore, the Commission considers that a distinction
could potentially be made between a market for cellular based B2C messaging
services and a market for OTT B2C messaging services.
(116) The Commission then considered whether, aside from the distinction between
cellular-based and OTT B2C messaging services, a separate market for rich cellular-
based B2C messaging channels could be distinguished (in particular Google RCS
and Apple Business Chat) or whether these services should be considered to be part
of the market for OTT B2C messaging services. As set out in footnote 20 above,
these channels are directly available on mobile phones and are intended to replace
SMS by adding functionalities of OTT messaging channels. However, unlike OTT
messaging channels, they do not require an account or registration (other than the
general account with which the user has signed up for the mobile phone to run on the
respective OS) and messages can be sent using a cellular connection. For example,
messages sent between iOS and Android devices would currently be sent as regular
SMS messages without the additional functionalities of OTT messaging channels.
(117) Most responding customers indicate that it would be technically possible to switch to
using a rich cellular-based B2C messaging service in the case where the cost of using
OTT B2C messaging services for communicating with end customers would increase
permanently by 5%. However, this would only be possible partly.
139
A majority of
responding customers indicate that it would not be possible from a commercial
perspective to switch to rich cellular-based B2C messaging services or only to a
limited extent.
140
One respondent points out that Google RCS and Apple Business
Chat are not available everywhere
141
, and another states that “contrary to SMS, RCS
is a new technology that is not yet fully deployed and the effectiveness of the RCS
channel is not yet fully demonstrated”.
142
In this regard, another respondent states
that they “don’t want to be switching channels and then need to “recruit” consumers
to those channels. We will follow where consumers will engage.”
143
The observations
regarding the size and roll-out of rich cellular-based B2C messaging channels are
mirrored in the fact that only 8% of the responding customers indicate that they used
either Google RCS or Apple Business Chat in the past year.
144
(118) The Commission considers that it can be left open whether a separate market for rich
cellular-based B2C messaging services should be distinguished, in particular because
of the limited presence of those services at present.
(119) The Commission further considered whether, within a market for OTT B2C
messaging services, a distinction could be made between OTT B2C messaging
services and webchat. In the scenario where the cost of using OTT B2C messaging
services would increase permanently by 5%, most of responding customers indicated
that while it would be technically possible to switch to using webchat they could
only do so in part.
145
One respondent states that while it is technically possible to
139
Replies to questionnaire 5 to business customers, question 5.1.
140
Replies to questionnaire 5 to business customers, question 5.2.
141
Reply to questionnaire 5 to business customers, question 5.2.1 (ID: 1584).
142
Reply to questionnaire 5 to business customers, question 5.2.1 (ID: 1484).
143
Reply to questionnaire 5 to business customers, question 5.2.1 (ID: 1571).
144
Replies to questionnaire 5 to business customers, question 2.
145
Replies to questionnaire 5 to business customers, question 7.1.
33
switch, “webchat is mainly customers within desktop browser, while OTT B2C lives
largely on mobile”.
146
Another respondent explains further that “webchat can be used
to engage with customers and non-customers, it is a reactive bi-lateral
communications channel – it cannot be used to launch campaigns or engage
proactively with potential customers”.
147
From a commercial perspective, most of the
responding customers indicate that they would not or only to a limited extent be able
to switch to webchat although several respondents indicate they would be partly or
completely able to switch.
148
As a result, the Commission considers that for the
purpose of this investigation it can be left open whether a distinction should be made
between OTT B2C messaging services and webchat.
(120) The Commission also considered whether the market for OTT B2C messaging
services included public comments on social media platforms, such as on Facebook
or on Instagram. Such public comments are integrated in Kustomer’s CRM software
tool, and can be used as a means of communicating with business, e.g., where such
business has a Facebook or Instagram page, or posts an advertisement on such a
platform. While a majority of market participants considered Facebook and
Instagram public comments to be OTT B2C messaging channels, and moreover, that
it was important for the purposes of customer service and support for a business to be
able to respond to such public comments, e.g., via a private OTT messaging
channel,
149
some of the more substantiated responses from market participants were
more nuanced.
150
A leading CRM software provider noted that “from a consumer
standpoint, they may not see a difference between Facebook and Facebook
Messenger (i.e. they see it as “Facebook”)”
151
while another CRM software provider
indicated that “public pages could be considered a common entry point for a B2C
conversation … with the business following up with the user via a private or direct
message to begin a conversation”.
152
A business customer indicated that “Facebook
and Instagram public comments are normally one time communications and not
channels to have a long-term conversation (possibly due to being public)”
153
Twitter,
the largest OTT B2C communication service provider after Meta (formerly
Facebook) indicated that they “would not consider that comments are the same as
OTT messaging channels. One has an expectation that messaging is a one to one
exchange and a private communication”
154
while a competitor of Meta (formerly)
Facebook indicated that “it is conceivably possible, but likely rare, that there may be
B2C communication through a public post.”
155
For the purposes of this Decision, the
Commission considers that it can be left open whether the market for OTT B2C
communication channels includes public posts or comments on social media
platforms such as Facebook and Instagram, in particular since, in the case of
Facebook and Instagram public comments, the associated private OTT B2C
messaging channels, Messenger and Instagram messaging respectively, each include
a feature which enables a business to read a public comment and respond to it via the
146
Reply to questionnaire 5 to business customers, question 7.1.1 (ID: 1533).
147
Reply to questionnaire 5 to business customers, question 7.1.1 (ID: 1496).
148
Replies to questionnaire 5 to business customers, question 7.2.
149
Replies to questionnaire 8 (Commitments Market Test), question 17.2.
150
Replies to questionnaire 8 (Commitments Market Test), question 17.1.
151
Reply to questionnaire 8 (Commitments Market Test), question 17.1.1 (ID: 2543).
152
Reply to questionnaire 8 (Commitments Market Test), question 17.1.1 (ID: 2634).
153
Reply to questionnaire 8 (Commitments Market Test), question 17.1.1 (ID: 2540).
154
Reply to questionnaire 8 (Commitments Market Test), question 17.1.1 (ID: 2628).
155
Reply to questionnaire 8 (Commitments Market Test), question 17.1.1 (ID: 2562).
34
relevant private channel.
156
Accordingly, whether or not public comments on social
media platforms are included within the market for OTT B2C messaging channels
will not impact the competitive assessment.
(121) Finally, with respect to a potential segmentation between OTT B2C messaging
services available on desktops or mobile devices, the results of the market
investigation indicate that such a segmentation of a potential market for OTT B2C
messaging services would not be warranted despite the fact that the services might
differ. Customers of Meta (formerly Facebook) and Kustomer indicate that
messaging services available on desktops and mobile devices differ in terms of use,
price or characteristics.
157
Respondents clarify that the messaging experience on
mobile devices is better, that mobile devices are more convenient for consumers and
that the user base for mobile devices is different.
158
However, a majority of
responding B2C communication channel providers and online ads service providers
do not recognise these differences.
159
Also as regards the question whether B2C
messaging services available on desktops or mobile devices are interchangeable, the
results of the market investigation are inconclusive. A majority of respondents do
consider that there are no significant costs involved for a business to offer B2C
messaging services on either a desktop or on mobile devices.
160
Overall, a majority
of respondents to the market investigation considered that no separate markets for
B2C messaging services on desktops or on mobile devices could be identified within
a broader potential market for asynchronous B2C communication services.
161
(122) As to a potential distinction between OTT B2C messaging services available on
desktops or mobile devices, the Commission notes that many if not all OTT B2C
messaging services are available both on desktops and mobile devices. For example,
all of Meta (formerly Facebook)’s messaging services are available both on desktops
and mobile phones and the same applies to other large OTT B2C messaging services
(such as webchat, Twitter, WeChat, Signal, Telegram, Twitter and Viber). The
Commission considers, for the purpose of this Decision, that the potential segments
for OTT B2C messaging services available respectively on desktops and on mobile
devices are part of the same overall market for OTT B2C messaging services. While
the messaging experience might differ between desktops and mobile devices, there
are no differences in terms of costs. Because of this, and on the basis of the results of
the market investigation, the Commission considers, for the purpose of this Decision,
that the exact product market definition can be left open.
6.2.1.4. Conclusion
(123) In light of the above, the Commission concludes that a broad overall product market
for B2C communication services exists. For the purpose of this Decision, the
Commission considers that, within this overall market for the supply of B2C
communication services, a separate intermediate product market for asynchronous
B2C communication services can be considered, while for the purpose of this
156
Parties’ reply to RFI 21, Annex 1.
157
Replies to questionnaire 4 to business customers, question 37.
158
Replies to questionnaire 4 to business customers, question 37.1 (IDs: 768, 776 and 812).
159
Replies to questionnaire 3 to B2C communication service providers and online ads service providers,
questions 30 and 32.
160
Replies to questionnaire 3 to B2C communication service providers and online ads service providers,
question 34; replies to questionnaire 4 to business customers, question 39.
161
Replies to questionnaire 3 to B2C communication service providers and online ads service providers,
question 35; replies to questionnaire 4 to business customers, question 40.
35
Decision it can be left open whether a further segmentation between quasi-
synchronous and asynchronous B2C communication services could be distinguished.
Within an intermediate market for asynchronous B2C communication services, the
Commission considers, for the purpose this Decision, that a separate narrower
product market for OTT B2C messaging services can be considered, which may
include rich-cellular-based B2C messaging services, webchat and public comments
on social media platforms. The Commission will carry out its competitive assessment
on the basis of a market for B2C communication services, while taking into account,
within that market, both the plausible intermediate market for asynchronous B2C
communication services as well as the plausible narrower market for OTT B2C
messaging services.
6.2.2. Geographic market definition
6.2.2.1. Commission precedents
(124) As regards the markets for consumer communication services and enterprise
communication services, the Commission considered previously in Microsoft/Skype,
that the geographic scope of both markets is at least EEA-wide, but ultimately left
open the exact geographic scope of the market.
162
The market investigation results
indicated that a majority of respondents do not consider there are particular legal or
technical barriers that restrict the use or trade of communication services worldwide
and that services are generally perceived as not different in terms of quality, price
and features worldwide (with the exception of some geographic areas).
163
The
Commission further concluded that, as regards consumer communication services,
suppliers provide their services on a global level with limited differentiation in the
quality and features in the EEA (and to a lesser extent worldwide).
164
(125) In Microsoft/Nokia, the Commission concluded that the geographic scope of the
market for consumer communication services is at least EEA-wide, if not worldwide,
but ultimately left open the exact scope of the geographic market.
165
(126) In Facebook/WhatsApp, the results of the market investigation revealed no major
differences in the offering of consumer communications apps across the world.
166
However, given the different degree of penetration of consumer communications
apps in certain world regions (i.e., WhatsApp being widespread in the EEA, but not
so much in the United States), the Commission considered that this is an indication
of different competitive dynamics, which could be explained by different consumer
preferences, regulatory environments and marketing efforts.
167
The Commission
further investigated the possibility of a narrower geographic definition within the
EEA. In this regard, in that case, the Commission concluded that while the degree of
penetration might differ, the apps of Meta (formerly Facebook) (e.g., Messenger and
WhatsApp) tended to be the number 1 and 2 in most large Member States. Further,
while WhatsApp was not free in every Member State, the Commission considered
this was not sufficient to justify the definition of national markets. Finally, the
Commission considered that the main factors relevant for the assessment of the
transaction (including closeness of competition, customer switching and entry
162
Commission decision of 7 October 2011 in case M.6281 – Microsoft/Skype, paragraphs 67-68.
163
Commission decision of 7 October 2011 in case M.6281 – Microsoft/Skype, paragraph 65.
164
Commission decision of 7 October 2011 in case M.6281 – Microsoft/Skype, paragraph 66.
165
Commission decision of 4 December 2013 in case M.7047 – Microsoft/Nokia, paragraph 81.
166
Commission decision of 3 October 2014 in case M.7217 – Facebook/WhatsApp, paragraph 37.
167
Commission decision of 3 October 2014 in case M.7217 – Facebook/WhatsApp, paragraph 38.
36
barriers) do not differ according to the Member State concerned.
168
The Commission
concluded in light of this that the geographic scope of the market for consumer
communication apps is at least EEA-wide, if not worldwide.
(127) Finally, as regards the geographic scope of the market for A2P SMS services, the
Commission considered in Syniverse/MACH that the market could have a larger
scope than national or regional, possibly worldwide. However, ultimately the
Commission left the geographic scope of the market for A2P SMS services
open.
169
6.2.2.2. The Notifying Party’s view
(128) The Notifying Party submits that the geographic scope of the market for B2C
communications is worldwide.
170
(129) The Notifying Party submits that (i) many B2C communication channels offer their
services on a global basis (e.g., email and particularly OTT channels), (ii) many
businesses procure B2C communication services on a global basis to allow
communications with consumers to occur cross-border (in particular email and
messaging channels), (iii) the functionalities of B2C communication services do not
vary materially across different regions and there are no material regulatory or
technical differences between particular regions or countries.
171
(130) As regards linguistic differences, the Notifying Party considers that while the
language of conversations will vary geographically, most B2C communication
channels offer support in multiple languages. Further, the Notifying Party submits
that such language differences do not determine the varying use or prevalence of
different messaging channels across countries.
172
6.2.2.3. Commission’s assessment
(131) The results of the market investigation do not differ significantly from the findings in
Facebook/WhatsApp. All responding CRM providers indicate that they obtain API
access to B2C communication channels at a worldwide level.
173
The large majority
of responding B2C communication service providers indicate moreover that they
make available the APIs for the B2C communication channels on a worldwide
level.
174
Finally, the large majority of responding customers of Meta (formerly
Facebook) and Kustomer indicate that they procure the B2C communication
channels they use on a worldwide level.
175
(132) Moreover, respondents did not raise the existence of any particular legal or technical
barriers that restrict the use or trade of B2C communication channels.
176
(133) Similarly as in Facebook/WhatsApp, the Commission finds that the degree of
penetration, at least for Meta (formerly Facebook)’s B2C messaging channels, differs
per EEA Member State and that, overall, the degree of penetration of WhatsApp,
168
Commission decision of 3 October 2014 in case M.7217 – Facebook/WhatsApp, paragraphs 41-43.
169
Commission decision of 29 May 2013 in case M.6690 – Syniverse/MACH, paragraph 183.
170
Form CO, paragraphs 6.27 and 6.29.
171
Form CO, paragraph 6.27 (i)-(iii).
172
Form CO, paragraph 6.27 (iv).
173
Replies to questionnaire 1 to CRM software providers, question 37.
174
Replies to questionnaire 3 to B2C communication service providers and online ads service providers,
question 44.
175
Replies to questionnaire 4 to business customers, question 51.
176
Replies to questionnaire 3 to B2C communication service providers and online ads service providers,
question 44.2.
37
Messenger and Instagram on an EEA-level is higher than the degree of penetration
worldwide.
177
However, the Commission considers that the main factors that might
be relevant for the assessment of the Transaction do not differ significantly according
to the Member State concerned.
6.2.2.4. Conclusion
(134) In light of the above, for the purpose of this Decision, the Commission considers that
the geographic scope of the market for B2C communication services, or any
segmentation thereof, is at least EEA-wide, if not worldwide
6.3. Online Display Advertising
(135) Meta (formerly Facebook) is active in the market for online advertising services, and
provides these services almost entirely on its own social networks as well as almost
entirely in the form of online display advertising. Meta (formerly Facebook) provides
advertising service for both mobile and desktop, as well as in both video and non-
video formats. Kustomer is not active in online advertising.
6.3.1. Product market definition
6.3.1.1. Commission precedents
(136) In previous decisions, the Commission considered the market for online advertising
to be separate from offline advertising. It also considered possible further
segmentations between online search and non-search (i.e., display) advertising, on
the basis of the platform (PC versus mobile), between online display ads on social
networks or off social networks, or between online display video and non-video
advertising.
178
Following the Google AdSense decision,
179
the Commission
concluded in Google/Fitbit that there were separate relevant product markets for the
supply of online search advertising services (and potential segments) on the one
hand, and the supply of online display advertising services (and potential segments)
on the other hand.
180
6.3.1.2. The Notifying Party’s view
(137) The Notifying Party considers that the relevant market includes all ads services
(i.e., including both online and offline advertising services).
181
(138) The Notifying Party submits that the various forms of media (e.g., news,
entertainment, games and social media) across a variety of channels (e.g., print, TV,
online and offline) all compete to attract and retain user engagement and advertisers.
According to the Notifying Party, advertisers allocate their budget across different
venues in order to maximise the additional sales and profits that they may generate
and where one advertising venue becomes more expensive, they will switch towards
another one where they can get a better return on investment, regardless of the media
or channel involved. Moreover, there is a tendency for advertising format and venues
177
Parties’ reply to RFI 6, question 22.
178
Commission decision of 3 October 2014 in case M.7217 – Facebook/WhatsApp, paragraphs 74-79;
Commission decision of 6 December 2016 in case M.8124 – Microsoft/LinkedIn, paragraphs 159-161;
Commission decision of 6 September 2018 in case M.8788 – Apple/Shazam, paragraphs 133-135;
Commission decision of 17 December 2020 in case M.9660 – Google/Fitbit, paragraphs 148-155.
179
Commission decision of 20 March 2019 in case AT.40411 – Google AdSense, paragraphs 121-183.
180
Commission decision of 17 December 2020 in case M.9660 – Google/Fitbit, paragraph 155.
181
Form CO, paragraph 6.34.
38
to converge due to its rapid technological progress (e.g., TV has made improvements
in the targeting and measurement options offered to advertisers).
182
(139) The Notifying Party also considers it is not appropriate to segment the market by
(i) specific channel (e.g., search vs. display ads), (ii) on-social network/off-social
network context, (iii) advertising format (video/non-video ads), or (iv) device
(desktop vs. mobile).
183
(140) As regards search vs. display ads, the Notifying Party submits that both forms of
advertising share characteristics in terms of format, buying, targeting functionalities
and market objectives.
(141) As regards on/off-social media networks, the Notifying Party considers that ads can
reach the same consumers on-social and off-social networks. In addition, the same
ads buying, targeting functionalities and marketing objectives are available in both
channels.
(142) As regards video/non-video, according to the Notifying Party, almost all advertising
channels offer a variety of ads formats, including image and video-based ads, which
are generally sold and purchased in the same way as other digital advertising.
(143) As regards desktop vs. mobile devices, the Notifying Party submits that all online
advertising channels offer ads on both desktops and mobile devices. Moreover, the
same formats, buying, targeting functionalities and marketing objectives are
available across both devices.
(144) The Notifying Party also submits that while CTM ads have distinct features, they are
substitutable with other ad formats and channels for the same aforementioned
reasons.
184
(145) In the Article 6(1)(c) Response, the Notifying Party argues that the Commission
arrives at narrow segmentations of the market for online display advertising services
based on a functional characteristics approach and without applying the
substitutability test as set out in the Market Definition Notice. In this regard, the
Notifying Party submits that market shares alone are not sufficient to determine
whether an undertaking is dominant without further evidence or analysis.
185
(146) However, the Notifying Party considers that, since the Transaction does not raise
competitive concerns under any plausible market definition, the exact scope of the
product markets for advertising services can be left open.
186
6.3.1.3. Commission’s assessment
(147) Contrary to the Notifying Party’s views that an overall market for ads services
including both online and offline advertising exists, the Commission considers in line
with recent precedents that a separate market for online ads services exists and that
within this market, a distinction can be made between online search and online
display advertising services. Several participants explained that search advertising
and display advertising had different use cases and purposes, noting that “there are a
number of differences between search-based online advertising and non-search
182
Form CO, paragraphs 6.35 and 6.36.
183
Form CO, paragraphs 6.37 and 6.38.
184
Form CO, paragraph 6.39.
185
Article 6(1)(c) Response, paragraph 2.3.
186
Form CO, paragraph 6.41.
39
online advertising”
187
and “search advertising is really focused on the bottom of the
funnel (intent and conversion). Display advertising can serve the entire funnel
(awareness, consideration, intent, and conversion)”.
188
Further, a majority of
respondents that indicated a view (including both competitors and business
customers) confirmed a difference in terms of targeting capabilities, user
engagement, return on investment, reach, price/cost, planning/marketing objectives,
ad format, design and content and ad tech services required.
189
(148) The Commission further investigated the likelihood for business customers to switch
from purchasing online display advertising services to online search advertising
services in case of a permanent price increase by 5%. The results of the market
investigation are mixed whether a switch to online search advertising services would
be technically possible. While a majority of the customers expressing an opinion
indicate that switching is entirely or partly possible from a technical perspective,
more than a third of those respondents indicate that this is not the case.
190
A majority
of online ads service providers expressing an opinion indicate that while switching
would be possible from a technical perspective
191
, this would not be possible or only
to a limited extent from a commercial perspective
192
. One respondent states that
“[…] as display ads rely on known characteristics of the users it can be more ideal
for targeted advertising which are general and not action specific. 5% increase
won’t be significant to cause the drastic change […]”.
193
The majority of customers
expressing an opinion confirm that from a commercial perspective they would not
switch to online search advertising services in this scenario.
194
One respondent states
in this regard that “from a commercial perspective this would imply to lose one
commercial channel, and therefore to lose market reach and customer leads”
195
,
whereas another respondent observes that “the two channels play different roles with
different performance metrics”
196
. Several respondents further indicate that a 5%
permanent price increase would not be sufficient for them to switch.
197
(149) As a result, the Commission maintains its view, in line with recent precedents, that a
distinction should be made between the market for online search advertising services
and the market for online display advertising services.
(150) Within a market for online display advertising services, the Commission investigated
a potential segmentation between online display advertising on and off social
networks. A majority of both competitors and business customers expressing an
opinion stated there are differences in terms of targeting capabilities, user
187
Reply to questionnaire 3 to B2C communications service providers and online ads service providers,
question 36 (ID: 864).
188
Reply to questionnaire 3 to B2C communications service providers and online ads service providers,
question 36 (ID: 870).
189
Replies to questionnaire 3 to B2C communication service providers and online ads service providers,
question 37; replies to questionnaire 4 to business customers, question 42.
190
Replies to questionnaire 5 to business customers, question 8.1.
191
Replies to questionnaire 6 to B2C communication service providers and online ads service providers,
question 11.1.
192
Replies to questionnaire 6 to B2C communication service providers and online ads service providers,
question 11.3.
193
Reply to questionnaire 6 to B2C communication service providers and online ads service providers,
question 11.4 (ID: 1704).
194
Replies to questionnaire 5 to business customers, question 8.2.
195
Reply to questionnaire 5 to business customers, question 8.2.1 (ID: 1493).
196
Reply to questionnaire 5 to business customers, question 8.2.1 (ID: 1571).
197
Replies to questionnaire 5 to business customers, question 8.2.1 (IDs: 1339, 1496 and 1571).
40
engagement, ad format, design and content and ad tech services required.
198
Respondents noted that “advertising on social networks has specific requirements for
format, content, and creative, which apply to the networks policies”
199
and “social
network[s] reach a greater number of people, and [achieve] more engagement from
those people”.
200
Furthermore, the Commission notes that third party industry reports
consider ads on social networks and off social networks separately, suggesting that
for advertisers these may be separate markets or segments and thus it may be
informative to have separate information in that respect when deciding how to spend
their advertising budget. This is in particular the case for the reports compiled by
IDC, a global provider of market intelligence that develops industry standards,
conducts research, and provides legal support for the online advertising industry.
201
(151) In the scenario where the price of online display advertising services on social
networks would permanently increase by 5%, a majority of both competitors and
business customers indicate that it would be technically possible for businesses to
switch to online display advertising off social networks.
202
A majority of both
competitors and business customers expressing an opinion indicate however that
business customers would not, or only to a limited extent, be able from a commercial
perspective to switch to online display advertising off social networks.
203
Responding competitors point in particular to the fact that advertising on social
networks can sometimes be more targeted
204
and state that “[d]ue to the different
audiences (advertise where your customers are) – the use of social networks is so
prevalent as part of a cohesive marketing strategy, it would be highly difficult to
completely move away from social network advertising into online (non social)
display advertising”
205
. Similar responses have been given by business customers, in
particular stating that there is “[…] a difference in the reach and the quality of the
audience for both prospecting and retargeting campaigns. Besides that the targeting
options will be limited (e.g. no interests targeting since it’s only available on social
networks”.
206
(152) The Commission considers that, for the purposes of this Decision, it can be left open
whether a separate market for online display advertising services on social networks
should be distinguished, as this would not change the outcome of the competitive
assessment in the present case.
(153) The Commission further investigated whether, within the market for online display
advertising services, different segments should be distinguished for online video
display advertising services and online non-video display advertising services. A
majority of respondents expressing an opinion (including both competitors and
198
Replies to questionnaire 3 to B2C communication service providers and online ads service providers,
question 39; replies to questionnaire 4 to business customers, questions 44.
199
Reply to questionnaire 3 to B2C communication service providers and online ads service providers,
question 38 (ID: 723).
200
Reply to questionnaire 4 to business customers, questions 43 (ID: 717).
201
Parties’ reply to RFI 4, question 1a.
202
Replies to questionnaire 6 to B2C communication service providers and online ads service providers,
question 12.1; replies to questionnaire 5 to business customers, question 9.1.
203
Replies to questionnaire 6 to B2C communication service providers and online ads service providers,
question 12.3; replies to questionnaire 5 to business customers, question 9.2.
204
Reply to questionnaire 6 to B2C communication service providers and online ads service providers,
question 11.4 (ID: 1704).
205
Reply to questionnaire 6 to B2C communication service providers and online ads service providers,
question 11.4 (ID: 1658).
206
Reply to questionnaire 5 to business customers, question 9.2.1 (ID: 1369).
41
business customers) confirmed a difference in terms of user engagement, return on
investment, price/cost, planning/marketing objective, ad format, design and content
and ad tech services required.
207
Respondents noted that “online video advertising
[…] offers the advertiser more opportunities for creativity, expression, and
engagement”
208
and “video ads, in particular, may be used when an advertiser wants
to employ a format better suited for storytelling, informing, and/or educating”.
209
Furthermore, similar to online display advertising on and off social networks, the
Commission notes that third party industry reports consider video and non-video ads
separately.
210
(154) However, the Commission considers that for the purposes of this Decision, it can be
left open whether a distinction should be made between online video and non-video
display advertising services, as this would not change the outcome of the competitive
assessment in the present case.
(155) Finally, the Commission investigated whether, within the market for online display
advertising services, different segments should be distinguished for online display
advertising services on mobile devices and desktops. A majority of respondents
expressing an opinion (including both competitors and business customers)
confirmed that a difference exists in terms of targeting capabilities, user engagement,
possibility to monitor effectiveness, return on investment, reach and ad format,
design and content.
211
A respondent noted that “the primary differences between
online display advertising services on desktops and online display advertising
services on mobile apps are that mobile apps may present advertisers with the
potential for greater reach, engaging full-screen experiences, and additional formats
available for mobile apps”.
212
Furthermore, similar to online display advertising on
and off social networks, the Commission notes that third party industry reports
consider ads on mobile and desktop separately.
213
(156) However, the Commission considers that for the purposes of this Decision, it can be
left open whether a distinction should be made between online display advertising
services on mobile devices and desktops, as this would not change the outcome of
the competitive assessment in the present case.
6.3.1.4. Conclusion
(157) In light of the above, for the purpose this Decision, the Commission considers that
the relevant product market is the market for the supply of online display advertising
services. The question whether the market for the supply of online display
advertising services should be further segmented by (i) on-/off-social networks, (ii)
video/non-video ads, and (iii) mobile/desktop can be left open in this Decision, as
this would not change the outcome of the competitive assessment in the present case.
207
Replies to questionnaire 3 to B2C communication service providers and online ads service providers
and questionnaire 4 to business customers, questions 41 and 46 respectively.
208
Reply to questionnaire 3 to B2C communication service providers and online ads service providers,
question 40 (ID: 855).
209
Reply to questionnaire 3 to B2C communication service providers and online ads service providers,
question 40 (ID: 873).
210
Parties’ reply to RFI 4, question 1a.
211
Replies to questionnaire 3 to B2C communication service providers and online ads service providers,
question 43; replies to questionnaire 4 to business customers, questions 48.
212
Reply to questionnaire 3 to B2C communication service providers and online ads service providers,
question 42 (ID: 882).
213
Parties’ reply to RFI 4, question 1a.
42
This applies equivalently to any further potential segmentations based on
combinations of these wider segments.
6.3.2. Geographic market definition
6.3.2.1. Commission precedents
(158) With reference to the geographic scope of the online advertising market and its
possible segments, the Commission found in previous cases that they should be
defined as national in scope or following linguistic borders within the EEA.
214
Following the Google AdSense decision,
215
the Commission concluded in
Google/Fitbit that the geographic scope of the online advertising market and its
possible segments is national or follows linguistic borders within the EEA.
216
6.3.2.2. The Notifying Party’s view
(159) The Notifying Party submits that although some advertising demand may be
national, the geographic scope of the advertising market is global and competitive
conditions in online advertising services do not vary significantly across geographic
areas.
217
(160) However, the Notifying Party considers that, since the Transaction does not raise
competitive concerns under any plausible market definition, whether the relevant
geographic market is narrower than global or EEA-wide can be left open.
6.3.2.3. Commission’s assessment
(161) The evidence in the Commission's file has not provided any indication which would
suggest that, in defining the relevant geographic market for online advertising (and
sub-segments thereof), it would be appropriate to deviate from its previous decisional
practice.
(162) The majority of respondents to the market investigation (both competitors and
business customers) indicate that the market for online advertising is
national/regional in scope, pointing in particular to language differences as well as
other factors such as different product offerings, cultural differences, regulatory
considerations and locus of control for the ad campaign.
218
Indeed, a majority of
respondents stated that country and/or language specificities (e.g., cultural,
regulatory) are important for the design of an online (display) advertising
campaign.
219
6.3.2.4. Conclusion
(163) In light of the above, for the purpose this Decision, the Commission considers that
the geographic scope of the market for online display advertising services (or
potential segments thereof) is either national in scope or following linguistic borders
within the EEA.
214
Commission decision of 3 October 2014 in case M.7217 – Facebook/WhatsApp, paragraphs 44 and 83;
Commission decision of 6 December 2016 in case M.8124 – Microsoft/LinkedIn, paragraphs 163-164;
Commission decision of 6 September 2018 in case M.8788 – Apple/Shazam, paragraphs 138-140.
215
Commission decision of 20 March 2019 in case AT.40411 – Google AdSense, paragraphs 203-217.
216
Commission decision of 17 December 2020 in case M.9660 – Google/Fitbit, paragraphs 158-160.
217
Form CO, paragraph 6.43.
218
Replies to questionnaire 3 to B2C communication service providers and online ads service providers,
questions 4; replies to questionnaire 4 to business customers, question 53.
219
Replies to questionnaire 3 to B2C communication service providers and online ads service providers,
question 46; replies to questionnaire 4 to business customers, questions 54.
43
7. COMPETITIVE ASSESSMENT
7.1. Introduction
7.1.1. Rationale for the Transaction
(164) The Notifying Party submits that, by acquiring Kustomer, Meta (formerly Facebook)
will have the opportunity to make it easier and more efficient for businesses and
consumers to communicate by accelerating the shift from more traditional channels,
such as phone and email, to messaging channels (such as Messenger, WhatsApp and
Instagram). CRM software could play a key role in facilitating this shift and allowing
businesses to make more and better use of B2C messaging. Kustomer in particular is
an ideal fit for Meta (formerly Facebook) because it has been designed to handle
messaging seamlessly alongside other communication channels. Meta (formerly
Facebook) [Business Plans], and use it to showcase the potential of its messaging
channels in B2C communications, to both businesses and other CRM software
providers.
220
(165) In order to facilitate and accelerate the shift from traditional channels to messaging
channels, the Notifying Party has a three-fold strategy depending on the size of a
business. “Tail businesses” (i.e., very small businesses) generally do not have the
scale or resources to justify the investment in paid CRM software. For these
businesses, Meta (formerly Facebook) has a free service available called Unified
Inbox that enables a business to aggregate messages that they receive on Messenger
and Instagram, and comments on Facebook and Instagram posts, in one consolidated
inbox. “Torso businesses” (i.e., SMBs), which have more interactions with customers
and that may have dedicated customer service personnel, need and often use CRM
software to manage those interactions. While the aim of Meta (formerly Facebook) is
to accelerate the shift to messaging channels more generally, [Business Plans].
Finally, “head businesses” (i.e., large enterprises) generally require a high level of
customisation from their CRM software. While Kustomer will continue to compete
for these businesses, they will typically turn to more sophisticated offerings from
large CRM providers. Meta (formerly Facebook) will also continue to support these
businesses through partnerships and the availability of APIs.
221
(166) The Notifying Party expects to gain in several ways from the Transaction. [Sources
of revenue for Meta resulting from the Transaction] presents a key revenue
opportunity for Meta (formerly Facebook). Meta (formerly Facebook) further
identified two ancillary revenue streams. [Sources of revenue for Meta resulting from
the Transaction].
222
7.1.2. Affected markets
(167) As set out above, the key rationale for the Transaction is the acceleration of the shift
to B2C messaging. In relation to this, the Transaction creates a vertical relationship
between the upstream market for the supply of B2C communication services (and the
narrower markets for asynchronous B2C communication services and OTT B2C
messaging services) and the downstream market for the supply of CRM software and
the narrower market for the supply of customer service and support CRM software.
Section 7.3 assesses the impact of a potential input foreclosure strategy of
Kustomer’s rivals through the restriction or degradation of API access to Meta
220
Form CO, paragraphs 1.5 and following.
221
Form CO, paragraph 3.2.5.
222
Form CO, paragraphs 3.2.11 and 3.2.12.
44
(formerly Facebook)’s B2C messaging channels on the market for the supply of
CRM software and the narrower market for the supply of customer service and
support CRM software.
(168) There is no meaningful horizontal overlap between the Parties’ activities. There may
be a limited horizontal overlap in the B2C communications market between Meta
(formerly Facebook)’s messaging channels and Kustomer’s webchat functionality.
However, this webchat functionality is offered only to businesses using Kustomer’s
CRM software (i.e., not on a standalone basis) and, in any case, Kustomer’s share of
the B2C communications market is less than [0-5%] in the EEA (with just [Customer
Information] of Kustomer’s EEA customers using its webchat functionality) and
even lower worldwide.
223
(169) Although the Transaction does not give rise to any horizontally affected markets in a
traditional sense, the Commission considers that, after the Transaction, Meta
(formerly Facebook) might be able to gain access to commercially exploitable data,
which can be obtained through Kustomer’s activities. Meta (formerly Facebook)’s
databases are not made available today to third parties and Kustomer, as a processor
under the GDPR, cannot on its own take a decision to give access to a database of its
business customers. Post-Transaction, customers of Kustomer will be able to opt-in
to share data with Meta (formerly Facebook) through Kustomer’s CRM software
which could add to the current portfolio of data exploitable by Meta (formerly
Facebook), in particular for online advertising purposes. In Section 7.4, the
Commission will therefore assess the impact of the potential accumulation of data by
Meta (formerly Facebook) on the market for online display advertising, and potential
segments of that market.
(170) According to the information provided by the Parties, there are no non-horizontally
affected markets between Meta (formerly Facebook)’s e-commerce activities and the
market for the supply of CRM software or the supply of customer service and
support CRM software.
224
(171) Finally, the Commission assesses in Section 7.5 whether the Transaction could give
rise to conglomerate effects by Meta (formerly Facebook) leveraging its position in
the market for online display advertising services into the market for the supply of
CRM software and the narrower market for the supply of customer service and
support CRM software (as well as any possible segments thereof).
7.2. Market Shares
(172) According to the Non-Horizontal Guidelines
225
, market shares provide useful first
indications of the market structure and of the competitive importance of both the
merging parties and their competitors.
7.2.1. CRM software
(173) Table 1 and Table 2 set out Kustomer’s and its main competitors’ market shares, in
value terms, within the market for customer service and support CRM software as
223
Form CO, paragraph 6.54 Tables 6.7 and 6.8.
224
Parties’ reply to RFI 20, question 6.
225
Guidelines on the assessment of non-horizontal mergers under the Council Regulation on the control of
concentrations between undertakings ("Non-Horizontal Guidelines"), OJ C 265, 18.10.2008,
paragraph 24.
50
(179) Finally, within the potential market for OTT B2C messaging services, as a sub-
segment of asynchronous B2C communication services, Meta (formerly Facebook)
has a share of more than [40-50%] in 2020, both worldwide and in the EEA. Meta
(formerly Facebook)’s market share has significantly and consistently increased in
the period from 2018 to 2020, both globally and in the EEA. B2C messaging is a
growing B2C communication channel, and discussions with market participants
indicate that it is expected to become the primary B2C communication channel in the
future.
239
Other messaging includes other OTT B2C messaging services (including
Google’s RCS-supported messaging channels and Apple Business Chat) and
webchat. The Notifying Party submitted that with regard to those channels, it was not
able to provide separate conversation volumes in a reliable manner. However, it did
provide market share estimates for other OTT B2C messaging services and webchat
on a worldwide market for 2020.
240
On this market, the market share for webchat
would be around [Market shares]. The market shares of other OTT B2C messaging
services would amount to approximately [10-20%].
241
As there are several providers
of webchat services, as well as several other OTT B2C messaging services, Meta
(formerly Facebook) appears to be the biggest single player in a potential market for
OTT B2C messaging services.
7.2.3. Online Display Advertising
(180) Meta (formerly Facebook) has not provided market shares for online display
advertising services and sub-markets/segments thereof. In previous decisions, the
Commission found that, whilst overall the online display advertising market is
fragmented, Meta (formerly Facebook) would be the largest player.
242
(181) The Commission has thus reconstructed Meta (formerly Facebook)’s market shares
for the national markets/sub-markets/segments for online display advertising, online
display advertising on-social networks, online display advertising off-social
networks, online display video advertising, online display non-video advertising,
online display advertising on desktops and online display advertising on mobile, on
the basis of data requested from the main market players on their revenues with
online advertising services on their own properties as well as data on the annual
value of the display advertising served to third parties publishers (i.e., including
revenues received by publishers). Data was collected from 18 market players. The
market size was calculated by summing the revenues of all respondents.
243
This may
under-estimate the actual size of the market given that not all players responded to
the Commission’s RFI, but there are no reasons in the present case to believe that the
estimated market size differs significantly from the actual size.
239
Agreed minutes of the call of 18 May 2021 with Zendesk, paragraph 10 (ID: 2090); agreed minutes of
the call of 14 June 2021 with Twilio, paragraph 12 (ID: 2029); agreed minutes of the call of 26 June
2021 with Salesforce, paragraph 8 (ID: 2005).
240
Form CO, Figure 6.1.
241
These figures have been calculated by the Commission using figures for the overall B2C
communications market provided by the Notifying Party in the Form CO, paragraph 6.54 Tables 6.7
and 6.8. These market shares are therefore estimates.
242
Commission decision of 6 December 2016 in case M.8124 – Microsoft/LinkedIn, paragraphs 171-173.
243
In cases where certain data were missing from players, the Commission has used certain assumptions
that over-estimate the corresponding market sizes, for instance by double-counting revenues across sub-
markets/segments of the online display advertising market. This may in turn under-estimate Meta
(formerly Facebook)’s market shares. Further, the Commission notes that the allocation of revenues
between sub-markets/segments of the online display advertising market has been made by the market
player themselves.
57
(183) As illustrated in the above Tables, in 2020 Meta (formerly Facebook)’s market
shares was:
Above 30%
245
:
– In the supply of online display advertising services in Austria, Belgium,
Bulgaria, Croatia, Czechia, Denmark, Estonia, Finland, France, Germany,
Greece, Hungary, Iceland, Ireland, Italy, Netherlands, Norway, Poland,
Portugal, Romania, Slovakia, Slovenia, Spain, and Sweden;
246
– In the supply of online display advertising services on social networks in
Austria, Belgium, Bulgaria, Croatia, Czechia, Denmark, Estonia, Finland,
France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Latvia,
Liechtenstein, Luxembourg, Malta, Netherlands, Norway, Poland, Portugal,
Romania, Slovakia, Slovenia, Spain, and Sweden;
247
– In the supply of online display video advertising services in Austria, Belgium,
Bulgaria, Croatia, Czechia, Denmark, Estonia, Finland, France, Germany,
Greece, Hungary, Iceland, Ireland, Italy, Netherlands, Norway, Poland,
Portugal, Romania, Slovakia, Slovenia, Spain, and Sweden;
248
– In the supply of online display non-video advertising services in Austria,
Belgium, Bulgaria, Croatia, Czechia, Denmark, Estonia, Finland, France,
Germany, Greece, Hungary, Iceland, Ireland, Italy, Lithuania, Netherlands,
Norway, Poland, Portugal, Romania, Slovakia, Slovenia, Spain, and
Sweden;
249
– In the supply of online display advertising services on mobile in Austria,
Belgium, Bulgaria, Croatia, Czechia, Denmark, Estonia, Finland, France,
Germany, Greece, Hungary, Iceland, Ireland, Italy, Netherlands, Norway,
Poland, Portugal, Romania, Slovakia, Slovenia, Spain, and Sweden;
250
and
– In the supply of online display advertising service on desktop in Austria,
Belgium, Croatia, Czechia, Denmark, Estonia, Finland, Greece, Hungary,
Norway, Portugal, Romania, Slovakia, and Slovenia.
251
Above 50%
252
:
– In the supply of online display advertising services in Austria, Belgium,
Croatia, Czechia, Denmark, Finland, France, Germany, Greece, Hungary,
245
Please note these lists include markets where Meta (formerly Facebook)’s share is above 50% and
above 75%.
246
Following linguistic borders within the EEA, this may also include Cyprus, Liechtenstein, Luxembourg,
and Malta.
247
Following linguistic borders within the EEA, this may also include Cyprus.
248
Following linguistic borders within the EEA, this may also include Cyprus, Liechtenstein, Luxembourg,
and Malta.
249
Following linguistic borders within the EEA, this may also include Cyprus, Liechtenstein, Luxembourg,
and Malta.
250
Following linguistic borders within the EEA, this may also include Cyprus, Liechtenstein, Luxembourg,
and Malta.
251
Following linguistic borders within the EEA, this may also include Cyprus, France, Germany,
Liechtenstein, Luxembourg, and the Netherlands.
252
Please note these lists include markets where Meta (formerly Facebook)’s share is above 75%.
58
Iceland, Italy, Norway, Poland, Portugal, Romania, Slovakia, Slovenia, Spain,
and Sweden;
253
– In the supply of online display advertising services on social networks in
Austria, Belgium, Bulgaria, Croatia, Czechia, Denmark, Estonia, Finland,
France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Latvia,
Liechtenstein, Malta, Netherlands, Norway, Poland, Portugal, Romania,
Slovakia, Slovenia, Spain, and Sweden;
254
– In the supply of online display video advertising services in Austria, Belgium,
Croatia, Czechia, Denmark, France, Germany, Greece, Hungary, Iceland, Italy,
Norway, Poland, Portugal, Romania, Slovakia, Spain, and Sweden;
255
– In the supply of online display non-video advertising services in Austria,
Belgium, Croatia, Czechia, Denmark, Finland, France, Greece, Hungary,
Iceland, Italy, Norway, Poland, Portugal, Romania, Slovakia, Slovenia, Spain,
and Sweden;
256
– In the supply of online display advertising services on mobile in Austria,
Belgium, Croatia, Czechia, Denmark, Finland, France, Germany, Greece,
Hungary, Iceland, Italy, Netherlands, Norway, Poland, Portugal, Romania,
Slovakia, Slovenia, Spain, and Sweden;
257
and
– In the supply of online display advertising service on desktop in Slovenia.
Above 75%:
– In the supply of online display advertising services on social networks in
Austria, Belgium, Croatia, Denmark, Finland, Greece, Hungary, Iceland, Italy,
Norway, Portugal, and Romania;
258
– In the supply of online display non-video advertising services in Belgium,
Greece, Hungary, and Slovenia;
259
and
– In the supply of online display advertising services on mobile in Belgium,
Greece, Hungary, Iceland, and Romania.
260
(184) In addition to display advertising on its own properties, Meta (formerly Facebook)
also serves display ads to third parties, via its intermediation service MAN. Although
the associated revenues are shared between Meta (formerly Facebook) and those
third party publishers, Meta (formerly Facebook) has a decisive role in determining
the allocation of ads it serves to third parties. As such, even though these revenues
253
Following linguistic borders within the EEA, this may also include Cyprus, Liechtenstein, Luxembourg,
and Netherlands.
254
Following linguistic borders within the EEA, this may also include Cyprus and Luxembourg.
255
Following linguistic borders within the EEA, this may also include Cyprus, Liechtenstein, Luxembourg,
and Netherlands.
256
Following linguistic borders within the EEA, this may also include Cyprus, Liechtenstein, Luxembourg,
Germany and Netherlands.
257
Following linguistic borders within the EEA, this may also include Cyprus, Luxembourg and
Liechtenstein.
258
Following linguistic borders within the EEA, this may also include Cyprus, France, Germany,
Liechtenstein, Luxembourg and the Netherlands.
259
Following linguistic borders within the EEA, this may also include Austria, Cyprus, France, Germany,
Liechtenstein, Luxembourg and the Netherlands.
260
Following linguistic borders within the EEA, this may also include Austria, Cyprus, France, Germany,
Liechtenstein, Luxembourg and the Netherlands.
59
have not been included in Meta (formerly Facebook)’s market shares above, they
may constitute an additional (aggravating) factor relevant to Meta (formerly
Facebook)’s market position within display advertising. In 2020, the share of the
national markets served by Meta (formerly Facebook) to third parties corresponded
to (i) between [0-5%] and [10-20%] of the total revenues with display advertising
off-social networks, (ii) between [0-5%] and [5-10%] of the total revenues with
display video advertising, depending on the EEA country, and (iii) below [0-5%] for
display advertising and all other potential segments thereof.
7.3. Assessment of vertical effects
7.3.1. Legal framework
(185) According to the Non-Horizontal Guidelines,
261
non-coordinated effects may
significantly impede effective competition as a result of a vertical merger if such
merger gives rise to foreclosure. Foreclosure occurs where actual or potential
competitors' access to supplies or markets is hampered or eliminated as a result of the
merger, thereby reducing those companies' ability and/or incentive to compete.
262
Such foreclosure may discourage entry or expansion of competitors or encourage
their exit.
263
(186) The Non-Horizontal Guidelines distinguish between two forms of foreclosure. Input
foreclosure occurs where the merger is likely to raise the costs of downstream
competitors by restricting their access to an important input. Customer foreclosure
occurs where the merger is likely to foreclose upstream competitors by restricting
their access to a sufficient customer base.
264
(187) Foreclosure may also take more subtle forms, such as the degradation of the quality
of input supplied. In its assessment, the Commission may consider a series of
alternative or complementary possible strategies.
265
(188) In assessing the likelihood of an anticompetitive foreclosure scenario, the
Commission examines, first, whether the merged entity would have, post-merger, the
ability to substantially foreclose access to inputs or customers, second, whether it
would have the incentive to do so, and third, whether a foreclosure strategy would
have a significant detrimental effect on competition. In practice, these factors are
often examined together as they are closely intertwined.
266
(189) Pursuant to Article 2(2) of the Merger Regulation, a concentration which would not
significantly impede effective competition, in the internal market or in a substantial
part of it, shall be declared compatible with the internal market. Conversely, pursuant
to Article 2(3) of the Merger Regulation, a concentration which would significantly
impede effective competition, in the internal market or in a substantial part of it, shall
be declared incompatible with the internal market. In that regard, it is sufficient for
the Commission to establish that the Transaction is likely to lead to a significant
impediment of effective competition. In that regard, where it has been notified of a
proposed concentration pursuant to the Merger Regulation, the Commission is, in
principle, required to adopt a position, either in the sense of approving or of
261
Non-Horizontal Guidelines, pages 6-25.
262
See Non-Horizontal Guidelines, paragraph 18.
263
See Non-Horizontal Guidelines, paragraph 29.
264
See Non-Horizontal Guidelines, paragraph 30.
265
See Non-Horizontal Guidelines, paragraph 33.
266
See Non-Horizontal Guidelines, paragraph 32.
60
prohibiting the concentration, in accordance with its assessment of the economic
outcome attributable to the concentration which is most likely to ensue.
267
It follows
that the Commission has to establish whether the Transaction is likely to lead to a
significant impediment of effective competition.
268
7.3.2. Foreclosure of access to APIs to Meta (formerly Facebook)’s messaging channels to
the detriment of competing providers of CRM software (input foreclosure)
(190) The Transaction gives rise to a vertical relationship
269
between Meta (formerly
Facebook)’s activities in the upstream market for the supply of OTT B2C messaging
services (and the broader market for B2C communication services) and Kustomer’s
activities in the downstream market for the supply of customer service and support
CRM software (and the broader market for the supply of CRM software overall).
(191) OTT B2C messaging channels, and indeed B2C communication channels more
generally, constitute an important input for the supply of customer service and
support CRM software. Customer service and support CRM software providers
obtain API access to OTT B2C messaging channels to help businesses manage
267
Case C-413/06 P, Bertelsmann a.o. v Impala, EU:C:2008:392, paragraph 52.
268
The Notifying Party argues that the standard of proof in order to establish a significant impediment to
effective competition is a standard of a “strong probability”, and refers in this respect to the judgment of
the General Court in Case T-399/16 CK Telecoms v Commission (EU:T:2020:217, paragraph 118) (SO
Response, paragraph 1.6(A)). The Commission disagrees that this is the correct test, and the judgment is
currently under appeal (Case C-376/20 P, Commission v CK Telecoms). The Commission considers
however that in any event, there is a strong probability that the Transaction would have significant
negative effects on competition in the market for customer service and support CRM software (see
Section 7.3.2.3 below).
269
The assessment would not materially change if examined through the lenses of conglomerate non-
coordinated effects. Foreclosure effects may arise when the combination of products in related markets
may confer on the merged entity the ability and incentive to leverage a strong market position from one
market to another closely related market by means of tying or bundling or other exclusionary practices.
Similarly to input foreclosure, conglomerate foreclosure may also take more subtle forms, such as the
degradation of the quality of a product supplied on a standalone basis. In certain circumstances such
practices may lead to a reduction in actual or potential competitors' ability or incentive to compete,
which may in turn reduce the competitive pressure on the merged entity allowing it to increase prices.
See Non-Horizontal Guidelines, paragraphs 91 and 93. As with vertical non-coordinated effects, in
assessing the likelihood of such other exclusionary practices in the form of foreclosure, the Commission
examines, first, whether the merged entity would have the ability to engage in such conduct by reducing
the ability to compete of its actual or potential competitors, second, whether it would have the economic
incentive to do so and, third, whether such strategy would have a significant detrimental effect on
competition, thus causing harm to consumers. See Non-Horizontal Guidelines, paragraph 94. Given that
in the case of exclusionary practices in the form of foreclosure the assessment framework for both
vertical and conglomerate non-coordinated effects is largely the same, the Commission’s assessment
below will focus on vertical effects. However, the assessment would be largely the same for
exclusionary practices under a conglomerate perspective. The link between Meta (formerly Facebook)’s
activities in B2C communication services and Kustomer’s activities in CRM software may also
constitute a conglomerate relationship in cases where businesses obtain API access to B2C
communications themselves directly from Meta (formerly Facebook) or other messaging channel
providers, or via a CPaaS provider, and separately procure a customer service and support CRM
solution from a CRM software provider. The Notifying Party argues that a conglomerate assessment is
therefore more appropriate. However, businesses that use CRM software generally integrate the B2C
communication channels (which they may have accessed via direct APIs or through CPaaS providers)
into their chosen CRM software, and CRM software providers actively facilitate such integration. As
such, for the purposes of this Decision, the Commission does not consider it to be necessary to carry out
a separate conglomerate effects assessment as the vertical assessment carried out by the Commission
addresses each of the various elements that may be necessary to determine the likelihood of foreclosure.
61
communications with consumers by unifying several interactions (phone, email,
webchat, SMS, Messenger, WhatsApp, Instagram, Twitter, etc.) under a single tool.
(192) While it is left open in Section 6.1.1.4 above whether there is a separate market for
the supply of customer service and support CRM software or whether it is part of the
overall market for the supply of CRM software, the Commission undertakes its
assessment of vertical effects on the narrower market for the supply of customer
service and support CRM software. The Commission considers that its assessment
holds even if it would be based on the broader market for CRM software overall.
This is in particular the case because, according to Gartner, […].
270
(193) In the market for OTT B2C messaging, Meta (formerly Facebook) holds a market
share of well above [40-50%] (with its WhatsApp, Messenger and Instagram
messaging channels), both at worldwide and EEA level. Meta (formerly Facebook)’s
share in this market has also been growing significantly over the past years.
Kustomer’s position on the downstream market is currently limited (less than [0-5%]
in CRM software, as well as in the narrower market for customer service and support
CRM software).
271
(194) The Commission has investigated whether Meta (formerly Facebook) has the ability
and incentive to foreclose competing customer service and support CRM software
providers, either through restricting API access to its channels, degrading rivals’ API
access to its channels, or upgrading Kustomer’s API access to its channels relative to
rivals’ access. The Commission also assessed whether such a foreclosure strategy
would have a negative impact on competition in the market for the supply of
customer service and support CRM software (and the broader overall CRM software
market).
7.3.2.1. Notifying Party’s views
272
(195) The Notifying Party submits that the merged entity will not have the ability or the
incentive to restrict or degrade access to its B2C communication channels for
downstream competitors of Kustomer post-Transaction (as an input into their CRM
software offering)
273
. Moreover, the Notifying Party submits that such strategy
would not have a significant impact in the CRM software market.
274
In relation to
this theory of harm, the Notifying Party has submitted an economic paper.
275
The
Notifying Party reiterated and expanded upon these points in the Article 6(1)(c)
Response and in the SO Response.
(a) Ability to foreclose
(196) The Notifying Party submits that Meta (formerly Facebook) does not have the ability
to foreclose Kustomer’s rival CRM providers.
270
See Kustomer’s internal document, […].
271
The Commission has therefore excluded possible standalone customer foreclosure concerns in the
immediate future.
272
In the Form CO, the Notifying Party submits its views on input foreclosure under the Section for
conglomerate effects (“Conglomerate Foreclosure: B2C communication services and CRM software
solutions (1) foreclosure of Kustomer’s CRM rivals”) as it considers there is significant overlap
between the assessment of any conglomerate foreclosure strategy relating to Kustomer’s rivals in CRM
software and any assessment of input foreclosure under vertical effects (Form CO, paragraph 8.84.)
273
Form CO, paragraph 6.86.
274
Form CO, paragraph 6.109.
275
Form CO, Annex 13.17, Acquisition by Facebook, Inc. of Kustomer, Inc. Analysis of Vertical Effects,
and Annex 13.18, Acquisition by Facebook, Inc. of Kustomer, Inc. Vertical arithmetic analysis.
62
(197) In the Form CO, the Notifying Party argues that it does not have any market power in
B2C communication services that it could use as leverage to impose material
switching to Kustomer from rival CRM providers. According to Meta (formerly
Facebook), there is an overall market for B2C communications, including phone and
email, in which Meta (formerly Facebook) would have a single-digit share ([5-10%]
globally, [0-5%] in the EEA).
276
(198) Moreover, the Notifying Party submits that Meta (formerly Facebook)’s messaging
channels are not “must-have” products for CRM software providers or their business
customers (now or in the near future), and provides examples of CRM providers that
do not integrate Meta (formerly Facebook)’s channels.
277
The Notifying Party also
considers that end-consumers multi-home between and switch across B2C
communication channels regularly and therefore do not consider Facebook or
messaging as a “must-have” B2C communication channel.
278
(199) According to the Notifying Party, these conclusions are not altered by the fact that
the importance of messaging channels for B2C communications is forecast to
increase significantly in the coming years.
279
(200) Further, the Notifying Party explains that there are strong alternative third-party
messaging providers, namely Google and Apple, who are investing heavily in B2C
messaging and whose products would represent attractive alternatives to Meta
(formerly Facebook)’s messaging channels in the event of an attempted
foreclosure.
280
(201) In its Article 6(1)(c) Response, the Notifying Party further argues that its B2C
messaging channels are not an important input for the downstream market for several
reasons. First, the Notifying Party explains that its channels do not represent a
significant cost factor, but rather around [Pricing Information] of a CRM’s costs.
Second, they are not a critical component for the sale of CRM software as Meta
(formerly Facebook)’s messaging channels are not the only channels that can be used
for B2C communications. Third, Meta (formerly Facebook)’s claims that its
messaging channels do not represent a significant source of product differentiation
for CRM software providers on the basis that integration of B2C messaging channels
does not represent an important driver of competition in the downstream market.
Fourth, Meta (formerly Facebook) claims that the cost of switching to alternative
inputs is nil or negligible as all CRM software providers integrate several OTT B2C
messaging channels, in addition to phone, email, SMS and other B2C channels.
281
(202) The Notifying Party argues that there is in reality “minimal usage of Facebook’s B2C
channels” compared to channels such as phone and email.
282
While accepting that
the OTT B2C messaging market is fragmented, the Notifying Party nonetheless
considers it to be dynamic with “strong and growing rivals such as Google [RCS]
and Apple [Apple Business Chat] and webchat providers”
283
that would exert a
competitive constraint on Meta (formerly Facebook)’s channels.
276
Form CO, paragraphs 6.88 - 6.92.
277
Form CO, paragraphs 6.91 and 6.92.
278
Form CO, paragraphs 6.98 and 6.99.
279
Form CO, paragraph 6.94.
280
Form CO, paragraph 6.95.
281
Article 6(1)(c) Response, paragraph 3.11.
282
Article 6(1)(c) Response, paragraphs 3.12-3.13.
283
Article 6(1)(c) Response, paragraphs 3.15-3.16.
63
(203) Contrary to the Commission’s conclusion in the Article 6(1)(c) Decision, the
Notifying Party considers that the high penetration rates of Meta (formerly
Facebook)’s messaging channels do not reflect market power in the OTT B2C
messaging market as they only relate to consumer usage (or peer-to-peer (“P2P”)
communications), and the “out-of-market” constraint exercised by channels such as
phone and email, which still account for the majority of B2C communications
overall, cannot be ignored.
284
The Notifying Party further argues that the Article
6(1)(c) Decision fails to consider the countervailing buyer power of major CRM
software providers such as Salesforce, SAP and Oracle, which already integrate
multiple OTT B2C messaging channels and could credibly threaten to switch within
a short timeframe. According to the Notifying Party, they have the ability to
influence businesses’ choice of which B2C channels to provide to their end
customers.
285
For all of these reasons, the Notifying Party considers that it would not
have the ability to engage in an input foreclosure strategy.
(204) The Notifying Party argues that, in addition to not having the ability to engage in
total input foreclosure (i.e., to deny access), Meta (formerly Facebook) would
likewise not have the ability to engage in partial input foreclosure (i.e., degradation
of access to third parties or granting inferior access to third parties compared to the
access that Kustomer would have). First, the Notifying Party points out that, as
outlined in the Form CO, since the average cost of integrating Meta (formerly
Facebook)’s WhatsApp for Business API (as its Messenger API is free) accounts for
a [Pricing information]) of the average price paid for Kustomer’s CRM product, a
strategy of raising rivals’ costs by increasing the price of API access to its messaging
channels would not lead to material switching, in particular — it is argued —
because there are “significant costs [involved in] switching CRM” providers.
286
(205) The Notifying Party similarly argues that any reduction in the quality of API access
relative to that of Kustomer would likely not result in successful foreclosure of
Kustomer’s rivals due to the “substantial costs involved in switching CRM” provider,
and would “likely lead to severe reputational damage” for Meta (formerly Facebook)
and push businesses to rival OTT B2C messaging channels rather than to Kustomer’s
CRM platform.
287
(206) Finally, the Notifying Party argues that Meta (formerly Facebook)’s ability to engage
in targeted foreclosure would be “limited” because there is “no group of customers
that is especially reliant on Facebook’s B2C messaging channels” and in any case
Meta (formerly Facebook) would “not be able to accurately identify a specific group
of businesses which would be more profitable to target”.
288
Even if it could identify
those businesses, the Notifying Party considers that it would not be able to target
them as it would not have the ability to “withdraw or degrade access to its B2C
channels for a subset of businesses that use” a rival CRM platform.
289
In any case,
the Notifying Party considers that there are “workarounds” available even if it was
able to target specific businesses, who could “regain access via some other
284
Article 6(1)(c) Response, paragraph 3.16.
285
Article 6(1)(c) Response, paragraph 3.16.
286
Article 6(1)(c) Response, paragraph 3.18.i.
287
Article 6(1)(c) Response, paragraph 3.18.ii.
288
Article 6(1)(c) Response, paragraph 3.19. See also Notifying Party Submission on Foreclosure Issues
dated 10 October 2021, footnote 1.
289
Article 6(1)(c) Response, paragraph 3.19.
64
means”
290
“for example, through a CPaaS provider or via integrations with a third-
party CRM provider”.
291
(207) In the SO Response, the Notifying Party restates a number of its arguments from the
Form CO and its Article 6(1)(c) Response, in addition to raising additional points
and putting forward some new evidence.
(208) First, the Notifying Party argues that the Commission unduly relies on responses
from CRM providers in concluding that Meta (formerly Facebook)’s channels are
important input, and ignores contrary evidence that a majority of business customers
do not consider them to be important.
292
The Notifying Party goes on to refer to
statistics previously submitted that Meta (formerly Facebook)’s B2C messaging
channels are not extensively used by business customers of CRM providers today,
pointing to the fact that of the c. [Customer information] of Kustomer’s customers
that use Meta (formerly Facebook)’s channels today, the [Customer information].
293
(209) Second, the Notifying Party argues that the Commission simply infers that Meta
(formerly Facebook) has market power from it “[…] share of B2C OTT
communications” without any other assessment and wrongly relies on the […] P2P
penetration rates of Meta (formerly Facebook)’s channels as being indicative of
market power in OTT B2C communications.
294
In particular, the Notifying Party
states that the Commission ignores “clear and probative evidence of the strength of
Facebook’s competitors”, in particular by taking a static view of the current positions
of Google RCS and Apple Business Chat today, which are very small, but—it is
submitted—ignoring the expected future growth of these rival channels, as well as
the penetration rates of Google and Apple in operating systems (“74% and 23%
respectively based on third party sources”).
295
(210) Third, the Notifying Party argues that the Commission ignores “clear evidence that
businesses can and do switch easily between B2C communication channels” given
that CRM provides typically integrate a range of different B2C communication
channels. In particular, the Notifying Party provides an analysis of a recent Meta
(formerly Facebook) outage on 4 October 2021, during which Meta (formerly
Facebook)’s messaging channels were entirely unavailable for both businesses and
end customers for approximately 6 to 8 hours. The Notifying Party argues that this an
analysis of [Information regarding customer behaviour during outage of Meta’s
messaging channels].
296297
(211) Fourth, the Notifying Party argues that the Commission fails to acknowledge that
beta testing
298
of new features and functionalities of a B2C messaging channel (e.g.,
290
Article 6(1)(c) Response, paragraph 3.19.
291
Notifying Party Submission on Foreclosure Issues dated 10 October 2021, footnote 1.
292
SO Response, paragraph 3.6.
293
SO Response, paragraph 3.8(A).
294
SO Response, paragraph 3.7.
295
SO Response, paragraph 3.8(C).
296
SO Response, Annex 3, paragraph 3.3.
297
SO Response, Annex 3, paragraph 3.3.
298
As explained by the Parties in reply to RFI 21, beta testing means testing of new API functionalities
with a small group of customers before rolling out to all customers, e.g., to check for bugs or other
software problems. For further background, alpha testing is the first phase of software testing - products
at this phase often lack key features or functionalities and may contain serious errors. Beta testing
occurs after alpha testing - products at this stage are likely to have more complete features and
functionalities than at the alpha phase, but may still contain bugs.
65
to check for bugs or software problems) is a “standard industry practice”.
299
The
Notifying Party points out that beta tests are typically only carried out with a subset
of CRM partners and, and only a small proportion of a participating CRM provider’s
business customers will be involved in such testing. This means beta tests are by
definition not must-haves whereas the Notifying Party argues that the Commission
“effectively asserts that, not only are Facebook B2C channels a critical input, every
single new feature of a Facebook B2C channel is a critical input (and before it has
even completed the beta testing process).”
300
(212) Fifth, the Notifying Party restates arguments that Meta (formerly Facebook) would
not be able to target close competitors of Kustomer as part of a targeted foreclosure
strategy due to an inability to identify which CRM providers are used by businesses
that connect to Meta (formerly Facebook)’s B2C messaging channels directly, or
indirectly via non-CRM providers, e.g., CPaaS providers, which account for the vast
majority of businesses using its channels.
301
(213) Sixth, and finally, the Notifying Party argues that it cannot be guaranteed that Meta
(formerly Facebook)’s partners, such as CPaaS providers, would cooperate with
Meta (formerly Facebook) in carrying out a targeted foreclosure strategy and that to
rely on “standard contractual language” to enforce such a strategy would be to
ignore “rational commercial considerations”.
302
(214) In the LoF Response, first, on the importance of Meta (formerly Facebook)’s B2C
messaging channels, the Notifying Party argues that in assessing the market
investigation results, the Commission wrongly excluded respondents that indicated
“don’t know”, ignored contrary evidence in qualitative responses, and also relied on
public statements by Meta (formerly Facebook)’s employees which it argues are less
probative than other data provided.
303
(215) Second, the Notifying Party argues that the Commission’s assessment ignores
evidence on the importance and growing competitive strength of Google RCS and
Apple Business Chat, such as Google’s “track record of successfully rolling out
Google RCS without the support of carriers [in the US and in Italy]”
304
and the fact
that it is “available…in all EEA member states”.
305
It is further argued that the
Commission ignored the importance of the “near full penetration of the entire mobile
user base in Europe [of Google and Apple’s operating systems] i.e. 73% and 25%
respectively”.
306
(216) Third, in relation to Meta (formerly Facebook)’s ability to engage in a targeted
foreclosure strategy of specific customer service and support CRM software
providers, the Notifying Party notes that this would “require a complex and
adversarial process”
307
and the examples of Meta (formerly Facebook) having
removed, threatened to remove or the allegations that it “removed API access to
299
SO Response, paragraph 3.13.
300
SO Response, paragraph 3.13.
301
SO Response, paragraph 3.16.
302
SO Response, paragraph 3.18(B).
303
LoF Response, paragraph 2.8-2.10.
304
LoF Response, paragraph 2.3(B)(iii).
305
LoF Response, paragraph 2.3(B)(v).
306
LoF Response, paragraph 2.4(A).
307
LoF Response, paragraph 2.5.
66
specific businesses [have] no bearing on its ability to establish whether specific
businesses are using specific CRMs”
308
.
(b) Incentive to foreclose
(217) The Notifying Party submits that Meta (formerly Facebook) does not have the
incentive to foreclose Kustomer’s rival CRM providers.
(218) First, the Notifying Party considers that to prevent or degrade businesses’ access to
its messaging channels is irreconcilable with Meta (formerly Facebook)’s strategy to
[Details of Meta’s strategy and growth expectations for B2C communications].
309
The Notifying Party expects almost [Details of Meta’s strategy and growth
expectations for B2C communications], and these upstream profits would be put at
risk as a result of foreclosure.
310
Meta (formerly Facebook) expects B2C messaging
rivals, such as Apple and Google, to capitalise on any foreclosure strategy and grow
at the expense of Meta (formerly Facebook).
311
(219) Second, in addition to lost sales in the B2C communications market, the Notifying
Party submits that such a foreclosure strategy would not result in substantially
greater sales in the CRM software market.
312
The Notifying Party expects that the
gains in downstream profits would be low, because there would be limited switching
by business customers to Kustomer from CRM software rivals. This is because
access to Meta (formerly Facebook)’s B2C channels is not a key driver of
businesses’ choice of CRM software (but one of many factors, e.g., cloud-based
systems, AI capabilities, price, and technical support), and so business customers
would switch B2C channels rather than CRM software providers; the latter of which
can often be resource intensive.
313
Further, the Notifying Party points out that
Kustomer’s offering is not suited to many businesses.
314
(220) As part of the submitted economic paper, the Notifying Party provided a model that
calculates the critical number of businesses that would need to switch to Kustomer
for this foreclosure strategy to be profitable (taking into account both the potential
losses and gains). The model concluded that this critical number of businesses is too
high for a foreclosure strategy to be realistically profitable.
315
In its Article 6(1)(c)
Response, the Notifying Party reiterates that its own model is robust and the
shortcomings highlighted by the Commission are unfounded, because (i) the model
accurately accounts for incremental ads revenues when they are envisaged to arrive,
(ii) the information from Kustomer which the Commission references is not credible,
and (iii) further modifications to the model are unfounded or do not change the
conclusion.
316
308
LoF Response, paragraph 2.6.
309
Form CO, paragraphs 6.102 and 6.105.
310
SO Response, paragraph 2.5.
311
Article 6(1)(c) Response, paragraph 3.22.ii.
312
Form CO, paragraphs 6.102, 6.106 and 6.107. Article 6(1)(c) Response paragraph 3.24.i.
313
Article 6(1)(c) Response paragraph 3.24.iii.
314
Article 6(1)(c) Response paragraph 3.24.iv.
315
Form CO, Annex 13.7, Acquisition by Facebook, Inc. of Kustomer, Inc. Analysis of Vertical Effects,
pages 21-24.
316
Article 6(1)(c) Response, paragraph 3.29.
67
(221) Third, the Notifying Party submits that foreclosure would be even less profitable in
the future, because B2C messaging related revenues are expected to grow
significantly over the next few years.
317
(222) Finally, the Notifying Party considers that partial or targeted foreclosure would be
even less profitable than total market-wide foreclosure, because the reputational
harm and negative impact on its other revenues streams would be greater under
partial foreclosure. It expects that attempting to degrade the interoperability of rival
CRM software with Meta (formerly Facebook)’s B2C channels would damage Meta
(formerly Facebook)’s reputation among messaging partners, most importantly the
leading CPaaS providers that are key to driving adoption of Meta (formerly
Facebook)’s messaging channels.
318
(223) In the SO Response, the Notifying Party restates a number of its arguments from the
Form CO and its Article 6(1)(c) Response, in addition to raising additional points
and putting forward some new evidence.
(224) First, the Notifying Party argues that the Commission’s assessment does not contain
any proper, reasoned and quantitative assessment of Facebook’s incentive to
foreclose, but instead relies on qualitative analysis;
319
and that the Commission offers
no explanation of why Parties quantitative evidence is not reliable.
320
(225) Second, the Notifying Party argues that the Commission’s assessment alternates
between different foreclosure strategies, and does not clearly assess the incentives to
foreclose for each foreclosure strategy separately.
321
(226) Third, the Notifying Party argues that the Commission is wrong to dismiss the
Parties’ historical switching data and the Parties’ estimated diversion ratios following
foreclosure based on these historical data. The Notifying Party argues these data are
indicative of switching post-Transaction and post-foreclosure.
322
(227) Fourth, the Notifying Party argues that the Commission makes a calculation error by
calculating costs based on the population of […], but calculates benefits based on
[…] business customers most likely to consider Kustomer.
323
(228) Fifth, the Notifying Party argues that the SO overstates each category of hypothetical
benefits from businesses that would switch to Kustomer, and understates each
category of costs, including wider costs of foreclosure such as reputational harm.
324
(229) Finally, the Notifying Party also argues that businesses would be constrained from
switching away from a chosen CRM provider due to the integrations and features
(other than OTT B2C messaging channels) offered by that CRM provider and which
are not available at Kustomer, noting that “for example, Kustomer offers [Details
about Kustomer’s integrations] of the integrations available in[Details about
Kustomer’s integrations], and offers only [Details about Kustomer’s integrations] of
the integrations available in[Details about Kustomer’s integrations]”.
325
The
317
Article 6(1)(c) Response, paragraph 3.27.
318
Form CO, paragraph 6.108.
319
SO Response, paragraph 4.4.
320
SO Response, paragraph 4.5.
321
SO Response, paragraph 4.6.
322
SO Response, paragraph 4.9.
323
SO Response, paragraph 4.9(B).
324
SO Response, paragraph 4.10 and 4.14.
325
SO Response, paragraph 4.20(B).
68
Notifying Party deduces from this that “the lack of overlap between apps integrated
by different CSCRM providers will make switching difficult or impossible.”
326
(230) In the LoF Response, the Notifying Party restates a number of its arguments from the
SO Response, but does not raise any additional points or put forward any new
evidence.
327
(c) Effects on competition
(231) The Notifying Party submits that a foreclosure strategy would at most lead to only a
limited diversion of sales to Kustomer, without material harm to its CRM rivals,
given that Kustomer is a very small player in the CRM market. The Notifying Party
explains that to the extent that a foreclosure strategy could cause some diversion, that
could not give rise to a significant detrimental effect on competition: businesses
would continue to be well served by Kustomer’s larger rivals and remain able to
easily compete for any customers in the event of any price increase or deterioration
in quality.
328
(232) In its Article 6(1)(c) Response, the Notifying Party further elaborates that such
strategy would not have a significant detrimental effect on competition because (i)
Kustomer is and will remain a tiny player (even post-foreclosure) and so the effect
on the downstream market would not be to create a dominant position or a significant
impediment of effective competition
329
, (ii) CRM software rivals will react by
promoting Apple and Google’s OTT B2C messaging channels and competing along
different parameters, such as the integrations, services and customisations offered,
thereby negating the anticompetitive effects of foreclosure
330
, (iii) smaller CRM
providers are not more innovative than the larger players who have significant
R&D
331
, and (iv) partial and targeted foreclosure must by definition have a smaller
effect than total and market-wide foreclosure.
332
(233) In its SO Response, the Notifying Party further argues that the SO does not show
competitive harm given that most rivals are totally non-foreclosed.
333
(234) The Notifying Party submits that, in the event of total foreclosure, the number of
close competitors mentioned by the Commission (in paragraph 308 of the SO) would
represent [Regulatory submission] of the market, and other small players around
[Regulatory submission]. The Notifying Party adds that the SO relies on the
assumption that the large non-foreclosed competition would be insufficient to
restrain Kustomer which is implausible and contradicts the Commission’s conclusion
in Microsoft/LinkedIn, where a foreclosure of less than 30% of the CRM market was
considered not to affect a sufficiently important proportion of Microsoft’s
326
SO Response, paragraph 4.21. In Annex 3 to the SO Response, paragraph 2.99 states that “Zendesk has
1,314 integrations compared to only 61 for Kustomer. This makes Zendesk businesses particularly
unlikely to switch away from Zendesk, as a substantial proportion of Zendesk’s business customers are
likely to have integration needs which are unmet by many of the potential alternative CSCRM providers
(and in particular, which are unmet by Kustomer).”
327
LoF Response, paragraphs 3.1-3.19.
328
Form CO, paragraph 6.109.
329
Article 6(1)(c) Response, paragraphs 3.36 and 3.37.
330
Article 6(1)(c) Response, paragraphs 3.38-3.41.
331
Article 6(1)(c) Response, paragraphs 3.42-3.44.
332
Article 6(1)(c) Response, paragraphs 3.45 and 3.46.
333
SO Response, paragraph 5.1. This argument as well as the ones presented in the following paragraphs
were also in the Notifying Party Submission on Foreclosure Issues dated 10 October 2021.
69
competitors to result in a price increase or reduction of market incentives to
innovate.
334
(235) The Notifying Party further submits that the SO fails to explain how an (undefined)
small group of smaller customer service and support CRM providers and the
innovation they bring stand out from their competitors in terms of impact on
competition. The Notifying Party argues the examples of innovation provided in the
SO were all introduced by larger players first (and one by Kustomer itself).
335
The
Notifying Party also submits that industry reports do not support the Commission’s
conclusion that smaller players have a disproportionate impact on competition
through innovation (as the Gartner’s “Magic Quadrant” Report 2020 and 2021
editions only refer to one smaller player as a “Visionary” and confirm that large
players are innovative).
336
Further, the Notifying Party argues there is also some out
of market innovation that the Commission does not take into account.
337
(236) The Notifying Party submits the SO does not explain why partially foreclosed rivals
could not continue to innovate on the parameters for competition indicated in the
results of the market investigation.
338
The Notifying Party also argues the SO does
not consider whether there are any other factors (beyond smaller customer service
and support CRM providers’ innovative role) to assess whether the foreclosed “close
competitors” play a significant competitive role in the market (such as their pricing
policies; market share evolution; number of new customers won after the launch of
innovations etc.).
339
(237) The Notifying Party adds that the market feedback the Commission received
confirms that the CRM market is highly competitive and is expected to remain so
post-Transaction.
340
(238) In its LoF Response, the Notifying Party submits that the additional evidence
presented by the Commission is not sufficient nor factually accurate, reliable and
consistent. In this regard, the Notifying Party submits that the feedback from larger
customer service and support CRM providers shows they serve SMB customers
341
and develop key innovations for both enterprises and SMB customers
342
but that the
Commission solely relies on the responses of two small CRM players.
343
(239) Furthermore, the Notifying Party argues that even though they are key innovators,
larger CRM players are necessarily excluded from Gartner’s “Cool Vendors” report
as well as from the “Visionary” category in Gartner’s “Magic Quadrant” report,
which require the vendor to be a small or not yet complete or established company,
respectively.
344
The Notifying Party also submits that many of the businesses listed
in the 2021 lists of CRM start-ups and companies in several EEA countries published
in the website Beststartup.eu (which are cited by the Commission in the LoF) no
longer exist or are not a CRM (they are either unrelated businesses or out-of-market
334
SO Response, paragraphs 5.10(A)-(C).
335
SO Response, paragraphs 5.16-5.17.
336
SO Response, paragraphs 5.17(B).
337
SO Response, paragraphs 5.17(C).
338
SO Response, paragraph 5.20.
339
SO Response, paragraph 5.24.
340
SO Response, paragraph 5.42.
341
LoF Response, paragraph 4.5(A).
342
LoF Response, paragraph 4.5(B)(ii).
343
LoF Response, paragraph 4.5(B)(i).
344
LoF Response, paragraph 4.5(C)(i).
70
actors that may integrate with a CRM).
345
Moreover, the Notifying Party submits that
the growth of the SMB segment (in the CRM software market) has no bearing on the
impact of smaller CRM players on innovation.
346
The Notifying Party reiterates the
alleged examples of innovations by small CRM players focussed on SMB customers
referred by the Commission are imitative and not innovative.
347
The Notifying Party
adds that the internal documents cited by the Commission [Details of internal
documents describing certain CRM software providers].
348
(240) Finally, according to the Notifying Party, the Transaction will create the following
efficiencies and customer benefits: (i) increase take-up of messaging for B2C
communications, (ii) drive innovation and competition in the B2C messaging space;
(iii) drive innovation and competition in the CRM market, and (iv) facilitate access
to higher quality ads services for businesses.
349
7.3.2.2. Commission’s Assessment
(241) For the reasons set out below and based on the results of the market investigation, the
Commission has reached the conclusion that it is likely, and there is even a strong
probability, that Meta (formerly Facebook) has the ability and will have the incentive
to foreclose competing providers of customer service and support CRM software by
restricting or degrading API access to its messaging channels. Moreover, if Meta
(formerly Facebook) engaged in such a foreclosure strategy, it is likely, and there is
even a strong probability, that the anticompetitive effects would be significant.
(a) Ability to foreclose
(242) For the reasons set out below, the Commission has reached the conclusion that Meta
(formerly Facebook) would have the ability to engage in a foreclosure strategy with
regard to competing providers of customer service and support CRM software. First,
API access to OTT B2C messaging channels is an important input for customer
service and support CRM software providers (and their business customers). Second,
Meta (formerly Facebook) has market power on the EEA-wide, if not worldwide,
upstream OTT B2C messaging market, and within a broader EEA-wide or
worldwide upstream market for B2C communications overall. Third, Meta (formerly
Facebook) has the ability — both technically and contractually — to restrict or
degrade API access to its messaging channels, including the ability to target such a
foreclosure strategy at Kustomer’s close rivals, which are typically smaller CRM
software providers focussed on serving SMBs.
(a.i) Important input for the downstream market
(243) According to the Non-Horizontal Guidelines, input foreclosure may raise
competition problems only if it concerns an important input for the downstream
product. This is the case, for example, when the input concerned represents a
significant cost factor relative to the price of the downstream product. Irrespective of
its cost, an input may also be important for other reasons. For instance, the input may
be a critical component without which the downstream product could not be
manufactured or effectively sold on the market, or it may represent a significant
345
LoF Response, paragraph 4.5(D).
346
LoF Response, paragraph 4.5(E).
347
LoF Response, paragraph 4.5(F).
348
LoF Response, paragraph 4.5(G).
349
Form CO, paragraph 9.1.
71
source of product differentiation for the downstream product. It may also be that the
cost of switching to alternative inputs is relatively high.
350
(244) Customer service and support software tools integrate a wide range of different B2C
communication channels, both traditional channels such as phone and email, as well
as more modern channels such as OTT messaging channels.
351
As outlined by a
leading CRM provider, “for most customers, the various B2C communications
channels are complementary and allow customers a range of options for
communicating with end-customers, and are thus not substitutable.”
352
Similar views
were expressed by businesses that use CRM software tools, with one such business
indicating that its “customers are demanding service via multiple channels and we
need to respond to that need.”
353
As a result, and as set out in Section 6.2.1 above,
there is not one overall market for B2C communication services in which phone,
SMS, Instagram, Twitter, etc. are substitutable for one another. Rather, there is
evidence to support the existence of a standalone product market for OTT B2C
messaging channels, excluding at least phone, email and SMS.
(245) The results of the market investigation were inconclusive as to whether webchat
should be included or excluded
354
so the Commission has — on a conservative basis
— carried out its assessment on the basis that webchat does form part of the relevant
market for OTT B2C messaging channels. Likewise, the market investigation and
other evidence — [Business Plans] — were mixed on whether rich cellular-based
B2C messaging channels, i.e., Google RCS and Apple Business Chat, and OTT B2C
messaging channels are substitutable with OTT B2C messaging channels.
355
The
Commission has nonetheless carried out its assessment on the basis that rich cellular-
based messaging channels and OTT B2C messaging channels form part of the same
relevant product market, in particular since this does not impact the Commission’s
conclusions given the current minimal share and uncertain future growth of such
channels, i.e., Google RCS and Apple Business Chat, as explained further in Section
7.3.2.2.(a.ii) below. In addition to the above, and contrary to what is argued by the
Notifying Party, the market investigation showed that API access to Meta (formerly
Facebook)’s messaging channels (which account for a large and important part of the
overall OTT B2C messaging market, as shown in Section 7.3.2.2.(a.ii) below) is
considered an important input for customer service and support CRM software
providers and their business customers.
356
In that regard, the vast majority of CRM
software providers that responded to the market investigation indicated that it would
be difficult to effectively sell their customer service and support CRM product
without such access.
357
One business customer stated that Meta (formerly
Facebook)’s channels are “must-have features” for communicating with its end
customers
358
, while a CRM provider said that “Facebook Messenger and WhatsApp
are the leading players in certain geographies where we sell our CRM software.
Having the ability to integrate with these B2C communication channels via APIs
350
See Non-Horizontal Guidelines, paragraph 34.
351
Replies to questionnaire 1 to CRM software providers, questions 49.1, 49.2, and 50.1.
352
Reply to questionnaire 1 to CRM software providers, question 32.1 (ID: 846).
353
Reply to questionnaire 5 to business customers, question 12.1 (ID: 1339).
354
See Section 6.2.1.3. above.
355
See Section 6.2.1.3. above.
356
Replies to questionnaire 1 to CRM software providers; replies to questionnaire 4 to business customers,
questions 50.1 and 61.1.
357
Replies to questionnaire 1 to CRM software providers, questions 51, 52 and 53.
358
Reply to questionnaire 4 to business customers, question 61.1 (ID: 527).
72
would allow us to effectively compete on the market for CRM software in those
geographies.”
359
Indeed, the Parties were unable to provide meaningful examples of
CRM software providers that integrate B2C communications channels but none of
Meta (formerly Facebook)’s messaging channels (the CRM software providers they
listed had very limited revenues and a combined share of only c. 0.1%).
360
Moreover,
the majority of business customers
361
and the vast majority of CRM software
providers
362
that expressed a view indicated that they consider that the integration of
Meta (formerly Facebook)’s messaging channels will become even more important
for businesses’ choice of customer service and support CRM provider in the coming
five years, with one business customer explaining that, in its view, “a growing
number of people will use Facebook’s messaging channels in the upcoming years.
Due to this and the goal of companies being to offer different and more innovative
ways to communicate with customers, an integration of Facebook's messaging
channels will become more important.”
363
(246) Contrary to what is argued in the SO Response, the Commission has taken into
account not only evidence from CRM software providers, i.e., the targets of any
foreclosure strategy, but also evidence from business customers to reach the
conclusion that that Meta (formerly Facebook)’s B2C messaging channels are
important inputs to CRM software providers. Indeed, this conclusion is supported by
the following evidence from business customers’ responses to the Commission’s
market investigation. Overall, in response to the question “How important is access
to Facebook’s messaging channels, through API-access or otherwise, to your
business now and in the near future (for communications with your customers)”, a
sizeable portion of business customers that expressed a view (30%) indicated that
they are essential or important, while a further 60% indicated that they are beneficial.
This compares to only 10% of business customers expressing a view which
considered that they were not important.
364
This was further supported by the
following narrative responses by businesses: “Facebook brings better user
engagement for customer care,”
365
, which is a particularly important observation
given the relevant downstream market is CRM software designed for customer
service and support, i.e., customer care, and “Facebook connects us to a larger
audience and allows us to interact with them, while also giving users a possibility to
contact us more conveniently.”
366
The importance of Meta (formerly Facebook)’s
channels did not mean they had to be used by businesses for all of their interactions
with their customers. As one business customer stated, “The Facebook messaging
channels are mainly used for interactions with community members, which is a small
but still important part of the customer base.”
367
Similarly, in response to the
question “To what extent is integration with Facebook’s messaging channels
359
Reply to questionnaire 1 to CRM software providers, question 50.1 (ID: 759).
360
Parties’ reply to RFI 1, question 30(b)(i).
361
Replies to questionnaire 5 to business customers, question 12.2.
362
Replies to questionnaire 7 to CRM software providers, question 38. This shift may have even been
accelerated by the Covid-19 pandemic, with one CRM software provider noting that “A lot of
traditional / physical small businesses moved to WhatsApp worldwide since COVID-19. I think this
trend will continue.” See reply to questionnaire 7 to CRM software providers, question 38.1 (ID: 1633).
363
Reply to questionnaire 5 to business customers, question 12.2.1 (ID: 1366).
364
Replies to questionnaire 4 to business customers, question 61. Note that Crisp’s response has been
excluded from the above, as Crisp is a provider of customer service and support CRM services.
365
Replies to questionnaire 4 to business customers, question 61.1 (ID: 842).
366
Replies to questionnaire 4 to business customers, question 61.1 (ID: 776).
367
Replies to questionnaire 4 to business customers, question 61.1 (ID: 830).
73
(including the ability to integrate with such channels or the quality of the
integration) an important factor in a business’ choice of customer service and
support CRM provider)”, a sizeable portion of business customers that expressed a
view (40%) indicated that such integrations are essential or important, while a further
48% indicated that they are beneficial. This compares to only 12% of business
customers expressing a view which considered that they were not important.
368
Again, many business customers that expressed a view echoed this view in narrative
responses: “We use Facebook channels so having them integrated would be
helpful”
369
, “Many of our customers do like to utilize Facebook products to
communicate”
370
, “Facebook grants access to relevant audiences (via Facebook,
Instagram, WhatsApp). From a company perspective, this enables more touch points
with customers. Having this data integrated allows for a more complete picture of
the customer and thus a better service (eg a conversation started by a customer via
Facebook can be continued via email)”
371
, “a lot of our users are active on social
media so it is important to be able to interact with them on a channel they are
familiar with and use frequently”
372
and “Customers are using this channel
frequently so they expect their loved brand to do it as well.
373
(247) Asked whether they expected the integration of Meta (formerly Facebook)’s
messaging channels to become more or less important for a business’ choice of
customer service and support CRM provider in the coming 5 years, the majority of
business customers that expressed a view (53%) indicated that such integrations
would become more important. This compares to only 12% of business customers
expressing a view which considered that this would become less important.
374
This
was further supported by the following narrative responses by businesses: “Social
media is mature and considered as a serious platform for business, therefore it is
important to integrate messaging channels to be in direct contact with your audience
”
375
and “A growing number of people will use Facebook’s messaging channels in
the upcoming years. Due to this and the goal of companies being to offer different
and more innovative ways to communicate with customers, an integration of
Facebook's messaging channels will become more important.”
376
In the LoF
Response, the Notifying Party argued that the Commission failed to take into account
responses of market participants that were uncertain about the future growth of Meta
(formerly Facebook)’s channels, such as Microsoft which said it would be “difficult
to predict” whether Meta (formerly Facebook)’s channels would grow in importance
over the coming five years.
377
However, the Commission did take account of all
responses to the market investigation in a holistic manner before drawing
conclusions. Where respondents express uncertainty or indicate that they do not
know the answer to a market investigation question, the Commission typically
368
Replies to questionnaire 5 to business customers, question 12.
369
Reply to questionnaire 5 to business customers, question 12.1 (ID: 1369).
370
Reply to questionnaire 5 to business customers, question 12.1 (ID: 1516).
371
Reply to questionnaire 5 to business customers, question 12.1 (ID: 1342).
372
Reply to questionnaire 5 to business customers, question 12.1 (ID: 1624).
373
Reply to questionnaire 5 to business customers, question 12.1 (ID: 1490).
374
Replies to questionnaire 5 to business customers, question 12.2.
375
Reply to questionnaire 5 to business customers, question 12.2.1 (ID: 1428).
376
Reply to questionnaire 5 to business customers, question 12.2.1 (ID: 1366).
377
LoF Response, paragraph 2.4. The Commission notes that Microsoft separately indicated that it
“expects digital channels to continue to increase in popularity [as a means of B2C communication] as
more customers spend more time online.” Reply to questionnaire 1 to CRM software providers,
question 47.2 (ID: 726).
74
weighs the responses which take a concrete position in the first instance, and then
does a control screening of qualitative reasons provided by respondents, including
those that did not express a concrete view, such as Microsoft in this case. In the
control screening that was performed in the present case (as set out in the paragraph
246 above), there were however a larger number of quantitative and well-reasoned
qualitative responses pointing to Meta (formerly Facebook)’s B2C messaging
channels remaining important or growing in importance in the coming five years,
such as the response of a rival B2C communication service provider which indicated
that “the Facebook B2C communications channels are expected to continue to rule
the market in 5 years”.
378
(248) In view of the fact that customer service and support CRM software providers
consider OTT B2C messaging channels, and in particular Meta (formerly
Facebook)’s B2C messaging channels, to be must-have inputs in order to effectively
compete downstream, it is not necessary, contrary to what the Notifying Party
argues, to also show that such inputs constitute a “significant cost factor” in
proportion to the overall cost or price of the downstream product.
379
The importance
OTT B2C messaging channels, and in particular Meta (formerly Facebook)’s B2C
messaging channels, for CRM software provider is also the same irrespective of
whether the downstream market is defined as EEA-wide or worldwide in scope.
(249) The Notifying Party also argues that OTT B2C messaging channels, and in particular
Meta (formerly Facebook)’s B2C messaging channels, do not constitute important
inputs since customer service and support CRM software products integrate several
OTT B2C messaging channels, as well as other B2C communication channels such
as phone, email and SMS, enabling businesses to quickly and easily switch between
channels at little cost. For the reasons outlined in Section 6.2.1.3 above, OTT B2C
messaging channels and other B2C communication channels do not form part of the
same relevant product market, as market participants see them as complementary
rather than substitutable. Pointing to the fact that a CRM tool integrates a range of
different B2C communication channels ignores the fact that businesses, and their
customers, may not view these as substitutable.
(250) In its SO Response, the Notifying Party further argues that the temporary outage on 4
October 2021 of all of Meta (formerly Facebook)’s messaging channels is “[Details
regarding Meta’s temporary outage].”
380
For a range of reasons, the Commission
considers that this temporary outage offers little to no indication of business
customer switching behaviour in the event of a foreclosure strategy aimed at
competitors of Kustomer. First, the likelihood of switching or the substitutability of
two products cannot be concluded from a temporary or transitory, and total,
unavailability of one of the products. It would instead be necessary to look at a
lasting, i.e., non-transitory, change in the relative price or quality of the products in
order to draw any clear conclusions. In any case, the Commission notes that [Details
regarding Meta’s temporary outage]
381382
. Second, during the outage all end users, all
businesses and all CRM software provides, including Kustomer, lost access to Meta
(formerly Facebook)’s messaging channels. There was therefore no control group
378
Reply to questionnaire 8 (Commitments Market Test), question 19.1 (ID: 2608).
379
Indeed, the Non-Horizontal Guidelines, paragraph 34, refer to this as just one “example” of a situation
where an input may be important for the downstream product.
380
SO Response, paragraph 3.8(D).
381
SO Response, Annex 3, paragraph 3.10.
382
SO Response, Annex 3, paragraph 3.10.
75
that remained using Meta (formerly Facebook)’s channels during the outage period
against which any comparisons could be made. This outage therefore does not shed
light on how businesses and end users would react in a situation where end users,
businesses and Kustomer retained access to Meta (formerly Facebook)’s channels,
but where only competing CRM software providers did not. Third, Meta (formerly
Facebook) itself acknowledged that the outage demonstrated how much businesses
rely on its channels for B2C communications, with the head of WhatsApp tweeting
that the outage was “a humbling reminder of how much people and organizations
rely on our app every day”
383
(emphasis added).
(251) Similarly, even if customer service and support CRM software providers integrate
third party OTT B2C messaging channels (e.g., Twitter or Google RCS), it is clear
from Section 7.3.2.2.(a.ii) below that Meta (formerly Facebook) has market power in
this market, since based on the shares, penetration rates, and the relative positions of
Meta (formerly Facebook) and other OTT B2C messaging channels, Meta (formerly
Facebook) accounts for a large and important part of the OTT B2C messaging
market. This, combined with the observations by market participants, including a
major CRM software provider, that, “in the customer service segment of CRM, it is
necessary to be where the customer is”
384
(i.e., to integrate the channels that
customers already use for P2P communications), and “the best customer experience
is one where everything happens on the same app … to change to another messaging
channel is to provide a bad customer experience”
385
, shows that the technical ability
to switch to other OTT B2C messaging channels, or other B2C communications
channels, does not detract from the importance of Meta (formerly Facebook)’s B2C
messaging channels in particular (and for OTT B2C messaging channels in general)
as inputs for customer service and support CRM software providers.
(252) Accordingly, the Commission concludes that the EEA-wide, if not worldwide,
upstream market for OTT B2C messaging channels (of which Meta (formerly
Facebook)’s channels represent a large and important part, as shown in
Section 7.3.2.2.(a.ii) below) constitutes an important input in order to be able to
effectively compete in the downstream customer service and support CRM software
market. In support of this conclusion, all CRM software providers that responded to
the market investigation indicated that they currently integrate OTT B2C messaging
channels into their CRM software, including in each case at least one of Meta
(formerly Facebook)’s B2C messaging channels.
386
(253) Finally, the Commission’s investigation has revealed that the market for OTT B2C
messaging channels is growing in importance relative to more traditional means of
B2C communications such as phone and email. This supports the conclusion that the
market for OTT B2C messaging channels will in fact become an increasingly
important input for companies active in the downstream customer service and
support CRM market. Indeed, the market participants that expressed a view
unanimously considered that the integration of B2C messaging channels into
customer service and support CRM software solutions will become a more important
factor in a business’ choice of customer service and support CRM software in the
383
See https://twitter.com/wcathcart/status/1445202016049844229, last accessed on 29 November 2021
(ID: 2454).
384
Agreed minutes of the call of 18 May 2021 with SAP, paragraph 15 (ID: 2008).
385
Agreed minutes of the call of 14 June 2021 with Twilio, paragraph 11 (ID: 2029).
386
Replies to questionnaire 1 to CRM software providers, question 46.
76
coming 5 years.
387
All else being equal, the Commission considers that Meta
(formerly Facebook)’s channels will likewise further increase in importance as inputs
for customer service and support CRM software providers in the future.
388
One CRM
software provider indicated that “there has been a generalized trend away from
email and phone to digital channels, social media in particular”,
389
while another
respondent noted that they have “seen a growth of conversations on Social
Messaging channels first dominated by Facebook Messenger and then WhatsApp”.
390
In a similar vein, a provider of B2C messaging services expressed the view, in
respect of B2C messaging channels, that “it seems intuitive that integration of
methods to communicate with consumers will become an increasingly important
factor in a business's choice of customer service and [support] CSM software in the
coming years.”
391
(254) On the basis of all these elements, and for the purposes of this Decision, the
Commission has reached the conclusion that the EEA-wide, if not worldwide
upstream market for OTT B2C messaging channels constitutes an important input for
companies active in the EEA-wide, if not worldwide, downstream market for
customer service and support CRM software.
(255) This conclusion would be the same for a potential broader EEA-wide or worldwide
downstream market for CRM software overall because the finding of such a potential
broader market would be based on supply-side substitutability arguments (as set out
in Section 6.1.1.3 above) and customer service and support CRM software
constitutes an important segment of an overall market for CRM software.
392
Therefore API access to OTT B2C messaging channels (and Meta (formerly
Facebook)’s B2C messaging channels in particular) would be a similarly important
input on the basis of such a broader market.
(256) This is moreover the case because, in the downstream customer service and support
CRM market it is important, if not essential, to carry a broad range of
communication channels to enable “omnichannel” communications between
businesses and their customers. As a major CRM software provider stated, “in the
customer service segment of CRM, it is necessary to be where the customer is, so
being present on all the channels and to have an ‘omnichannel’ offering is the
minimum entry requirement for the customer service CRM market. For SAP, access
to Facebook APIs is essential as it is for any CRM software provider active in the
customer service CRM market.”
393
This conclusion would not change if the customer
service and support CRM software market, or the overall CRM software market,
were to be segmented based on (i) business customer size, (ii) mode of deployment
387
Replies to questionnaire 6 to B2C communication service providers and online ads service providers,
question 14.
388
Several market participants that expressed a view also considered, aside from the growing in
importance of B2C messaging overall, that Meta (formerly Facebook)’s messaging channels
specifically will become more important for businesses’ choice of customer service and support CRM
provider in the coming five years. See replies to questionnaire 7 to CRM software providers,
question 37.
389
Reply to questionnaire 1 to CRM software providers, question 47.1 (ID: 726).
390
Reply to questionnaire 1 to CRM software providers, question 47.1 (ID: 762).
391
Reply to questionnaire 6 to B2C communication service providers and online ads service providers,
question 14.1 (ID: 1604).
392
Customer service and support software accounts for 36% of the overall CRM software market by
revenues in 2020, both on a worldwide basis and in the EEA. See Form CO paragraph 6.47
Tables 6.2-6.5.
393
Agreed minutes of the call of 18 May 2021 with SAP, paragraph 15 (ID: 2008).
78
stated by one respondent to the market investigation, “WhatsApp and Facebook are
two of the more popular platforms in the EEA, which could make it difficult for a new
or existing solution to gain market share”
398
since, in the customer service and
support CRM software market, it is necessary to integrate the channels that
customers already use for P2P communications (i.e., “to be where the
customer is”
399
).
(260) Remainder of the market is fragmented. The remainder of the OTT B2C messaging
market is rather fragmented. In the Form CO, the Parties indicated they were unable
to provide data for OTT B2C messaging on a standalone basis (or for any segments
of an overall B2C communications market).
400
However, indications — based on
[…], the Commission’s own investigation on the basis of publicly available
information and the results of the market investigation — are that other OTT
messaging channels (such as Twitter, Viber, WeChat, Telegram or Line) would each
have a modest share of the market for OTT B2C messaging services, and indeed that
the combined share of all other OTT messaging channels (excluding webchat) would
be at least three times smaller than the standalone share of Meta (formerly Facebook)
in the OTT B2C messaging market, both worldwide and in the EEA.
401
(261) It even appears that the market penetration rates of other OTT messaging channels is
not only lower than those of Meta (formerly Facebook)’s channels but also less
consistent across regions, in particular within the EEA. For example, Twitter, which
is likely to be the largest OTT B2C messaging channel after Meta (formerly
Facebook), appears to have a far lower market penetration than Meta (formerly
Facebook)’s channels. In the U.S., its main market, it has a market penetration of less
than 20% according to public sources
402
, and it would be reasonable to deduce that it
would therefore have an even lower market penetration in the EEA. Other OTT
messaging channels that are used for B2C communications (and which certain CRM
software providers indicated that they integrate into their offerings) may have
meaningful market penetration only in one or a small number of Member States, or
may have a primary focus outside the EEA. For example, Viber is more widely used
in Eastern Europe and the Middle East, WeChat is primarily used in China, Line’s
main market is Japan and Telegram appears to be most widely used outside the EEA
(e.g., in Russia).
403
(262) Webchat is a collection of individual channels with minimal standalone shares. To
the extent webchat may form part of the same market as OTT B2C messaging, it is
important to recall that webchat is not a single product but rather an umbrella term
398
Reply to questionnaire 3 to B2C communication service providers and online ads service providers,
questions 54.1 (ID: 886).
399
Agreed minutes of the call of 18 May 2021 with SAP, paragraph 15 (ID: 2008).
400
Form CO, Table 6.7.
401
According to the Notifying Party’s calculations, Meta (formerly Facebook)’s channels accounted for
[5-10%] of the worldwide B2C communications market in 2020, while all other OTT channels
accounted for just [0-5%] (Form CO, Figure 6.1, and Table 3 above). In a market for OTT B2C
Messaging (i.e., Table 5 above, which excludes phone and email), Meta (formerly Facebook)’s own
OTT B2C messaging channels account for more than [40-50%] of the market and (based on the
breakdown provided by the Notifying Party in the Form CO, Figure 6.1, for “Other messaging”) the
combined share of all non-Meta (formerly Facebook) OTT B2C messaging channels (excluding
webchat) accounts for less than [10-20%] the market, both on a worldwide basis and in the EEA.
402
See https://www.statista.com/statistics/183466/share-of-adult-us-population-on-twitter/, last accessed
on 15 December 2021 (ID: 2810).
403
See, for example, https://www.similarweb.com/corp/blog/research/market-research/worldwide-
messaging-apps/, last accessed on 15 December 2021 (ID: 2809).
79
covering a plethora of different individual webchat applications, each of which may
only enable communication with a specific business and — [Customer
Information]
404
— may have the disadvantage of requiring end customers in certain
circumstances to download a new app on mobile devices or tablets to speak to a
specific business (as opposed to Meta (formerly Facebook)’s channels, which end
customers may typically have already downloaded, e.g., for P2P communication
purposes, and which can be used more broadly to communicate with businesses).
Webchat applications would therefore individually have a minimal market share and
be unlikely as a result to have market power in a market for OTT B2C messaging
services.
(263) Google RCS and Apple Business Chat have minimal presence today with growth
highly uncertain. The picture does not significantly change if Google RCS and Apple
Business Chat are included, which the Commission understands currently have very
low market penetration rates in the EEA and there is significant uncertainty
regarding the likelihood and extent of their future growth.
405
(264) Evidence of the current negligible penetration rates and usage of Google RCS and
Apple Business Chat includes (i) direct evidence from Google and Apple,
406
(ii)
[Customer information]
407
and (iii) responses to the Commission’s market
investigation.
408
Accordingly, including these channels within the relevant market,
which the Commission has done in a conservative approach, will not have a material
impact on the assessment as they would account for a minimal share of the relevant
market, irrespective of the precise definition.
(265) It is moreover highly uncertain if such channels, in particular Google RCS (which is
currently only available as a built-in functionality on certain android devices and,
unlike OTT B2C messaging channels, cannot be installed as applications on most
mobile devices), will become more important in the coming years, since, as an OTT
B2C messaging provider indicated, “Google needs to build up different partnerships
with, for example, Apple, and carriers [i.e., mobile network operators], which takes
many years. The process is very complex compared to working with an OTT
messaging channel that works on any device, and on any network.”
409
Contrary to
what is argued in the SO Response
410
, the Commission has therefore factored the
potential future growth of Google RCS and Apple Business Chat in its assessment.
(266) In addition, Google, having responded “Don’t know / It depends” to the Phase II
market investigation questionnaire question of whether its own B2C messaging
channel (and that of Apple) would “become a more important means of B2C
404
Article 6(1)(c) Response, page 23 Figure 3.1 [Meta assessment of channel usage patterns by end
customers in communicating with businesses].
405
Replies to questionnaire 1 to CRM software providers and to questionnaire 4 to business customers,
questions 46 and 57.
406
Google said it was unable to estimate the penetration rate of RCS in the EEA, despite stating that it was
available in all EEA member states (Google’s reply to RFI dated 15 November 2021 (ID: 2307)). Apple
indicated that despite being “available in every EEA member state”, Apple Business Chat has a very
low penetration rate in the EEA today, namely “less than 50,000 … daily conversations” in 2020
(compared to over [Customer Information] daily conversation on Meta (formerly Facebook)’s OTT
B2C messaging channels in 2020 ) (Apple’s reply to RFI dated 15 November 2021 (ID: 2319)).
407
Article 6(1)(c) Response, page 23 Figure 3.1 [Meta internal assessment of RCS].
408
Replies to questionnaire 5 to business customers, question 5.2.1 (ID: 1484) (“RCS is a new technology
that is not yet fully deployed and the effectiveness of the RCS channel is not yet fully demonstrated.”).
409
Agreed minutes of the call of 16 September 2021 with Viber, paragraph 4 (ID: 2033).
410
SO Response, paragraph 3.8(C).
80
communication in the coming 5 years”
411
, later explained the reasons for its
uncertainty, such as the fact that the adoption of Google RCS “faced several
headwinds. The most important of these headwinds was the relatively limited
adoption by carriers, including in the EEA.” Google acknowledged that “the carrier-
led approach [which it has chosen to adopt] can lead to slow adoption,
fragmentation which produces operational complexity (e.g., brands and their
partners may need to go through multiple different verification processes depending
on the number of self-managed carriers in a given market) and potentially
misaligned ecosystem incentives as carriers may not be incentivised to promote RCS
over SMS.” It went on to note that “Despite being available for over 10 years, the
RCS standard itself has not achieved widespread adoption” (emphasis added).
Finally, Google indicated that it “hopes that its efforts will help accelerate the
adoption of the RCS standard in EEA” but has not been able to provide concrete
estimates of the extent of any such accelerated adoption, if it were to materialise.
412
(267) Apple expressed a similar level of uncertainty as regards the future growth of Apple
Business Chat. Apple indicated that despite being “available in every EEA member
state”, Apple Business Chat has a very low penetration rate in the EEA today,
namely “less than 50,000 … daily conversations” in 2020 (compared to over
[Customer information] daily conversation on Meta (formerly Facebook)’s OTT B2C
messaging channels in 2020
413
). In other words, based on Meta (formerly
Facebook)’s [40-50%] share of the EEA OTT B2C messaging market in 2020,
Apple’s share would be less than c. 0.2%. Apple also pointed to numerous “technical
requirements to participate in Apple Business Chat” including creating a “Messages
for Business account”, for which the business in question “must have an
asynchronous messaging platform staffed by live agents during business hours [and]
must use an Apple-approved [Messaging Service Provider]”.
414
In the Commission’s
view, given the current low adoption rate and the potential cumbersome steps
required to adopt Apple Business Chat for businesses of different sizes, it is
unsurprising that Apple indicated that it was “unable to speculate as to whether
Business Chat will gain "widespread adoption" in the EEA in the timeframe
indicated [i.e., within the coming five years].”
415
(268) Similar uncertainty about the future growth of rich cellular-based B2C messaging has
been expressed by respondents to the Commission’s market investigation. One such
respondent indicated that “To our knowledge, this technology has been available [i.e.
Google RCS has been available for several years] and that switch has not
happened.”
416
In the LoF Response, the Notifying Party argued that the Commission
ignored the fact that Google RCS is available in all EEA member states.
417
However,
this is rather suggests that the current low penetration rate and share of Google RCS
in the OTT B2C messaging market may not be explained by lack of availability.
Another respondent said that rich cellular-based B2C messaging channels would
411
Reply to questionnaire 6 to B2C communication service providers and online ads service providers,
question 8 (ID: 1667).
412
Google’s reply to RFI dated 15 November 2021 (ID: 2307).
413
Form CO, Table 6.8.
414
See https://register.apple.com/resources/messages/messaging-documentation/, last accessed on 29
November 2021 (ID: 2462).
415
Apple’s reply to RFI dated 15 November 2021 (ID: 2319).
416
Replies to questionnaire 6 to B2C communication service providers and online ads service providers,
question 8.1 (ID: 1513).
417
LoF Response, paragraph 2.3(B).
81
likely become even less important in the coming five years because it considered that
uptake will be driven primarily by the penetration rate in P2P communications,
which Google RCS and Apple Business Chat lack: “No it will become less important,
the prevalance [sic] and uptake of new messaging channels and their richness comes
to the fore-front. B2C will go where the highest denominator of their customer base
will be - this is most likely social media based channels for public communications
(Twitter, Instagram), and then direct messaging (Whatsapp, Viber, etc) for private
communications”
418
(emphasis added). This response makes a correlation between
P2P penetration rates and B2C penetration rates, a point also made by other market
participants. In the LoF Response, the Notifying Party argues that the operating
system penetration rates of Google and Apple should be similarly seen as providing a
foundation for a strong share in the B2C messaging market.
419
However, the
Commission considers that there is at best a weak correlation between the use of an
operating system and the use of a messaging channel on the operating system for
B2C communications, especially relative to the stronger correlation between the use
of a messaging channel for personal or P2P communications and the use of the same
messaging channel for B2C communications. If anything, the fact that Google RCS
and Apple Business Chat are limited to Google’s and Apple’s respective operating
systems may represent a significant disadvantage compared to OTT B2C messaging
channels which are not restricted to a particular operating system.
(269) Even assuming, in a conservative approach, that Google RCS and Apple Business
Chat experience significant growth in the coming years relative to their positions
today, and ignoring the structural ceiling on such growth as a result of the penetration
rates of Google and Apple in operating systems, given such channels each have a
minimal market presence today, any such growth would be expected to result in only
a modest share of the overall OTT B2C messaging market given the fact that Google
RCS and Apple Business Chat each have a minimal market presence today. Meta
(formerly Facebook)’s channels on the other hand, account for over [40-50%] of the
OTT B2C messaging market today, have experienced significant growth in the past
three years and are expected to continue to grow in the coming years. It is therefore
improbable that any future growth of rich cellular-based B2C messaging channels,
even if substantial relative to their position today, would result in Meta (formerly
Facebook)’s B2C messaging channels ceasing to represent a large and important part
of the overall OTT B2C messaging market. For the same reason, it is equally
improbable that any such growth of rich cellular-based B2C channels would
undermine the Commission’s conclusion that Meta (formerly Facebook)’s OTT B2C
messaging channels have, and will continue in the coming years to have, market
power in the upstream market for OTT B2C messaging services, as well as in the
broader upstream market for B2C communications overall.
(270) Interim conclusion based on multiple sources of evidence. In circumstances where, as
one market participant indicated, “Businesses increasingly need and expect to reach
and be available to their customers in popular OTT messaging channels, such as
WhatsApp and Facebook”,
420
an important element in usefulness of a messaging
channel for B2C communications (and therefore as an input for customer service and
418
Replies to questionnaire 6 to B2C communication service providers and online ads service providers,
question 8.1 (ID: 1658).
419
LoF Response, paragraph 2.4.
420
Reply to questionnaire 3 to B2C communication service providers and online ads service providers,
questions 55.1 (ID: 886).
82
support CRM software providers) is the popularity (as measured through the proxy
of market penetration) of that channel. By contrast to the other channels listed above,
as can be seen from Table 12 above and in line with the observations of one market
participant, “FB, Instagram and Whatsapp have a strong penetration rate in most
countries, making them a critical channel for business which wish to use B2C
messaging channels.”
421
(271) In view of the above, and contrary to what is argued in the SO Response, the
Commission has not simply inferred that Meta (formerly Facebook) has market
power from it “relatively higher share of B2C OTT communications” without any
other assessment.
422
Rather, based on the shares, penetration rates, and the relative
positions of Meta (formerly Facebook)’s and other OTT B2C messaging channels, it
is apparent that Meta (formerly Facebook) accounts for a large and important part of
the OTT B2C messaging market, both in the EEA and at worldwide level. On the
basis of all these elements, the Commission has reached the conclusion that Meta
(formerly Facebook) has market power in the EEA-wide, if not worldwide, upstream
market for OTT B2C messaging channels.
(272) Conclusion holds for the broader overall market for B2C communication channels.
Further, in view of the importance of Meta (formerly Facebook)’s messaging
channels for B2C communications more generally and the complementary nature of
different channels (as confirmed by several major CRM software providers in
response to the Commission’s market investigation
423
), the Commission also
concludes that Meta (formerly Facebook) has market power even on a broader EEA-
wide, if not worldwide, market for B2C communications channels. This is the case in
particular because, while Meta (formerly Facebook)’s own B2C messaging channels
are unified under a single undertaking, this broader market is highly fragmented
since phone (63%), email (18%), as well as webchat ([5-10%]), other OTT ([0-5%])
and SMS ([0-5%]) (each aggregated under “Other Messaging” in the data provided
by the Notifying Party) are in fact umbrella terms and each one is comprised of a
multitude of players of varying size, none of which the Commission expects to have
a material share or market power on an individual basis. Moreover, as shown in
Tables 3, 4 and 5 above, the shares of both phone and email have declined from 2018
to 2020 in the EEA and worldwide, while Meta (formerly Facebook)’s share has
increased over the same period, both in the EEA and worldwide, and is expected to
further increase in the future. This corroborates the market investigation results
evidencing a continuing decline in the use of more traditional B2C communication
channels such as phone and email towards an increased usage of more modern
channels, like the OTT B2C messaging channels offered by Meta (formerly
Facebook), even if there will remain certain instances where the use of phone and
email are more suitable.
424
In further support of the importance of Meta (formerly
421
Reply to questionnaire 3 to B2C communication service providers and online ads service providers,
questions 55.1 (ID: 741).
422
SO Response, paragraph 3.7.
423
See, e.g., replies to questionnaire 7 to CRM software providers, question 24.1.1 (ID: 1710) (“Most
communication channels are not wholly interchangeable. Each channel reaches slightly different
audiences and serves slightly different purposes. Customers expect providers of service CRM software
to provide a variety of digital messaging channels to reach their own customers.”); and replies to
questionnaire 7 to CRM software providers, question 24.1.1 (ID: 1655) (“We see the various
communication channels as complementary to one another, and not necessarily substituting each
other.”).
424
Replies to questionnaire 1 to CRM software providers, questions 47.1 and 47.2; replies to questionnaire
4 to business customers, questions 58.1 and 58.2.
83
Facebook)’s B2C channels, including within this broader market for B2C
communications overall, a major CRM software provider indicated that “it is
necessary to be where the customer is, so being present on all the channels and to
have an ‘omnichannel’ offering is the minimum entry requirement for the customer
service CRM market. For SAP, access to Facebook APIs is essential as it is for any
CRM software provider active in the customer service CRM market.”
425
(a.iii) Meta (formerly Facebook) has the ability to restrict or degrade access to its APIs
(273) As set out above (in particular in Section 5 on the Industry Overview), customer
service and support CRM software solutions will typically integrate a range of B2C
communications channels, including OTT B2C messaging channels (e.g.,
Kustomer’s CRM tool integrates 11 different communication channels). Such
integration is made possible through API access, which enables access to the
communication channel by multiple agents/employees in a business. OTT B2C
messaging channel providers (as well as other B2C communication channel
providers) may choose to offer free access to their channels, such as via an open,
public API access (which is currently the case for Messenger API).
426
They can also
opt to refuse access, to selectively grant access only to certain counterparties, and to
charge a fee for API access. As will be shown below, (A) Meta (formerly Facebook)
has the ability to restrict or block API access to its B2C messaging channels outright
or on a selective basis, (B) Meta (formerly Facebook) has the ability to degrade or
reduce the features or functionality provided by API access to its messaging
channels, including on a selective or targeted basis, (C) Meta (formerly Facebook)
would be able to identify a specific group of CRM providers to target, and (D)
foreclosed CRM providers would not be able to evade or workaround such targeting
by other means. However, for the reasons in part (E) below, Meta (formerly
Facebook) would likely not have the ability to obtain data on CRM competitors of
benefit to Kustomer.
A. Meta (formerly) Facebook has the ability to restrict access to its APIs outright or on
a selective basis
(274) While Meta (formerly Facebook)’s Messenger and Instagram APIs are currently
openly accessible, its WhatsApp for Business API (also referred to as “WABAPI”) is
not. For example, Meta (formerly Facebook)’s WhatsApp for Business API is
currently only provided to certain “channel partners”,
427
such as CPaaS providers
Twilio and MessageBird, which can then sell such access to businesses and CRM
software providers. There are however certain businesses (e.g., [Customer
Information] and [Customer Information]) and CRM software providers (e.g.,
[Customer Information]) to whom Meta (formerly Facebook) has opted to provide
direct API access to its WhatsApp for Business API.
428
(275) The fact that Meta (formerly Facebook) has opted to allow different degrees of
openness depending on the B2C messaging channel in question (and has even
adopted varying approaches to access for the same API, e.g., [Customer Information]
425
Agreed minutes of the call of 18 May 2021 with SAP, paragraph 15 (ID: 2008).
426
Meta (formerly Facebook)’s Messenger API is also available through “channel partners”. See Form CO,
paragraph 8.19.
427
Form CO, paragraph 6.105. This was confirmed by one such partner which said that “WABAPI is not
yet a publicly accessible API and its availability is limited through select providers like Twilio.”
(Agreed minutes of the call of 14 June 2021 with Twilio, paragraph 3 (ID: 2029)).
428
Form CO, paragraph 8.19.
84
and [Customer Information] were provided direct API access to WhatsApp for
Business, while many businesses need to go through an intermediary such as a
CPaaS provider) shows that Meta (formerly Facebook) has the ability — both
technically and contractually — to restrict API access to its OTT B2C messaging
channels should it so wish. This conclusion was corroborated and further confirmed
by the Commission’s market investigation, and in particular by the vast majority of
CRM software providers that expressed a view
429
, with one such provider stating
that, in its opinion, “Facebook could close off all of its open APIs at any time.”
430
(276) The Commission has also reached the conclusion that Meta (formerly Facebook) has
the ability to restrict API access to its messaging channels to specific rivals of
Kustomer (e.g., close competitors with a particular focus on SMB customers) and to
prevent new entrants in the customer service and support CRM software market from
obtaining such access (which may deter them from seeking to enter the market).
(277) As observed above, Meta (formerly Facebook) provides businesses and CRM
software providers with access to its messaging channels both directly, and through
channel partners, such as CPaaS providers like Twilio and MessageBird.
(278) Since Meta (formerly Facebook) is the one providing API access, its ability to grant
API access on a selective basis (and therefore to exclude access to others, such as
new entrants or close competitors of Kustomer) is clear. For example, Meta
(formerly Facebook) currently choses to make access to its Messenger and Instagram
APIs open, whereas it is more selective as regards to which businesses and partners it
grants access to its WhatsApp for Business API, which is currently only made
available to a relatively small number of “partners”.
(279) Where API access to Meta (formerly Facebook)’s messaging channels is obtained
not directly from Meta (formerly Facebook) itself but via a third party intermediary,
such as a CPaaS provider like Twilio or MessageBird (that may themselves have
obtained API access directly from Meta (formerly Facebook)), Meta (formerly
Facebook) would need to have the ability to control who CPaaS providers can offer
access to. Irrespective of whether Meta (formerly Facebook) has the technical ability
to do so, [Commercial Contratcs] that Meta (formerly Facebook) would, from a
contractual standpoint, have the means to prevent CPaaS providers from suppling
specific customer service and support CRM software providers, such as new entrants
or close competitors of Kustomer.
(280) In its agreements with CPaaS providers, a clause — which appears to be
[Information about Meta’s contractual rights].”
431
In view of the [Information about
Meta’s contractual rights], there would be nothing to prevent Meta (formerly
Facebook) from requiring its CPaaS providers to refuse API access to its messaging
channels to close competitors of Kustomer or new entrants in the customer service
and support CRM market.
432
In its SO Response,
433
[Information about Meta’s
contractual rights].”
434
However, the likelihood Meta (formerly Facebook) actually
429
Replies to questionnaire 7 to CRM software providers, question 50.
430
Reply to questionnaire 7 to CRM software providers, question 50.1 (ID: 1707).
431
See for example, clause 4.2 of the strategic partnership agreements that Meta (formerly Facebook)
entered into with [Commercial Contracts], provided as Meta (formerly Facebook)’s internal documents,
[Commercial Contracts].
432
[Confidential details on termination provisions of Meta contracts with CPaaS providers] See e.g., Meta
(formerly Facebook)’s internal documents, […] and […].
433
See Section 7.3.2.1.(a) above.
434
LoF Response, paragraph 2.5
85
engaging in a strategy that it has the technical ability to engage in depends primarily
on whether it would have an economic incentive to do so, and for the reasons set out
in Section 7.3.2.2.(b) below, the Commission concludes that Meta (formerly
Facebook) would have the incentive to engage in such a targeted foreclosure
strategy. The Commission further notes that Meta (formerly Facebook) [Information
about Meta’s contractual rights]
435
. Based on this agreement, Meta (formerly
Facebook) would have the contractual ability [Information about Meta’s contractual
rights].
436
(281) In view of the above, and irrespective of whether Meta (formerly Facebook) provides
API access to its B2C messaging channels (i.e., Messenger, Instagram and
WhatsApp for Business) directly or through third party intermediaries such as CPaaS
providers, the Commission concludes that Meta (formerly Facebook) has the ability
to restrict (i.e., to prevent) third parties from accessing the APIs to its B2C
messaging channels, including on a selective basis, e.g., by specifically targeting
close competitors of Kustomer or new entrants in the customer service and support
CRM market.
B. Meta (formerly Facebook) has the ability to degrade API access to its messaging
channels for all third parties or on a selective basis
(282) In addition to the ability to restrict API access, the Commission has also reached the
conclusion that Meta (formerly Facebook) would have the ability to degrade API
access to its B2C messaging channels. Meta (formerly Facebook) could do so on a
general basis, i.e., vis-à-vis all third parties that access its channels through APIs, or
on a selective basis by targeting certain businesses, certain competitors of Kustomer,
or certain regions.
(283) The ability of Meta (formerly Facebook) to degrade API access was confirmed by
the vast majority of respondents to the Commission’s market investigation.
437
In fact,
the Commission received confirmation that “Facebook has a wide variety of options
for how to degrade the quality of API access to its messaging channels. Facebook
can introduce rate limits to allow fewer transactions in a given period of time, can
restrict the features that are possible do through the API, and can limit types of data
that can be sent through the API.”
438
(284) Internal documents of Meta (formerly Facebook) [Business Plans Regarding Pricing
Strategy]
439
, [Business Plans Regarding Pricing Strategy]
440
(emphasis added).
(285) An EEA-based CRM software provider and competitor of Kustomer that took part in
the Commission’s market investigation suggested that Meta (formerly Facebook)
may even have previously degraded API access to its B2C messaging channels:
“This already happened to us in the past: Facebook removed some access to
Messenger API data essential to provide customer support […] - Profile pictures -
435
See Meta (formerly Facebook)’s internal document, […], clause 4.2.
436
See, e.g., Form CO, paragraph 6.67 Table 6.13. Separately, at Form CO, paragraph 9.16, the Notifying
Party indicated that following the Transaction [Meta strategic plans in relation to growth of Kustomer
post-Transaction].
437
Replies to questionnaire 1 to CRM software providers, question 51.
438
Reply to questionnaire 1 to CRM software providers, question 52.1 (ID: 762).
439
Parties’ reply to RFI 7, question 3.
440
Meta (formerly Facebook)’s internal document, Attachment […]. Meta (formerly Facebook), in its
reply to RFI 7, question 3, indicated that […] is a member of the “Deal Support Team” and was one of
[…] … who evaluated the Kustomer transaction from a business or strategic perspective, […].”
86
Audio messages - User timezone - User spoken language.”
441
Internal documents of
Meta (formerly Facebook) [Business Plans Regarding API Availability].”
442
[Business Plans Regarding API Availability]
443
, the Commission concludes that this
is nonetheless evidence of Meta (formerly Facebook)’s ability to degrade API access
to or the functionalities of its B2C messaging channels for third parties, even on the
basis of geographic location should it wish to do so.
(286) Other similar examples include Meta (formerly Facebook)’s removal of certain
specific APIs or features of the Meta (formerly Facebook) B2C messaging channels
(i.e., degrading access to such channel) for specific business users considered to be
non-compliance with Meta (formerly Facebook)’s applicable policies, e.g.,
[Information about User Removal from Messaging Channels]
444
(emphasis added).
Relatedly, in the case of the consumer version of WhatsApp, Facebook previously
threatened users that did not accept its updated policy that the “messaging app will
gradually degrade and eventually cease to function if [the users] don’t accept the
changes”
445
, which may point to an ability to adopt a similar approach vis-à-vis
businesses and CRM software providers that access the relevant channels via APIs.
Indeed, Meta (formerly Facebook) has faced and is facing antitrust scrutiny in the
United States, in which the charges refer to past practices of allegedly degrading API
access to players it perceived as competitors to its messaging channels.
446
The
Notifying Party, in the LoF Response, argues that these examples are not relevant to
Meta (formerly Facebook)’s ability to identify which CRM software provider to
target for a selective foreclosure strategy, since they have “no bearing on its ability
to establish whether specific businesses are using specific CRMs”.
447
This is not
however what the Commission is arguing. As explained by the Parties during
technical sessions in the week of 20 September 2021, for a business to use a B2C
messaging channel via a CRM software tool, it is necessary for the CRM software
provider to itself have an integration with the relevant B2C communication channel,
e.g., Meta (formerly Facebook)’s B2C messaging channels. Therefore, if Meta
441
Reply to questionnaire 4 to business customers, question 70.1 (ID: 527).
442
The Commission understands that a similar email may have been sent to a wide range partners
(including CRM software providers and CPaaS providers). See, by way of example, Meta (formerly
Facebook)’s internal documents, […].
443
See Meta (formerly Facebook)’s internal document, […].
444
Meta (formerly Facebook)’s internal document, […].
445
See https://www.wired.com/story/whatsapp-privacy-policy-facebook-data-sharing/, last accessed on
2 December 2021 (ID: 2517).
446
See Case No.: 1:20-cv-03590-JEB, Federal Trade Commission v Facebook, Inc,, First Amended
Complaint For Injunctive And Other Equitable Relief, filed 19 August 2021 with the U.S. District Court
for the District of Columbia, e.g., paragraph 154 (“In or around April 2013, Facebook terminated
Path’s access to key API functionality, and Path’s growth subsequently slowed significantly.”) and
paragraph 157 (“a. In January 2013, Facebook cut off key API access to Voxer, a mobile messaging
app featuring voice communication, shortly after Messenger launched competing voice functionality.
…b. In February 2013, messaging app MessageMe soared in popularity and achieved nearly one
million users within a week of release. But shortly after MessageMe reached one million users,
Facebook shut off key API access … In August 2013, Facebook undertook an enforcement strike
against a number of messaging apps simultaneously, with the Head of Developer Enforcement directing
colleagues to restrict them from “accessing any read APIs beyond basic info[,]” instructing that “we
will not be communicating with the [developers] in any way about these restrictions. … d. In October
2016, Facebook cut off certain API functionality to Tribe, a video-messaging app that was generating
buzz around that time” (emphasis added), available at:
https://www.ftc.gov/system/files/documents/cases/ecf_75-1_ftc_v_facebook_public_redacted_fac.pdf,
last accessed on 29 November 2021 (ID: 2515).
447
LoF Response, paragraph 2.6(B).
87
(formerly Facebook) were to target such CRM software providers, just as it engaged
in targeted foreclosure in the above examples, Meta (formerly Facebook) would not
necessarily need to identify ex ante the business users using or wishing to use that
CRM software solution. Even if such a business user had its own access to the
relevant channel already (which the Commission also understands from the technical
sessions to typically be the case), that business would not be able to have Meta
(formerly Facebook)’s B2C channels integrated into the foreclosed provider’s CRM
software tool, or would only be able to have them integrated in a degraded manner,
i.e., depending on the foreclosure strategy adopted by Meta (formerly Facebook).
(287) Meta (formerly Facebook) could alternatively, or in addition, upgrade the quality of
API access that Kustomer would have relative to third parties, or require data in
return for entering into API access partnerships.
448
. [Business Plans Regarding API
Availability]
449
(emphasis added). The Commission considers that this shows […],
Meta (formerly Facebook) considered that following its acquisition of Kustomer it
would have the ability to engage in a strategy of differentiated API access as between
Kustomer and rivals.
(288) There also appears to be evidence of Meta (formerly Facebook) having provided a
version of its APIs with enhanced features to certain partners and not others in the
past. Following technical sessions in the week of 20 September 2021, the Notifying
Party submitted information clarifying that its WhatsApp for Business API and
Messenger API are in fact each [Business Plans Regarding API Availability]
450
Despite statements that such features [Business Plans Regarding API Availability],
the Commission considers that Meta (formerly Facebook) has the ability to restrict
access to certain features to certain businesses. The fact that Meta (formerly
Facebook) [Details about Meta’s beta testing]
451452
appears to support this
conclusion.
(289) Although the Notifying Party points out that [Details about Meta’s beta testing].
453
454455
Contrary to what is argued in the SO Response
456
, the Commission does not
consider that genuine beta testing, which the Notifying Party states is a common
industry practice, would of itself raise competition concerns. Rather, given the
duration of such testing is uncertain, and at the discretion of Meta (formerly
Facebook) and not necessarily public knowledge
457
, the Commission considers that
third parties (e.g., competitors of Kustomer) are unlikely to have visibility of whether
a new API feature — even if initially tested with Kustomer — would be ready for
wider roll-out, and there would be nothing to prevent Meta (formerly Facebook)
from delaying (or deciding against) such a wider-roll out following successful testing
if it provided an advantage to Kustomer compared to rival CRM software providers.
448
Replies to questionnaire 4 to business customers, question 77.
449
Meta (formerly Facebook)’s internal document, […].
450
Parties’ reply to RFI 21, question 4. A list of the [Regulatory Submission] APIs that make up the
Messenger (40 APIs), Instagram (16 APIs) and WhatsApp ([Regulatory Submission] APIs) APIs
respectively were provided in the Parties’ reply to RFI 24, Annex 1.
451
Meta (formerly Facebook) has also sometimes rolled out new features to all API users at the same time,
without prior testing with a sub-set of partners. See Parties’ reply to RFI 21, question 5(c)(iii).
452
Parties’ reply to RFI 21, question 5(a).
453
Parties’ reply to RFI 21, question 4.
454
Parties’ reply to RFI 21, question 5(c)(i).
455
Parties’ reply to RFI 21, question 5(j).
456
See Section 7.3.2.1(a) above.
457
Replies to questionnaire 8 (Commitments Market Test), question 13.
88
(290) While the Notifying Party points out that [Details about Meta’s beta testing]
458459
Market participants that the Commission spoke to during its market investigation
referred to additional examples: “One specific example of this is the WhatsApp API
provided to the airline KLM. KLM through this API has access to unique WhatsApp
features (regarding boarding passes) that are not available through the API the rest
of the world has access to (e.g., via Facebook partners such as CPaaS providers). In
other words, Facebook has the ability to create tailored APIs with additional
features for certain partners. It is therefore entirely possible that Facebook may offer
Kustomer additional features not available to CRM competitors.”
460
Accordingly,
the Commission considers that Meta (formerly Facebook) would likely have the
ability to roll out new (tailored) features to its messaging channels to Kustomer, and
certain select partners and businesses (e.g., larger customer service and support CRM
software providers to whom Kustomer may not be a particularly close competitor),
while refusing to make the same features available to Kustomer’s close rivals or new
entrants into the customer service and support CRM software market.
(291) The Notifying Party argues, without providing supporting evidence, that such a
strategy “would not give any material “advantage” to Kustomer …[but even if it
did]…any such advantage would be truly de minimis”
461
. The Commission notes
however that Meta (formerly Facebook) is [Meta’s plans for new features to be
introduced in its messaging channels].
462
If some or all of these were restricted to
Kustomer the Commission considers that these additional functionalities could
provide a material advantage to Kustomer over competitors in the customer service
and support CRM software market. [Internal details regarding reasons for customer
switching]
463
[Internal details regarding reasons for customer switching]. Contrary to
the Notifying Party’s argument in its SO Response
464
, the above conclusions do not
require the Commission to demonstrate that each individual feature of a particular
B2C channel is, in itself, an important input. Rather, in view of the fact that OTT
B2C messaging channels are defined by reference to their richness of features, the
fact, or even the perception in the market, that rivals of Kustomer could only access a
less rich versions of Meta (formerly Facebook)’s channels or versions lacking certain
features, could over time result in a material advantage for Kustomer, in particular in
view of the fact that, first, and as outlined above, Meta (formerly Facebook)’s
channels are expected to further increase in importance for B2C communications in
the coming years and, second, as one business customer pointed out, the features that
may be important are typically determined by the end customer, rather than by the
CRM software provider or the business customer: “The most convenient feature will
be preferred by the customer and develop to an expected functionality over time. It
can be estimated that companies would consider switching to CS CRM provider that
is able to offer these functionalities or features to avoid unsatisfied customers.”
465
(292) In addition to differentiation on the basis of features, the Commission considers that
Meta (formerly Facebook) would have the ability to provide for a differentiated
degree of functionality or quality for all or a sub-set of third parties using its APIs as
458
Parties’ reply to RFI 21, question 5(f).
459
Meta (formerly Facebook)’s internal document, […].
460
Agreed minutes of the call of 9 September 2021 with Zendesk, paragraph 3 (ID: 2085).
461
Parties’ reply to RFI 21, question 5(j).
462
Parties’ reply to RFI 21, question 5(i).
463
Parties’ reply to RFI 21, question 16(6).
464
SO Response, paragraph 3.13.
465
Reply to questionnaire 8 (Commitments Market Test), question 10 (ID: 2622).
89
compared to the quality or level of functionality enjoyed by Kustomer. This point
was made by several respondents to the Commission’s market investigation, for
example, by differentiating, “the speed, data rate limits, and parity with natively
available features.”
466
[Details regarding Meta’s envisaged strategies in relation to
the Transaction] In an internal Meta (formerly Facebook) document from 2019
entitled [Details regarding Meta’s envisaged strategies in relation to the
Transaction]
467
(293) In view of the above, the Commission concludes that Meta (formerly Facebook) has
the ability to introduce important differences in the quality of API access to its
channels enjoyed by Kustomer compared to the access granted to third parties,
including on the basis of feature availability and quality or level of functionality
(e.g., rate limits). Meta (formerly Facebook) would be able to do this by degrading
the features and/or functionality provided to third party API access seekers, or by
improving the features and/or functionality made available to Kustomer, or through a
combination of such strategies. It could also engage in such a relative degradation
strategy in a targeted way, focusing on certain CRM software providers, e.g., close
competitors of Kustomer or new entrants for the reasons outlined in part A. above.
C. Meta (formerly Facebook) would be able to identify a specific group of CRM
providers to target
(294) As outlined above, the Notifying Party argues that Meta (formerly Facebook)’s
ability to engage in targeted foreclosure would be “limited” because there is “no
group of customers that is especially reliant on Facebook’s B2C messaging
channels” and in any case Meta (formerly Facebook) would “not be able to
accurately identify a specific group of businesses which would be more profitable to
target”.
468
Similar arguments were made in the SO Response, as summarised in
Section 7.3.2.1.(a) above, in particular that Meta (formerly Facebook) cannot identify
which CRM software providers are used by businesses that access its channels
directly or via third party intermediaries, such as CPaaS providers. Even if it could
identify those businesses, the Notifying Party considers that it would not be able to
target them as it would not have the ability to “withdraw or degrade access to its
B2C channels for a subset of businesses that use” a rival CRM platform.
469
(295) The Commission however considers that Meta (formerly Facebook) would have the
ability to identify and selectively target, for example, certain close competitors of
Kustomer (and even certain businesses using those CRM software providers) with a
sufficient degree of certainty.
(296) First, the targets of foreclosure are rival customer service and support CRM
providers, not business users, and Kustomer would be expected to be able to identify
its own close competitors. More specifically, through using Kustomer’s (albeit
historic) win/loss data, together with its general industry knowledge, and the large
variety of public sources and industry reports (e.g., Gartner) that analyse the
customer service and support CRM software market and often include profiles on
specific providers (e.g., whether they focus on e-commerce businesses, SMB clients,
466
Reply to questionnaire 1 to CRM software providers, question 52.1 (ID: 782).
467
Meta (formerly Facebook)’s, internal document].
468
Article 6(1)(c) Response, paragraph 3.19. See also Notifying Party Submission on Foreclosure Issues
dated 10 October 2021, footnote 1.
469
Article 6(1)(c) Response, paragraph 3.19.
90
etc.), the merged entity would be able to identify close competitors of Kustomer,
470
which could be included in any targeted foreclosure strategy. In addition, many CRM
software providers publish relatively extensive lists of their customer on their
websites, in many cases with the option to filter by industry and/or by region.
471
All
of this could also inform a targeted foreclosure strategy. Second, there would be
nothing to prevent Meta (formerly Facebook), which has a significant number of
businesses on its platforms (and using its B2C messaging channels) today, from
conducting a survey of the businesses that have a Facebook page and that use its
messaging channels in interacting with their customers (which Meta (formerly
Facebook) could verify) through which it asks that they identify their current
customer service and support CRM software provider allow Facebook to identify the
relevant. This could enable Meta (formerly Facebook) to determine if such
incumbent CRM software provider is a (close) competitor of Kustomer and if so,
whether to engage in a targeted foreclosure strategy toward that CRM software
provider. The Commission notes that the SO Response did not contradict this
observation. Indeed, Meta (formerly Facebook) has indicated that [Business Plans]
472
for which it would be expected to draw on data it holds or could readily obtain from
businesses that use it B2C messaging channels.
(297) In light of the above, the Commission concludes that contrary to what the Notifying
Party argues, Meta (formerly Facebook) would have the ability to identify, for
example, certain close competitors of Kustomer (and even certain businesses using
those CRM software providers) in order to carry out a targeted foreclosure strategy.
D. Foreclosed CRM providers would not be able to evade or workaround such targeting
by other means
(298) The Notifying Party also considers that there are “workarounds” available even if it
was able to target specific customer service and support CRM software providers
and/or businesses, who could “regain access via some other means”
473
“for example,
through a CPaaS provider or via integrations with a third-party CRM provider”.
474
(299) As outlined above, its agreements with partners such as [Information about Meta’s
contractual partners], give Meta (formerly Facebook) the contractual ability to
[Information about Meta’s contractual rights]
475
In view of the [Information about
Meta’s contractual rights], there would be nothing to prevent Meta (formerly
Facebook) from requiring its CPaaS providers to refuse API access to its messaging
channels to close competitors of Kustomer or new entrants in the customer service
and support CRM market, thereby preventing CPaaS providers being used as a
“workaround” as the Notifying Party argues. In the SO Response, the Notifying
Party argues that this would require the cooperation of such CPaaS providers.
476
However, the Commission considers that the ability of Meta (formerly Facebook) to
terminate any of it agreements with channel partners such as CPaaS providers,
470
See, e.g., Form CO, paragraph 6.67 Table 6.13.
471
See, e.g., HubSpot (https://www.hubspot.com/case-studies, last accessed on 15 December 2021 (ID:
2804)), Zendesk (https://www.zendesk.com/why-zendesk/customers/, last accessed on 15 December
2021 (ID: 2808)) and Freskdesk (https://freshdesk.com/customers, last accessed on 15 December 2021
(ID: 2803)).
472
Form CO, paragraph 9.16, indicates that following the Transaction [Meta strategic plans in relation to
growth of Kustomer post-Transaction
].
473
Article 6(1)(c) Response, paragraph 3.19.
474
Notifying Party Submission on Foreclosure Issues dated 10 October 2021, footnote 1.
475
See for example, […]. Meta (formerly Facebook)’s internal documents, […] and […].
476
SO Response, paragraph 3.18.
91
[Information about Meta’s contractual rights], would likely act as a strong incentive
for such partners to cooperate with any requests by Meta (formerly Facebook) to
cease providing access to specific CRM software providers.
477
(300) The Notifying Party also argues that an integration with a third-party CRM provider
could be used as a workaround in the case of targeted foreclosure.
478
In that regard,
the Commission notes however that, first, this would require that the third-party
CRM provider in question would not be foreclosed by Meta (formerly Facebook),
that the third party CRM provider in question offers an integration with Meta
(formerly Facebook)’s B2C communication services (the Commission understands
that this may not universally be the case) and that Meta (formerly Facebook) would
not have a similar contractual ability as it does with CPaaS providers to prohibit
them from granting access to certain customers (e.g., it has such a contract with
Zendesk). In addition to the fact that it is uncommon for businesses to multi-home as
regards CRM software with the same functionality (such as customer service and
support), the Notifying Party’s argument also ignores the fact that business
customers, in particular SMBs, which are the main focus of Kustomer’s offering, are
particularly price sensitive
479
and are unlikely to pay for two customer service and
support CRM tools at the same time. Indeed, as one market participant indicated, a
main obstacle to multi-homing is “subscription fees with different providers.”
480
(301) Contrary to what the Notifying Party argues
481
, it would not be possible for a
foreclosed CRM software provider to regain access to the particular Meta (formerly
Facebook) B2C messaging channel directly from Meta (formerly Facebook). Even if
their business customers have access to the relevant channels, Meta (formerly
Facebook) accepted that businesses — in particular SMBs that tend to favour off-the-
shelf CRM software solutions and which, unlike large enterprise businesses, would
typically not have the resources or the need to build a customised in-house solution
— usually require that “those B2C channels are then generally integrated in the
chosen CRM”,
482
which presumably would not be possible where Meta (formerly
Facebook) has blocked or degraded the API integration available to the relevant
CRM (e.g., using its “gating capability [to ensure that the API] …is not
inadvertently accessed by a non-participating developer”
483
, or through other
means). This means that even if the relevant business user can regain access to Meta
(formerly Facebook)’s channels via other means (e.g., direct access), it would not be
able to integrate it into the software of the CRM provider. SMBs would in many case
also not have the resources or inclination to build their own CRM software tool, or to
use a fragmented customer service set-up where only some B2C communication
channels are integrated into the CRM solution and other channels (e.g., those of Meta
(formerly Facebook)) are not, as this would be less efficient and more costly to
operate. The more plausible outcome is that business customers, and in particular
SMBs, i.e., the main focus of Kustomer’s offering, which are particularly price
477
[Confidential details on termination provisions of Meta contracts with CpaaS providers] See e.g., Meta
(formerly Facebook)’s internal documents, Attachments […] and […].
478
See Section 7.3.2.1(a) above.
479
The vast majority of business customers indicated that price is an important or very important factor in
a business’ choice of customer service and support CRM provider. See replies to questionnaire 5 to
business customers, question 15.
480
Reply to questionnaire 4 to business customers, question 11.2 (ID: 842).
481
See Section 7.3.2.1.(a) above.
482
Meta (formerly Facebook)’s Comments on State of Play Meeting, 22 July 2021, paragraph 5.1(B).
483
Parties’ reply to RFI 21, question 5(a).
92
sensitive
484
, would switch to using a customer service CRM software tool in which
all B2C communication channels, including in particular Meta (formerly
Facebook)’s OTT B2C messaging channels, are integrated, namely Kustomer.
(302) In light of the above, the Commission concludes that CRM software providers (and
their business customers) would not be able to evade or workaround a targeted
foreclosure strategy.
E. Meta (formerly Facebook) is unlikely to have the ability to obtain data on CRM
competitors to provide a competitive advantage to Kustomer
(303) For completeness, the Commission investigated whether Meta (formerly Facebook)
may be able to obtain data on competitors of Kustomer, e.g., directly or via partners
such as CPaaS providers, which it could use to provide a material advantage to
Kustomer.
(304) Meta (formerly Facebook) confirmed that while it [Confidential information on
Meta’s ability to obtain data].
485
Meta (formerly Facebook) further pointed out that it
“[Confidential information on Meta’s ability to obtain data].”
486
(305) Where Meta (formerly Facebook)’s messaging channels are accessed via third
parties, [Commercial Contracts].”
487
(306) While some of this information could theoretically be competitively sensitive and/or
capable of providing a strategic advantage to Kustomer, such a scenario appears to
be remote. Rather, based on the explanations provided by Meta (formerly Facebook),
it appears that there may be an objective reason for Meta (formerly Facebook)
requiring that such information be provided, such as ensuring integrity of the service,
confirming the identity and authenticity of the API access seeker or user, to allow for
the provision of the service (e.g., for a business’ name and identifying information to
be displayed in the relevant messaging channel as part of a conversation with an end
customer, that information would need to be communicated to Meta (formerly
Facebook).
(307) In light of the above, the Commission concludes that Meta (formerly Facebook) is
unlikely to have the ability to obtain data on CRM competitors to provide a
competitive advantage to Kustomer.
(a.iv) Conclusion
(308) On the basis of the assessment in Section 7.3.2.2.(a) above, the Commission
concludes that Meta (formerly Facebook) would have the ability to foreclose access
by customer service and support CRM software providers, both in the EEA and
worldwide, to a large and important part of the EEA-wide, if not worldwide, OTT
B2C messaging market. This conclusion also holds for the broader EEA-wide, if not
worldwide, upstream market for B2C communication services overall, given the
importance of Meta (formerly Facebook)’s channels for CRM software providers and
their business customers, as well as the complementarity of different channels in
484
The vast majority of business customers indicated that price is an important or very important factor in
a business’ choice of customer service and support CRM provider. See replies to questionnaire 5 to
business customers, question 15.
485
Parties’ reply to RFI 24, question 3(d).
486
Parties’ reply to RFI 24, question 3(e).
487
[Commercial Contracts].
93
such a wider market
488
, as well as the broader downstream EEA-wide, if not
worldwide, market for CRM software overall, given the importance of customer
service and support CRM software within an overall CRM software market.
489
As
outlined in Section 7.3.2.2.(a.i) above, the Commission’s conclusion would not
change on the basis of the potential segments of the customer service and support
CRM software market, or the overall CRM software market, based on (i) business
customer size, (ii) mode of deployment or (iii) the industry sector of the business
customer because OTT B2C messaging channels would remain an equally important
input irrespective of such further segmentation. This conclusion remains the same
irrespective of whether the relevant upstream and downstream markets, and potential
segments thereof, are defined as EEA-wide or worldwide in geographic scope.
(b) Incentive to foreclose
(309) The Commission has reached the conclusion that the merged entity would have the
incentive to engage in a foreclosure strategy with regard to competing providers of
customer service and support CRM software. A foreclosure strategy would, on the
one hand, increase the number of businesses using Kustomer (resulting in gains to
the merged entity), and, on the other hand, decrease the use of Meta (formerly
Facebook)’s messaging channels by those businesses that do not switch to Kustomer
(resulting in losses to the merged entity). Sub-section (b.i) concludes that the gains to
the merged entity from a foreclosure strategy appear numerous, diverse and
significant. This includes gains from SaaS revenue, additional data for online ads
purposes, and CTM ads, as well as longer-term benefits from steering businesses into
the Meta (formerly Facebook) ecosystem of products. Sub-section (b.ii) concludes
that the losses to the merged entity from a foreclosure strategy can be limited to a
sufficient degree by Meta (formerly Facebook) by targeting Kustomer’s close
competitors (sub-section A) and through further aggravating factors (sub-section B),
thereby sufficiently minimizing the number of business customers that do not switch
to Kustomer and the associated losses. In light of the significantly larger gains
relative to losses and the available evidence on business customer switching post-
foreclosure, the Commission concludes that the level of switching to Kustomer
would be sufficient so that the merged entity would have the incentive to engage in a
foreclosure strategy.
(310) The Commission considers that in the context of this case a purely quantitative
assessment of the exact gains from a potential input foreclosure strategy is not
possible. Whilst the Commission undertakes a quantitative assessment of the
quantifiable gains and losses, a qualitative approach is required to analyse the
unquantifiable gains. The Commission does not consider there to be any
unquantifiable losses, and so the quantitative assessment, which does not take into
account certain unquantifiable gains, is therefore conservative. Indeed, during the
488
For example, one respondent to the market investigation indicated that “We see the various
communication channels as complementary to one another, and not necessarily substituting each other”
(reply to questionnaire 1 to CRM software providers, question 31.1 (ID: 759). For the same reasons as
for the broader market for B2C communication services overall, the conclusion equally holds on the
basis of the potential intermediate product market for asynchronous B2C communication services. Any
conclusion in this Decision in relation to the broader market for B2C communication services overall
should be read as also referring to the potential intermediate market for asynchronous B2C
communication services.
489
In 2020, customer service and support CRM software accounted for a large portion (c. 36%) of the
overall CRM market on both a worldwide and EEA basis. See Form CO, paragraph 6.45 and
paragraph 6.47 Tables 6.2 to 6.5.
94
deal process, the Notifying Party itself [Information about deal valuation].
490
Further,
during the investigation, the Notifying Party has still not been able to quantify all the
benefits and costs associated with such a foreclosure strategy. This includes, for
example, the reputational costs that the Notifying Party argues would significantly
reduce its incentive to undertake such a strategy.
491
(311) Firstly, as will be demonstrated in the remainder of this Section, certain gains from
foreclosure are inherently forward-looking and uncertain, since they are likely to
only become salient after a number of years. It is therefore difficult to meaningfully
quantify them. In addition, any indirect positive spill overs to Meta (formerly
Facebook)’s other businesses are similarly difficult to meaningfully quantify.
492
(312) Secondly, it is also unclear how many business customers would, post-Transaction
and as a result of foreclosure, switch to Kustomer downstream and away from Meta
(formerly Facebook) upstream. The Notifying Parties’ internal estimates of the
potential number of business customers that could switch to Kustomer [Information
about parties’ switching estimates].
493
Further, it remains difficult to quantify the
number of business customers that would switch to Kustomer as a result of
foreclosure, because a foreclosure strategy may occur in parallel with other Meta
(formerly Facebook) strategies to encourage increased use of Kustomer, for example,
offering Kustomer for free or below market price. The Notifying Party concludes that
business customers would not switch sufficiently to Kustomer based on historical
win/loss data.
494
The Commission considers that these data are not informative in
this respect because there has not been an example in the past where the owner of a
leading B2C channel owns a CRM (i.e., the pre-Transaction scenario is significantly
different to the post-Transaction scenario). As such, in all cases where a business
customer decided to switch for reasons other than to get a better integration to OTT
B2C messaging channels (e.g., AI functionalities, third-party integrations, price, etc.)
historical switching rates would not be indicative of switching following a
foreclosure strategy, i.e., in response to a deterioration of Meta (formerly Facebook)
access for some of Kustomer’s competitors but not for Kustomer, and can at best
provide a lower bound for the switching rates to Kustomer in response to such a
deterioration. Indeed, accounting for some of these shortcomings overturns the
Parties’ conclusion that foreclosure could not be profitable.
(313) Overcoming the degree of uncertainty inherent with certain benefits of foreclosure
would require the use of strong assumptions. As such, any conclusions drawn from a
pure quantitative analysis conducted on this basis would likely lack reliability and
meaning.
(314) In this regard, the Commission considers that the vertical arithmetic analysis
submitted by the Notifying Party does not provide a reliable and sufficiently
complete assessment of the incentive for the merged entity to engage in a foreclosure
strategy.
495
As part of the economic submission, Meta (formerly Facebook) provided
a model, based on a standard vertical arithmetic framework, which calculates the
critical number of businesses that would need to switch to Kustomer for a total
490
See, for example, Meta (formerly Facebook)’s internal document, Attachment E0548 – […].
491
Article 6(1)(c) Response, paragraph 3.29.iii.
492
The Commission’s quantitative analysis therefore (conservatively) excludes these unquantifiable gains.
493
See, for example, Form CO, Annex 13.17, Acquisition by Facebook, Inc. of Kustomer, Inc. Analysis of
Vertical Effects, footnote 64; Meta (formerly Facebook)’s internal document, Attachment E0551 – […].
494
SO Response, Annex 3, Figure 1.
495
Form CO, Annex 13.17, Acquisition by Facebook, Inc. of Kustomer, Inc. Analysis of Vertical Effects.
95
foreclosure strategy to be profitable. However, whilst the Parties’ analysis mirrors in
large parts the Commission’s assessment closely (with both relying on the same
sources and data), the Commission’s assessment below demonstrates that firstly the
assumptions underlying the Parties’ model are not consistent with the available
evidence, most importantly with respect to the level of switching that will occur
following a targeted foreclosure strategy, and that secondly the model overstates the
amount of losses Meta (formerly Facebook) would incur following foreclosure
(specifically with respect to CTM ad revenues). By addressing some of these
shortcomings in the Parties’ quantitative assessment, the Parties conclusion that a
foreclosure strategy would be unprofitable for the merged entity is overturned.
Moreover, the Parties’ analysis does not assess (qualitatively or otherwise) any of the
unquantifiable gains identified by the Commission.
(315) In light of the above, the Commission considers the most appropriate approach to
assess the Parties’ incentives to engage in input foreclosure consists of a qualitative
assessment of both the quantifiable and unquantifiable gains and losses, as well as a
quantitative assessment of only the quantifiable gains and losses, which is itself
conservative since only certain gains are unquantifiable.
(316) As outlined in Section 7.3.2.2.(a), Meta (formerly Facebook) has multiple avenues
through which it could restrict or degrade CRM competitors’ API access to its
channels. This includes the ability to totally or partially foreclose rivals, as well as to
target specific CRM software providers, potential entrants and even specific business
customers. Whilst these various potential foreclosure strategies would change the
magnitude of gains and losses from foreclosure, they would not significantly impact
the question of whether the merged entity has the incentive to foreclose, i.e., whether
the gains outweigh the losses, because the benefits would remain significant relative
to the limited losses across the different foreclosure strategies. For example, and as
demonstrated below, a total foreclosure strategy compared to a partial foreclosure
strategy would not significantly change the gains and losses per business customer,
or the ratio with which business customers would switch from foreclosed CRM
software providers to Kustomer or non-foreclosed CRM software providers. As such,
the relative gains and losses would remain largely the same between these two
foreclosure strategies, i.e., if total foreclosure is profitable (with the gains
outweighing the losses), then partial foreclosure would also be profitable, and vice
versa. In other words, whilst total foreclosure would encourage more switching away
from foreclosed CRM software providers (and therefore provide more gains in
absolute terms to Meta (formerly Facebook)), whether the gains outweigh the losses
depends on the gains and losses per business customer and the ratio of switching to
Kustomer versus non-foreclosed CRM software providers, which would be similar
between total and partial foreclosure strategies; and hence the direction of
profitability would be the same between these different foreclosure strategies. As
such, the analysis of benefits and costs of foreclosure proceeds with each benefit and
cost in turn, and noting in each instance whether the different foreclosure strategies
available to the merged entity would significantly impact the analysis of the benefit
or cost. On the basis of the assessment, and as noted at the beginning of this Section,
the Commission considers that the benefits from foreclosure are significant, and that
losses would be sufficiently limited, because Meta (formerly Facebook) is able to
sufficiently steer business customers towards Kustomer. This conclusion holds for
both total and partial foreclosure strategies, as well as irrespective of the specific
CRM providers or business customers targeted by the foreclosure.
96
(b.i) Benefits from foreclosure are numerous, diverse and significant
(317) The Commission considers that the benefits to the merged entity from an input
foreclosure strategy, i.e., from steering business customers towards Kustomer, are
numerous, diverse and significant. Meta (formerly Facebook) is a large business both
in terms of its market share (e.g., in the online display advertising market or OTT
B2C messaging market) and in absolute (revenue) terms.
496
Accordingly, the benefits
the merged entity is able to realise from an input foreclosure strategy are significant,
precisely because of the scale and network effects it can build upon. Meta (formerly
Facebook)’s products are also numerous and diverse as well as large in scale, and
include, for example, its online advertising business and ecommerce products. Meta
(formerly Facebook)’s many different closely related products provide it with many
avenues from which to benefit from foreclosure, and in particularly from business
customers switching to Kustomer. This is the case in the short-term (with the
quantifiable gains), but the Commission considers in light of the many longer-term
gains (which are themselves difficult to quantify), that a longer timing horizon must
also in parallel be taken into account (e.g., within the next 5 years). Sub-sections A-
D proceed by discussing in detail each of the benefits from foreclosure in turn,
demonstrating that each of the numerous and diverse benefits to the merged entity
from steering a business customer to Kustomer is significant (per business customer).
This applies equally to steering a business customer to Kustomer by foreclosing
current CRM software competitors and potential entrants into the market. Sub-
section (b.ii) subsequently demonstrates that the number of businesses switching to
Kustomer will be sufficiently large such that the overall benefit to the merged entity
from foreclosure will be significant and thereby also the losses limited.
A. SaaS revenues
(318) A successful foreclosure strategy would mean that the merged entity gains market
share downstream, as rivals’ CRM offerings become less attractive compared to
Kustomer’s offering. The merged entity would gain from businesses that switch
away from their current CRM provider to Kustomer as a result of foreclosure. The
most direct of these benefits comes from SaaS revenues generated when a business
customers purchases Kustomer’s CRM software.
(319) [Information about Transaction Rationale].
497
A review of the Notifying Parties’
internal documents indicates that [Information about Transaction Rationale].
498
(320) However, the Commission notes that sales from CRM software could nevertheless
constitute a significant and growing stream of revenue for the merged entity.
Kustomer’s internal data for 2020 indicates that Kustomer would earn on average
USD [Revenue Information] profit a year from each business customer that switches
to Kustomer.
499
Meta (formerly Facebook) has estimated that Kustomer may be a
feasible CRM software option for [Customer Information] businesses.
500
This can be
496
See, for example, Commission decision of 17 December 2020 in case M.9960 – Google/Fitbit,
paragraph 463.
497
See, for example, Meta (formerly Facebook)’s internal document, Attachment E20130 - […]; Parties’
reply to RFI 9, question 26.
498
Meta (formerly Facebook)’s internal documents discussing the deal rationale [Confidential information
on Meta’s valuation of Kustomer]. See Form CO, paragraphs 3.2.11-3.2.12.
499
Parties’ reply to RFI 9, question 26. In addition, this assumes a [Revenue Information] % gross margin
for Kustomer’s CRM software services, which is taken from Meta (formerly Facebook)’s internal
document, […].
500
Meta (formerly Facebook)’s internal document, Attachment E20130 – […].
97
contrasted to Meta (formerly Facebook)’s projection of growing Kustomer’s
customer portfolio from [Customer Information] businesses today to [Customer
Information] businesses in [years].
501
If all the customers for which Kustomer may
be a feasible solution were to switch to Kustomer as a result of the foreclosure
strategy, this would imply a maximum possible gain of over USD [Revenue
Information] year from CRM SaaS alone as a result of such a foreclosure strategy.
While the actual gain depends on the number of business customers who would
switch to Kustomer as result of the foreclosure strategy, it shows at least that there is
a large pool of businesses out of which Meta (formerly Facebook) would potentially
be able to gain new customers by engaging in a foreclosure strategy. And, as outlined
in Section 7.3.2.2.(b.ii) below, it seems likely that a significant proportion of the
relevant market would switch to Kustomer following a foreclosure strategy. The
Commission notes that Kustomer’s [Business Plans], which would imply a higher
maximum possible gain from foreclosure.
502
(321) In addition, the Commission notes that revenues from SaaS are likely to grow
exponentially. A model from Meta (formerly Facebook), used to estimate the SaaS
revenue to be gained as a result of the Transaction, assumes that revenue from
business customers increases by [Revenue Information] every 12 months as a result
of “upselling”, i.e., from either purchasing a higher quantity of Kustomer’s products
or a more expensive version of them.
503
For illustration, over five years, an annual
increase of [Revenue Information] leads to a more than [Revenue Information]
increase in revenues. Indeed, this model predicts that by [Revenue Information]. As
such, one expects that this benefit from foreclosure will become more and more
significant in the future. Indeed, the slides from the technical session dated 22
September 2021 note that “[Business Plans]”.
504
One would expect any SaaS
revenues the merged entity gains through foreclosure to grow similarly.
(322) Lastly, as discussed in Section 7.3.2.2.(b.ii) below, the merged entity might combine
a foreclosure strategy with parallel aggravating factors designed to encourage
business customers to switch to Kustomer, which – even if legitimate on their own –
would aggravate the impact of the foreclosure. For example, Meta (formerly
Facebook) may (in line with historical behaviour) offer Kustomer for free or below
market price. In such a scenario, the merged entity would therefore forgo the SaaS
revenues from Kustomer. However, as will be shown, the benefits from foreclosure
to Meta (formerly Facebook)’s current business activities likely far outweigh these
SaaS revenues.
501
Meta (formerly Facebook)’s internal document, Attachment E029 – […].
502
Form CO, Attachment E0551. This document prepared by Kustomer includes strategies for Kustomer to
win customers following its possible acquisition by Meta (formerly Facebook), and estimates Kustomer
could potentially gain between [40,000-300,000] Torso customers post-Transaction, which is more than
the critical number of businesses estimated by any version of the model submitted by the Parties. In the
Parties’ reply to RFI 10, Meta (formerly Facebook) stated that it does not consider Kustomer’s growth
projections as set out in Attachment E0551 to be a realistic or reliable benchmark against which to
assess the extent to which it would be able to profitably acquire new customers post-Transaction (with
or without foreclosure).
503
Meta (formerly Facebook)’s internal document, Attachment E20130 – […].
504
PowerPoint presentation session 2 “Facebook’s proposed acquisition of Kustomer - Kustomer and the
CRM Market” of the site visit held on 22 September 2021, slide 11.
98
B. Additional data for online ads purposes
(323) By engaging in a foreclosure strategy and steering businesses away from their
current CRM provider to Kustomer, Meta (formerly Facebook) would also gain
additional data from these business customers. [Business Plans].
505
(324) Meta (formerly Facebook) [Business Plans in relation to Data Sharing].
506
(325) Meta (formerly Facebook) currently offers businesses a number of tools to share data
with Meta (formerly Facebook) for ads purposes. Meta (formerly Facebook)
Business Tools include Facebook Pixel, Conversions API, App Events via Meta
(formerly Facebook) software development kits (“SDKs”), Offline Conversions API,
and App Events API. Data shared with Meta (formerly Facebook) via Meta (formerly
Facebook) Business Tools generally originates outside Meta (formerly Facebook)’s
channels or surfaces. Depending on the relevant solution and the features used,
business customers may choose to share different types of information with Meta
(formerly Facebook).
507
(326) In its valuation model of Kustomer, Meta (formerly Facebook) [Business Plans].
508
[Business Plans].
509
(327) In this regard, Meta (formerly Facebook) has estimated that in [Year] it would earn
on average USD [Revenue information] profit a year from each business customer
that switches to Kustomer.
510511
Meta (formerly Facebook) has estimated that
Kustomer may be a feasible CRM software option for [Customer information]
businesses.
512
This can be contrasted to Meta (formerly Facebook)’s projection of
growing Kustomer’s customer portfolio from [Customer information] businesses
today to [Customer information].
513
If all the customers for which Kustomer may be
a feasible solution were to switch to Kustomer as a result of the foreclosure strategy,
this would imply a maximum possible gain of over USD [Revenue Information] a
year from additional data alone as a result of such a foreclosure strategy. As noted
above, while the actual gain depends on the number of business customers who
would switch to Kustomer as result of the foreclosure strategy, it shows at least that
there is a large pool of businesses out of which Meta (formerly Facebook) would
potentially be able to gain new customers by engaging in a foreclosure strategy. And,
as outlined in Section 7.3.2.2.(b.ii) below, it seems likely that a significant proportion
of the relevant market would switch to Kustomer following a foreclosure strategy.
The Commission notes that Kustomer’s own estimates for the number of potential
business customers it could serve post-Transaction are far higher than those
505
See, for example, Meta (formerly Facebook)’s internal document, Attachment E029 – […]; Parties’
reply to RFI 1, question 33.
506
Parties’ reply to RFI 1, questions 8. 34 and 35; Parties’ reply to RFI 6, question 9.
507
Parties’ reply to RFI 16, question 12.
508
Parties’ reply to RFI 10, question 20.
509
Parties’ reply to RFI 10, question 20.
510
Parties’ reply to RFI 9, question 26. [Future revenue calculation methodology and assumptions]. See
Parties’ reply to RFI 1, question 33.
511
It is worth noting that the model assumes that [Business Plans in relation to Data], but, as demonstrated
later in this Section, this data is highly valuable for online advertising and therefore may have a higher
incremental value than Facebook assumes for the purposes of its deal valuation.
512
Meta (formerly Facebook)’s internal document, Attachment E20130 – […].
513
Meta (formerly Facebook)’s internal document, Attachment E029 – […].
99
estimated by Meta (formerly Facebook), which would imply a higher maximum
possible gain from foreclosure.
514
(328) Meta (formerly Facebook) included in its valuation model of Kustomer the benefits
from [Business Plans about Data Sharing].
515
[Business Plans about Data Sharing]. A
market participant notes that “online advertising service providers cannot easily
access this kind of data through other means. There is some sharing between
vendors, but not a lot, and it is becoming more difficult to share data because of
regulation and legislation around the world, and because of the privacy policies of
firms like Apple or Google (Android) which limit the ability of apps on a phone to
exchange third-party data between them” and “in the current, privacy-focused,
environment, the Transaction is a way for Facebook to get direct access to this
data”.
516
In light of the decreasing availability of third-party data, the value of
gaining additional data through Kustomer is significant and growing. These benefits
from additional data would similarly come about as a result of foreclosure, and to a
greater extent if a foreclosure strategy encourages a greater level of switching to
Kustomer (than absent any foreclosure strategy).
(329) The next part of this Section discusses in detail, first the data Meta (formerly
Facebook) could potentially get access to through Kustomer, and second the
relevance of that data to Meta (formerly Facebook)’s online advertising business. In
light of this information, the Commission considers that the data Meta (formerly
Facebook) could obtain following the acquisition of Kustomer is highly relevant and
valuable for improving Meta (formerly Facebook)’s targeting capabilities for the
online display advertising services it provides. Additional data would therefore
constitute a significant benefit from foreclosure.
(330) Meta (formerly Facebook) qualifies the types of data that it would potentially get
access to through the acquisition of Kustomer as Customer Transaction Data and
Other Event data, pursuant to the terms for Meta (formerly Facebook) Business
Tools.
(331) In this context, Customer Transaction Data (or conversions data) relates to data on
completed transactions that businesses may share with Kustomer. More specifically,
this data could include, for example, personal data such as customer contact
information, their gender or date of birth, together with order and purchase history,
payment information, return data and customer interaction data.
(332) Customer interaction data, meaning conversations between a particular customer and
a customer service agent of a business, are uniquely presented in Kustomer’s CRM
software. A conversation may take place over several B2C communication channels
and Kustomer’s CRM software shows this conversation in a single timeline:
514
Form CO, Attachment E0551. This document prepared by Kustomer includes strategies for Kustomer to
win customers following its possible acquisition by Meta (formerly Facebook), and estimates Kustomer
could potentially gain between [Business Plans] Torso customers post-Transaction, which is more than
the critical number of businesses estimated by any version of the model submitted by the Parties. In the
Parties’ reply to RFI 10, Meta (formerly Facebook) stated that it does not consider Kustomer’s growth
projections as set out in Form CO, Attachment E0551 to be a realistic or reliable benchmark against
which to assess the extent to which it would be able to profitably acquire new customers post-
Transaction (with or without foreclosure).
515
Attachment E0548 – […].
516
Agreed minutes of the call of 3 June 2021 with a market participant, paragraph 20 and 21 (ID: 2104).
102
business’ back-end system, and through widgets that are included in for example e-
commerce platforms.
519
(336) Finally, Kustomer’s CRM software may also separately generate some of its own
data. For example, Kustomer’s CRM software calculates a “customer sentiment”
score. This score allows a business to evaluate how a customer’s current sentiment
compares to that customer’s average sentiment, or what the impact is of a customer
service agent interaction on the customer’s sentiment:
Source: Kustomer Timeline, Product Overview video.
520
(337) The data that is available to a business on Kustomer’s CRM software may be used
for specific searches and can be used for a specific advertisement campaign. One
Kustomer live demo shows for example how a search can be done among VIP
customers of a business that did not make a purchase in the last 30 days, that spent
more than a certain amount with the business in its lifetime, and that had a negative
sentiment during the last customer service agent interaction.
521
(338) The Commission assessed the volume, value, variety and velocity of the data that
Meta (formerly Facebook) could obtain on the basis of its review of internal
documents as well as on the results of the market investigation.
(339) First, as regards the volume of data, it must be noted that, at the time of notification,
Kustomer had around [Customer Information] customers worldwide. According to
Meta (formerly Facebook)’s growth plans, this would increase to approximately
519
Kustomer live demo dated 24 April 2019, available at: https://kustomer.wistia.com/medias/aj05oeenqj,
last accessed on 21 July 2021 (ID: 2801).
520
Kustomer Timeline, Product Overview video, available at: https://www.kustomer.com/demo-videos/,
last accessed on 21 July 2021.
521
Kustomer live demo dated 24 April 2019, available at: https://kustomer.wistia.com/medias/aj05oeenqj,
last accessed on 21 July 2021.
103
[Business Plans] customers by 2025.
522
However, no information is available on the
total number of customers with whom business customers of Kustomer have
recorded interactions. For each customer of a business customer of Kustomer, there
are many data points that Meta (formerly Facebook) could obtain as set out above,
including both Customer Transaction Data and Other Event Data (as described
above).
(340) The Notifying Party submits that through the acquisition of Kustomer, Meta
(formerly Facebook) would not get access to any new types of data that it could not
access before the Transaction.
523
In particular, the Notifying Party states with regard
to the custom object data that businesses may decide to upload into their Kustomer
CRM software, that businesses are already able to share such data with Meta
(formerly Facebook) for ads purposes through the various Meta (formerly Facebook)
Business Tools available to them (e.g., SDKs, Facebook Pixel, Conversions API,
Offline Conversions API, App Events API, etc.).
524
(341) The Commission considers that, while Meta (formerly Facebook) might already have
access to similar types of data, it may not currently have a full overview of customer
transactions (both before, during and after) on and off Meta (formerly Facebook)’s
platforms. This is confirmed by one respondent to the market investigation, stating
that getting access to data through the acquisition of Kustomer “will help to close the
loop on the customer funnel from Awareness – Conversion – Repurchase”.
525
(342) Moreover, a presentation from the CEO of Kustomer states the following regarding
the fact that joined data creates new ads opportunities: “[Business Plans]?”
526
The
same presentation contains the following overview of data that might be available to
respectively Meta (formerly Facebook) and Kustomer:
[Business Plans]
Source: Form CO, Attachment E.
527
(343) [Regulatory Submission]. Moreover, any customer interaction that takes place
between a customer and a business on any B2C communication channel other than
those owned by Meta (formerly Facebook) captured by Kustomer (during the
Inquiry, Delivery and Support stages mentioned above) would also become available
to Meta (formerly Facebook).
(344) [Business Plans about Data Sharing], the Notifying Party submitted that it currently
already has access to some of this data. For example, Meta (formerly Facebook) now
also has a Shopify integration and businesses using Shopify can choose to share their
Shopify data relating to “[Business Plans about Data Sharing]” with Meta (formerly
Facebook).
528
Moreover, Meta (formerly Facebook) submitted that data relating to
“[Business Plans about Data Sharing]” can be shared by businesses as Other Event
Data through Meta (formerly Facebook)’s existing Meta (formerly Facebook)
Business Tools, such as Facebook Pixel.
529
522
Form CO, paragraph 3.2.10.
523
Form CO, paragraph 6.122.
524
Form CO, footnote 227.
525
Reply to questionnaire 4 to business customers, question 66.1 (ID: 852).
526
Form CO, Attachment E, […].
527
Form CO, Attachment E, […].
528
Reply to RFI 17, question 8(d).
529
Reply to RFI 21, Table 12.1.
104
(345) The Commission considers however that, while Meta (formerly Facebook) can
receive in theory most of these types of the data mentioned through its Meta
(formerly Facebook) Business Tools as Other Event Data, this does not mean that
Meta (formerly Facebook) actually receives this data. In that regard, the Notifying
Party [Regulatory Submission]
530
Moreover, the Commission cannot rule out that the
content of messages (i.e., unstructured customer interaction data) from B2C
communication channels other than those of Meta (formerly Facebook), which is not
regularly shared with Facebook today, could be a potentially rich source of
information.
(346) The Notifying Party further submits that any accretive data signal that could be
available to Meta (formerly Facebook) via Kustomer would be reliant on the
agreement of the businesses making use of Kustomer.
531
This could have an impact
on the volume of data that would become available to Meta (formerly Facebook)
following the Transaction. More specifically, the Notifying Party submits that end-
customer’s consent is first and foremost governed by the terms of the relevant
business with which that end-customer is transacting.
532
As regards data that is stored
within Kustomer on business customer’s behalf, the Notifying Party submits that the
relevant consents are governed by the contracts and policies concerning Kustomer’s
use of that data.
533
(347) Internal documents show [Information from internal documents about ads data]
534
[Information from internal documents about ads data]. In this regard, the Notifying
Party submits that Meta (formerly Facebook) [Information from internal documents
about ads data]
535
(348) The results of the market investigation confirm that respondents do not consider that
the need to obtain such an agreement would hinder Meta (formerly Facebook) in its
ability to obtain a significant volume of data through its acquisition of Kustomer. All
B2C communication channel providers and online ads service providers expressing
an opinion indicate that Meta (formerly Facebook) could encourage businesses using
Kustomer to share more data through its ownership of Kustomer.
536
Moreover, a
majority of customers of Meta (formerly Facebook) and Kustomer that responded to
the market investigation indicate that, following the Transaction, they would be
willing as a customer of CRM software to share data with Meta (formerly Facebook)
for online display advertising purposes.
537
(349) Similarly, all B2C communication channel providers and online ads service
providers expressing an opinion consider that access to data through Kustomer would
provide Meta (formerly Facebook) with an increment of data that materially
improves its ability to personalize its online display advertising services for end
530
Reply to RFI 21, Table 12.1.
531
Form CO, paragraph 6.122.
532
Form CO, footnote 231.
533
Meta (formerly Facebook)’s comments on the State of Play Meeting, 22 July 2021, paragraph 4.2.
534
Form CO, Attachment E2344.
535
Form CO, paragraph 6.122.
536
Replies to questionnaire 3 to B2C communication service providers and online ads service providers,
question 70.
537
Replies to questionnaire 4 to business customers, question 64.
105
customers having had interaction with the businesses using Kustomer’s CRM
software.
538
(350) Taking into account all of the above the Commission considers that Meta (formerly
Facebook) could obtain access to a significant volume of new types of valuable data
or at least significant new volumes of existing types of valuable data through its
acquisition of Kustomer.
(351) As regards the value of the data Meta (formerly Facebook) would obtain, the
Commission considers, and the Notifying Party does not dispute, that any Customer
Transaction Data could be highly relevant for improving Meta (formerly Facebook)’s
advertisement targeting capabilities as well as for getting better insight in
conversions from advertising campaigns to actual sales (the return on investment). It
appears that these data that Meta (formerly Facebook) might obtain are particularly
valuable as regards the profiling and targeting relating to the end-customers
interacting with the businesses that use Kustomer’s CRM software. However, these
data that Meta (formerly Facebook) might obtain are also valuable for modelling and
profiling of users at large (i.e., extrapolating or ‘optimising’ data obtained for the
profiling of users that are not interacting themselves with businesses that use
Kustomer’s CRM software). Finally, Other Event Data could potentially be used for
very targeted, product-specific ads that might achieve high conversion ratios.
(352) The value of the data as it relates to individual customers is [Business Plans]. It is
stated that Meta (formerly Facebook) [Business Plans].
539
(353) In another internal document, [Business Plans] is discussed. [Business Plans].
540
(354) The results of the market investigation confirm the value of the data shared by
business customers with a customer service and support CRM provider such as
Kustomer. One responding CRM provider indicates that data stored on customer
service and support CRM software is generally useful for online display advertising:
“[d]ata in a customer service CRM platform can be useful for online display
advertising because it can reveal customer brand loyalty and sentiment, propensity
to spend wealth, specific intent (e.g. item added to shopping cart), post-purchase
experience (e.g. was item returned?), etc. This data would be valuable to online
advertisers so they could place a more accurate value on an advertising
impression.”
541
(355) Furthermore, a large majority of the respondents indicate that the access to this data
is important or very important for Meta (formerly Facebook).
542
One respondent
states that it “it could substantially increase their targeting ability and be a game
changer in differentiating their platform as an ads solution even further than it is
now”.
543
Another respondent adds that “[b]y using Kustomer’s CRM data, Facebook
538
Replies to questionnaire 3 to B2C communication service providers and online ads service providers,
question 68.
539
Form CO, Attachment E, […].
540
Form CO, Attachment E14223, page 2.
541
Reply to questionnaire 1 to CRM software providers, question 61.1 (ID: 762).
542
Replies to questionnaire 3 to B2C communication service providers and online ads service providers,
question 65; replies to questionnaire 4 to business customers, question 67.
543
Replies to questionnaire 3 to B2C communication service providers and online ads service providers,
question 65.1 (ID: 741).
106
will be able to build “super-profiles” about its users, enabling Facebook to increase
its already dominant market share in the online advertising market”.
544
(356) Several respondents also refer to the impact of the rapidly evolving privacy
landscape for the availability of customer transaction data. One respondent states in
this regard that “due to changes in the regulatory landscape and the shift towards
differential treatment of third party cookies (e.g., by browsers), having access to this
level of first party data is increasingly valuable and important for companies that
provide targeted advertising”.
545
Further, the ability of certain companies to block
Meta (formerly Facebook)’s access to data through, for example, Apple’s new
privacy policy relating to its mobile operating system iOS is also considered relevant
by respondents. One respondent states: “[i]t has a large impact, consumers need to
opt int [sic] to be targeted and tracked with relevant metadata in advertising
targeting. Opting out, through IOS, removes the ability to target users based on
specific interests incurred within the mobile application.”
546
(357) On this basis, the Commission considers that the value of Kustomer’s data may only
increase over time in particular because Meta (formerly Facebook) would receive
data directly from businesses without being dependent on third parties, such as CRM
software competitors. After all, the Customer Transaction Data and Other Event Data
that Meta (formerly Facebook) currently collects through the use of its Meta
(formerly Facebook) Business Tools (including e.g., Facebook Pixel, Conversions
API, Offline Conversions API, App Event API etc.)
547
might become less readily
available due to more limited possibilities to track end-users outside Meta (formerly
Facebook)’s own platforms, comparatively increasing the value of a direct source of
data such as Kustomer.
(358) In addition, the Commission notes that Kustomer itself generates data from
Kustomer’s own proprietary webchat, i.e., anything an end-customer types into the
webchat. Meta (formerly Facebook) would likely have access to this data following
the Transaction, and could use it to target advertising to those end-customers or other
end-customers more generally. The Parties have [Business Plans]. Kustomer states
that [Business Plans].
548
[Business Plan²]. This data may therefore constitute a
valuable addition for Meta (formerly Facebook).
(359) Taking into account all of the above, the Commission considers that the data that
Meta (formerly Facebook) could obtain through the acquisition of Kustomer is
highly valuable to improve its online ads services.
(360) Also as regards the variety of data, the Commission mentioned above that the data
that Meta (formerly Facebook) could obtain through its acquisition of Kustomer
would permit it to “close the loop”, by having access to data at all stages of a
transaction (discovery, inquiry, purchase, delivery, support and retargeting/upsell).
Further, the Transaction may allow Meta (formerly Facebook) to get access to data
relating to businesses that might not (yet) be active on some or all of Meta (formerly
Facebook)’s platforms, to data relating to customers of those businesses that might
544
Reply to questionnaire 3 to B2C communication service providers and online ads service providers,
question 65.1 (ID: 886).
545
Reply to questionnaire 3 to B2C communication service providers and online ads service providers,
question 65.1 (ID: 873).
546
Reply to questionnaire 3 to B2C communication service providers and online ads service providers,
question 65.2 (ID: 723).
547
Form CO, footnote 227.
548
Parties’ reply to RFI 17, question 5.
107
not be active on (all of) Meta (formerly Facebook)’s platforms, or interactions
between businesses and their customers that take place outside Meta (formerly
Facebook)’s platforms. Finally, Meta (formerly Facebook) might also get access to
the entire purchase and interaction history of a particular customer. The Commission
therefore considers that the data that Meta (formerly Facebook) could obtain through
the acquisition of Kustomer is highly varied.
(361) Finally, as regards the velocity of data, the Commission notes that, by its nature, data
that is stored in a business’ CRM software is the most complete and the most recent
(where it relates to direct interactions) data that can be collected. The Commission
therefore considers that the data that Meta (formerly Facebook) could obtain through
the acquisition of Kustomer is significant in terms of volume, value, variety and
velocity.
(362) On the basis of all of the above, the Commission considers that the additional data
Meta (formerly Facebook) could obtain following the acquisition of Kustomer is
highly relevant for improving Meta (formerly Facebook)’s targeting capabilities for
the online display advertising services it provides. Additional data would therefore
constitute a significant benefit from foreclosure.
C. CTM ads
(363) By engaging in a foreclosure strategy and steering businesses away from their
current CRM provider to Kustomer, Meta (formerly Facebook) would expect
[Business Plans].
549
(364) CTM ads are a format of advertising available to advertisers. CTM ads are generated
when end-customers click on an ad and are taken to a WhatsApp, Messenger or
Instagram messaging experience. The Parties state that the key driver of this
incremental revenue is that Kustomer’s business customers will be more likely to
buy CTM ads where they are using a CRM tool that is optimised to work well with
messaging channels (i.e., a business is more likely to buy CTM ads if that business is
well-positioned to handle the resulting messages).
550
(365) Meta (formerly Facebook) expects to [Business Plans].
551
[Business Plans].
(366) In this regard, Meta (formerly Facebook) has estimated that in [Year] it would earn
on average USD [Revenue information] profit a year from each business customer
that switches to Kustomer.
552
Meta (formerly Facebook) has estimated that Kustomer
may be a feasible CRM software option for [Customer Information] businesses. This
can be contrasted to Meta (formerly Facebook)’s projection of growing Kustomer’s
customer portfolio from [Customer Information] businesses today to [Customer
Information].
553
If all the customers for which Kustomer may be a feasible solution
were to switch to Kustomer as a result of the foreclosure strategy, this would imply a
maximum possible gain of over USD [Revenue Information] a year from CTM ads
alone as a result of such a foreclosure strategy. As noted above, while the actual gain
549
See, for example, Meta (formerly Facebook)’s internal document, Attachment E029 – […]; see Parties’
reply to RFI 1, question 33.
550
Parties’ reply to RFI 1 question 33.
551
Form CO, Figure 3.4; Parties’ reply to RFI 1 question 33.
552
This figure assumes that the incremental ads revenue [Internal Assumptions and Ad Revenue
Projections]. See Parties’ reply to RFI 1, question 33. For the avoidance of doubt, this revenue [Internal
Assumptions and Ad Revenue Projections], which is taken from Meta (formerly Facebook)’s internal
document, Attachment E20128 – […].
553
Meta (formerly Facebook’s internal document, Attachment, E029 – […].
108
depends on the number of business customers who would switch to Kustomer as a
result of the foreclosure strategy, it shows at least that there is a large pool of
businesses out of which Meta (formerly Facebook) would potentially be able to gain
new customers by engaging a foreclosure strategy. And, as outlined in
Section 7.3.2.2.(b.ii), it seems likely that a significant proportion of the relevant
market would switch to Kustomer following a foreclosure strategy. The Commission
notes that Kustomer’s own estimates for the number of potential business customers
it could serve post-Transaction are far higher than those estimated by Meta (formerly
Facebook), which would imply a higher maximum possible gain from foreclosure.
554
(367) CTM ad revenues are also growing. Meta (formerly Facebook) currently earns USD
[Revenue Information] globally from CTM ads and in fact estimates that [Revenue
Information] of its revenues related to B2C messaging will come from CTM ads.
555
By [Year], Meta (formerly Facebook) forecasts generating USD [Revenue
Information] globally from CTM ads.
556
It is unclear however how reliable these
forecasts are. Meta (formerly Facebook) separately provided an estimate that its
[Year] CTM ad revenues would reach USD [Revenue Information], but when asked
to provide detail on the basis for these forecasts, Meta (formerly Facebook) conceded
that they “are contingent on a range of assumptions that are subject to considerable
uncertainty and hence there is significant uncertainty around the forecasts
themselves.”
557
A model from Meta (formerly Facebook), used to estimate the SaaS
revenue to be gained as a result of the Transaction, assumes that the growth in
business messaging will [Business Plans] year-on-year
558
, and, in relation to this
estimate, the Parties confirmed that [Business Plans].
559
For illustration, an annual
[Business Plans] growth corresponds to an almost [Business Plans] increase over
[Year]. As such, one expects that this benefit from foreclosure will become more and
more significant in the future.
(368) In a similar regard, Meta (formerly Facebook) has indicated that only Twitter and
Snapchat may currently offer CTM ads as part of their respective ads portfolios.
560
The benefits from CTM ads would therefore be almost unique to Meta (formerly
Facebook). Business customers would have little to no alternative for the supply of
CTM ads. As such, one expects that this benefit from foreclosure will remain
significant in the future.
(369) The Form CO and the Parties’ response to various RFIs further supports the
conclusion that there is a causal link between businesses switching to Kustomer and
incremental CTM ad revenue for Meta (formerly Facebook), irrespective of the
554
Form CO, Attachment E0551. This document prepared by Kustomer includes strategies for Kustomer to
win customers following its possible acquisition by Meta (formerly Facebook), and estimates Kustomer
could potentially gain between [40,000-300,000] Torso customers post-Transaction, which is more than
the critical number of businesses estimated by any version of the model submitted by the Parties. In the
Parties’ reply to RFI 10, Meta (formerly Facebook) stated that it does not consider Kustomer’s growth
projections as set out in Attachment E0551 to be a realistic or reliable benchmark against which to
assess the extent to which it would be able to profitably acquire new customers post-Transaction (with
or without foreclosure).
555
Parties’ reply to RFI 1, question 43.
556
Parties’ reply to RFI 1, question 31.
557
Parties’ reply to RFI 22, question 2.
558
Form CO, Attachment E0686.
559
Parties’ reply to RFI 10, question 21.
560
Parties’ reply to RFI 6, question 15.
109
reason for or cause of such switching (i.e., whether as a result of foreclosure or
otherwise).
(370) In the Form CO, the Notifying Party notes that: [Regulatory Submission].
561
(371) In response to RFI 1, the Notifying Party notes that: “to the extent Kustomer’s
customers do choose to do these things as a result of the Transaction, that is because
those choices benefit their business by providing a greater return on investment for
CTM ad purchases and ultimately benefit consumers by providing them with more
personalized ads”.
562
(372) In response to RFI 6, the Notifying Party notes: “CTM ads are generated when end-
customers click on an ad and are taken to a WhatsApp, Messenger or Instagram
messaging channels. The better businesses can manage messaging traffic and build
consumer trust and engagement through those channels, the more they may decide to
use messaging to capture queries that can be converted to sales following a B2C
conversation - for example, through increased usage of CTM ads. Where a business
is using CRM software that is optimised to work well with messaging channels they
may have the capacity and resources to better manage the increased messaging
volume that one would hope to receive as a result of a CTM ads campaign”.
563
The
Notifying Party notes further that: “while most major CRM providers offer some
degree of integration with messaging channels (e.g. providing the ability to read and
respond to messages within the CRM software), the integration between messaging
channels, other communication channels and businesses’ systems is often clunky”
564
,
which may indeed explain why the Notifying Party considers that switching to
Kustomer (as a result of foreclosure) will result in increased CTM ad revenues.
(373) In response to RFI 9, the Notifying Party notes that “Facebook estimates the rate of
CTM ad incrementality within a range of [Regulatory Submission]”.
565
(374) In response to RFI 10, the Notifying Party notes that it forecasts receiving “an uplift
in CTM revenue. In particular, Facebook forecasts that it will generate [Revenue
Information] through increased sales of CTM ads to businesses using Kustomer.
[Business plans regarding CTM Revenue]”.
566
The Notifying Party goes on to state
that “The vast majority of businesses use email, phone and SMS to communicate with
customers. CRM software from more established providers like Salesforce, Oracle
and SAP, as well as some CPaaS software, is built around these channels,
….[while]… Kustomer, on the other hand, is a messaging-first customer support
CRM software by design and [Business plans regarding CTM Revenue]. The better
businesses can manage messaging traffic and build consumer trust and engagement
through those channels, the more they may decide to use messaging to capture
queries that can be converted to sales following a B2C conversation - for example,
through increased usage of CTM ads”.
567
The Notifying Party goes on to conclude
that “This is not because those businesses [i.e., using established CRM software
providers] would be restricted in any way from purchasing CTM ads but simply
because they may be less equipped to handle inbound enquiries from a Facebook
561
Form CO, footnote 239.
562
Parties’ reply to RFI 1, question 36c.
563
Parties’ reply to RFI 6, question 24.
564
Parties’ reply to RFI 6, question 24.
565
Parties’ reply to RFI 9, question 2.
566
Parties’ reply to RFI 10, question 10.
567
Parties’ reply to RFI 10, question 10.
110
CTM ad campaign if they do not have a way of dealing with these at scale (e.g.
through CRM, CPaaS or in-house technology)”.
568
(375) On the basis of the above, the Commission considers that the benefits Meta (formerly
Facebook) could gain from CTM ads by a foreclosure strategy are significant.
D. Meta (formerly Facebook)’s ecosystem of products
(376) By engaging in a foreclosure strategy and steering businesses away from their
current CRM provider to Kustomer, Meta (formerly Facebook) will also be able to
gain by bringing these businesses into the Meta (formerly Facebook) “ecosystem”,
i.e., Meta (formerly Facebook)’s suite of business and personal products (including
Meta (formerly Facebook)’s ecommerce products, e.g., Marketplace and Shops, and
Workplace). The Transaction would provide Meta (formerly Facebook) with another
avenue through which it could bring businesses into the Meta (formerly Facebook)
ecosystem and purchase its products, i.e., cross-selling. In this case, Meta (formerly
Facebook) would make profits directly from business customers switching to
Kustomer, not solely from services provided by Kustomer but also other services
provided by Meta (formerly Facebook) within its ecosystem (e.g., Facebook
Marketplace; Facebook Shops, Workplace; advertising services). Meta (formerly
Facebook) therefore has an incentive to steer businesses to Kustomer, and thereby
into the Meta (formerly Facebook) ecosystem. This applies equivalently to steering
businesses to Kustomer through input foreclosure of competing CRM software
providers. Further, Meta (formerly Facebook) would have an incentive to both bring
new business customers into the Meta (formerly Facebook) ecosystem through
Kustomer, as well as bring existing business customers further into the Meta
(formerly Facebook) ecosystem through Kustomer, in order to encourage additional
use of its ecosystem and in particular of its ecommerce products by those existing
business customers.
(377) Meta (formerly Facebook) offers two key ecommerce products in the EEA,
Facebook Marketplace and Shops. Marketplace allows users to browse and search
for items, and users can post listings for items they want to sell. Shops is a virtual
storefront in which businesses can showcase their inventory, create collections to
attract customers and operate a shared storefront across Facebook and Instagram.
Meta (formerly Facebook) has indicated that the Transaction is part of a wider
strategy (intended to create better experiences for users as they interact with
businesses) that also includes innovations and initiatives with Shops.
569
(378) Meta (formerly Facebook)’s internal documents indicate that [Business Plans]
570
[Business Plans].
(379) Similarly, it appears that a foreclosure strategy, steering business customers into
Meta (formerly Facebook)’s ecosystem, would be consistent with Meta (formerly
Facebook)’s public statements around its future business decisions. Indeed, Meta
(formerly Facebook) makes links across its range of products when talking about
“the trajectory of the company over the long term”, for example, Mark Zuckerberg,
CEO of Meta (formerly Facebook), states that “this modern commerce system is
568
Parties’ reply to RFI 10, question 10.
569
Parties’ reply to RFI 20, question 7.
570
See, for example Meta (formerly Facebook)’s internal document, Attachment E0548 – […]; Meta
(formerly Facebook)’s internal document, Attachment E0419 – […]; Meta (formerly Facebook)’s
internal document, Attachment E0409 – […]; Meta (formerly Facebook)’s internal document,
Attachment E0936 – […].
111
going to bring together a number of areas where we either already have strong
offerings – like in ads, community tools, and messaging -- with areas like Shops,
business messaging, and payments that we're focused on ramping up now”.
571
(380) In this regard, and in the context of Meta (formerly Facebook)’s ability and incentive
to grow Kustomer, Zendesk (a close competitor of Kustomer) draws together and
mentions the various business activities of Meta (formerly Facebook): “the
[T]ransaction will have a big impact on ecommerce in general and the open web. It
will result in a company that has an overview over every step of the buying
process/customer journey, which is a big value proposition to companies seeking to
advertise on Facebook. There is a real prospect for Facebook to use its ads business
to subsidize its CRM offering. Possibly this will result in a substantially below
market rate, and Kustomer being positioned as part of their advertising solution.
This could especially be destructive in B2C ecommerce.”
572
(381) The Commission notes that the merged entity may have additional longer-term gains
from steering business customers to Kustomer. In particular, there may be longer-
term benefits from increasing Meta (formerly Facebook)’s ability to defend its
position in the highly valuable B2C messaging market in the future (if the
foreclosure strategy enables Meta (formerly Facebook) to achieve a significant
market power on the downstream market). However, these longer-term benefits are
difficult to quantify, and have therefore been omitted from the Commission’s
analysis (since they are not necessary to reach the Commission’s conclusions).
(382) In light of the above, the Commission considers that steering businesses into the
Meta (formerly Facebook) ecosystem, to then cross-sell them Meta (formerly
Facebook)’s other products, constitutes a significant benefit in the long-term from
foreclosure. The Commission notes that this gain is not quantifiable, and is therefore
(conservatively) not included in the Commission’s quantitative analysis.
(b.ii) Meta (formerly Facebook) can limit losses from foreclosure
(383) The Commission considers that Meta (formerly Facebook) has the ability to limit its
losses from an input foreclosure strategy. As outlined in Section 7.3.2.2.(a.iii) above,
Meta (formerly Facebook) exercises a high level of control over the API access to its
channels. And, as outlined in Section 7.3.2.2.(a.i) above, Meta (formerly Facebook’s
channels are an important input for not only CRM software providers but also their
business customers.
573
As such, the Commission considers that Meta (formerly
Facebook) will have the ability to limit its losses from foreclosure by targeting
foreclosure to Kustomer’s close competitors (including potential entrants who would
become close competitors). Doing so would maximize the number of business
customers switching to Kustomer (downstream) whilst minimizing the number of
business customers switching away from Meta (formerly Facebook) (upstream). Sub-
section A proceeds by first outlining the potential losses to the merged entity
following foreclosure. Sub-section B then proceed by demonstrating that the
available evidence suggests a significant proportion of the business customers of
571
Transcript of Meta (formerly Facebook)’s First Quarter 2021 Results Conference Call, April 28th,
2021; available at: https://s21.q4cdn.com/399680738/files/doc_financials/2021/Q1/FB-Q1-2021-
Earnings-Call-Transcript.pdf, last accessed on 15 December 2021 (ID: 2815).
572
Agreed minutes of the call of 9 September 2021 with Zendesk, paragraph 12 (ID: 2085).
573
Zendesk notes that “Facebook’s messaging channels are becoming increasingly important, in
particular, for consumer brands” and “a significant portion of the market strongly values access to
Facebook’s B2C messaging channels” (Agreed minutes of the call of 9 September 2021 with Zendesk,
paragraph 6 (ID: 2085)).
112
Kustomer’s close competitors would switch to Kustomer, whilst the business
customers of less close rivals would remain unaffected.
574
Second, the Commission
considers that there are potentially further aggravating factors that would allow Meta
(formerly Facebook) to further encourage switching to Kustomer and thereby
increase the incentive for Meta (formerly Facebook) to foreclose. These factors are
discussed in the Sub-section C below.
A. Losses from foreclosure
(384) By engaging in a foreclosure strategy, Meta (formerly Facebook) would potentially
lose (i) revenues from providing API access to its messaging channels or from usage
of the API (for those APIs that Meta (formerly Facebook) offers against payment
today or in the future); and (ii) messaging-related revenues (i.e., CTM ads and paid
messaging services associated with businesses who stop using Meta (formerly
Facebook)’s messaging channels as a result of the foreclosure).
(385) As stated by the Parties, there may also be damage to Meta (formerly Facebook)’s
reputation following a foreclosure strategy, which may in turn lead also to losses in
messaging-related revenues (and is therefore accounted for in the Commission’s
assessment of messaging-related losses below).
575
However, regarding wider-
reaching reputational damage outside of losses in messaging-related revenues,
evidence submitted by third parties indicates that any further losses outside
messaging-related revenues would likely be borne by third parties and not Meta
(formerly Facebook), because the reputational harm is more likely to fall on the
foreclosed CRM software provider and business customers than on Meta (formerly
Facebook) further upstream. Salesforce notes that “[input foreclosure] could, in
principle, have a negative impact on both business users and third-party CRM
providers. First, [input foreclosure] could have a reputational impact on business
users since end customers might blame them for a bad customer experience, not
knowing it was because Facebook had altered API features. Second, it could have a
reputational impact on third-party CRM providers. […]. Were such a strategy to be
pursued, it could in principle result in Salesforce’s customers switching away to
Kustomer.”
576
(386) Regarding losses of revenues from providing API access to its messaging channels,
Meta (formerly Facebook) [Revenue Information], specifically it received USD
[Revenue Information] in revenue in 2020 from WABAPI, which accounts for less
than [Revenue Information] of Meta (formerly Facebook)’s revenue worldwide.
Meta (formerly Facebook) has stated that its forecast WABAPI revenues will reach
USD [Revenue Information] by 2022 and that its gross margin on WABAPI is
[Business Plans], which implies a maximum loss of USD [Business Plans].
577
However, Meta (formerly Facebook) earns this profit from many businesses, around
[Customer Information]. And so, the average loss per business customer switching
574
Business customers who may switch to these less close rivals from Kustomer’s close competitors would
also have their access to Meta (formerly Facebook)’s B2C messaging unaffected, and so would (all else
being equal) not alter their use of Meta (formerly Facebook)’s B2C messaging channels.
575
As such, for a total foreclosure strategy, any losses associated with a damaged reputation will be
covered by the losses in messaging-related revenues. This is in line with the Parties’ approach taken in
their economic submission, which does not include any incremental loss associated with a damaged
reputation (beyond any losses in messaging-related revenues).
576
Agreed minutes of the call of 28 September 2021 with Salesforce, paragraph 4 (ID: 2151).
577
Parties’ reply to RFI 9, Annex 9.002.
113
away from Meta (formerly Facebook) would only amount to USD [Revenue
Information].
578
(387) Regarding losses of messaging-related revenues, Meta (formerly Facebook) indicated
that it earned USD [Revenue Information] in revenue globally from CTM ads
in 2020,
579
but clarified that [Revenue Information] of this was generated from “tail”
businesses […], i.e., which typically do not need, and would not pay for a CRM tool
(and would therefore likely not be the targets of foreclosure).
580
Therefore, the
remaining [Revenue Information] or USD [Revenue Information] in 2020 CTM ad
revenues were earned through non-“tail” businesses (i.e., which could be potential
target customers of Kustomer). The remaining analysis conservatively assumes that
all non-“tail” business customers generate CTM ad revenues via a CRM software
provider,
581
and therefore that all CTM ad revenue associated with these business
customers could potentially be lost as a result of foreclosure . However, the
Commission notes that the actual losses are therefore likely below those estimated in
the following analysis. By 2022, Meta (formerly Facebook) forecasts generating
USD [Revenue Information] globally from CTM ads.
582
Assuming the proportion of
“tail” businesses will stay roughly the same between 2022 and 2020, USD [Revenue
Information] of these CTM ad revenues will be generated by business customers that
would use CRM software. Meta (formerly Facebook)’s internal documents indicate
the gross margin on its ads services is [Business Plans], which implies a maximum
loss of USD [Revenue Information].
583
Meta (formerly Facebook) has indicated it
earns this profit from over [Customer Information] businesses,
584
and so the average
loss per business customer switching away from Meta (formerly Facebook) would
only amount to USD [Revenue Information].
585
The Commission notes that the
578
Parties’ reply to RFI 9, Annex 9.002.
579
Notifying Party Submission on Foreclosure Issues dated 10 October 2021, Figure 2.2.
580
Notifying Party Submission on Foreclosure Issues dated 10 October 2021, paragraphs 1.18 and 2.18.
The Commission notes that a further […]% was generated from “head” and “hair” businesses [Business
Plans], for which Kustomer’s “torso-focussed [i.e., SMB and mid-sized business focused] product is not
suitable”. The following figures and calculations conservatively include these business customers.
581
As explained in paragraph 391 below, [Confidential Meta information on CTM ad spending with Meta
by businesses using CRM software].
582
Parties’ reply to RFI 22, question 2. When asked to provide detail on the basis for these forecasts, Meta
(formerly Facebook) conceded that they “there is significant uncertainty around the forecasts
themselves.”
583
Parties’ reply to RFI 9, Annex 9.002.
584
Parties’ reply to RFI 23, question 2. These calculations are based on the best and most granular data that
Meta (formerly Facebook) was able to provide. While the average upstream loss was calculated across a
larger sample of businesses than the corresponding calculation of average downstream gains, there is no
indication that the estimates would be biased in any direction as a result. In particular, the Commission
has seen no evidence to suggest that the average CTM ad spend of the approximately [Smaller Target
Customer Group] business customers most addressable by Kustomer (which were the sample size used
to calculate downstream losses) is significantly higher than the average CTM ad spend of the
approximately [Larger Target Customer Group] business customers identified as “Torso advertisers”
(which was the sample size used to calculate upstream losses, and which most likely includes the
[Smaller Target Customer Group] business customers most addressable by Kustomer).
585
Parties’ reply to RFI 9, Annex 9.002. The Commission notes that these calculations imply a
significantly higher downstream gain from incremental CTM ad spend compared to the downstream
loss of baseline CTM ad spend. However, first, the Commission notes that these figures were calculated
using the Parties’ own data. Second, whilst the figures imply a […] increase in CTM ad spend (post-
switching to Kustomer), in absolute terms the increase in CTM ad spend is only approximately USD
[…] (i.e., less than the amount the same business would typically spend on CRM software, i.e.,
approximately USD […]). Third, the Parties’ own calculations imply a roughly […] increase in CTM ad
spend (post-switching to Kustomer). See Parties’ reply to RFI 9, Annex 9.002, Q26, Scenario 1 (also
114
Parties’ analysis attributes all the revenues earned from these [Customer
Information] businesses to the [Customer Information] business customers most
addressable by Kustomer, which thereby drastically overstates the average loss per
business customer switching away from Meta (formerly Facebook).
586
This single
extreme assumption drives in large part the Parties’ conclusion that foreclosure
would not be profitable for the merged entity.
(388) As outlined in the Parties’ economic submission
587
, Meta (formerly Facebook)’s
losses from foreclosure are limited to those business customers that switch away
from Meta (formerly Facebook) (or use its channels less, and including switching
away to potential entrants), i.e., those business customers that do not switch away
from the foreclosed CRM software competitors to Kustomer (or CRM software rivals
that are not foreclosed).
(389) In this context, the Commission notes that it would not require a high degree of
switching for losses from foreclosure to be outweighed by the gains, i.e., for
foreclosure to be profitable. Even conservatively accounting for only the quantifiable
gains and losses (which excludes some of the significant gains), the merged entity
gains substantially more per business customer switching to Kustomer than it would
lose per business customer switching away from Meta (formerly Facebook)’s B2C
channels, USD [Revenue Information] and USD [Revenue Information]
respectively.
588589
This implies that, for foreclosure to be unprofitable (i.e., the losses
to outweigh the gains), for each business customer switching to Kustomer there
would need to be more than [40-50] business customers switching away from Meta
(formerly Facebook)’s B2C channels, i.e., a [40-50]:1 critical switching ratio. In
other words, if more than [0-5] of business customers of foreclosed CRM software
providers switch to Kustomer, foreclosure would be profitable. This figure (i.e., the
critical switching percentage for foreclosure to be profitable) would be even lower if
assuming that incremental ads revenue will be split between additional data and CTM ad revenues
consistently over time).
586
The Parties assume the entire CTM ad revenue from [Business Plans] businesses come from […]
business (which are most likely to consider Kustomer), but provide no evidence as to why the revenue
is distributed in such an extreme way. See Form CO, Annex 13.17, Acquisition by Facebook, Inc. of
Kustomer, Inc. Analysis of Vertical Effects, footnote 72. LoF Response, paragraph 3.7, argues (with no
sources or additional evidence) that the […] businesses include smaller businesses that may have a
smaller CTM ad spend, which would skew the estimate for the average CTM ad spend for the […]
businesses (which are most likely to consider Kustomer) downwards. However, the Parties’ analysis
takes this to the extreme by assuming that less than [0-5]% of businesses included in this figure earn
[90-100]% of the revenue, i.e., that [90-100]% of these “Torso” businesses have no CTM ad spend. In
contrast, the same data indicates that the average CTM ad spend of “tail” businesses (who would have a
smaller CTM ad spend than the smaller businesses included in the “Torso” segment) is USD […].
587
Form CO, Annex 13.7, Acquisition by Facebook, Inc. of Kustomer, Inc. Analysis of Vertical Effects,
pages 21-24.
588
Parties’ reply to RFI 9, question 26. This also includes USD […] profit which Meta (formerly
Facebook) estimates it would on average earn from incremental WABAPI revenues from each business
customer that switches to Kustomer.
589
The Notifying Party implicitly argues in the SO Response, paragraph 4.8(B)(ii)(d), that the CTM ad
gain for every new business customer of Kustomer should be roughly equal to the CTM ad loss for
every business customer lost by Meta (formerly Facebook) that was using a third party CRM software
provider. However, this is misleading. The Commission notes that the CTM ad losses are incurred from
business customers not using Kustomer but a third party CRM software provider; whereas the CTM ad
gains are from business customers using Kustomer. This is consistent with the Transaction rationale that
business customers using Kustomer (will) spend more on CTM ads than Meta (formerly Facebook)’s
average business customer or a business customer using a third party CRM software provider. As such,
one would not expect the CTM ad gain to be roughly similar to the CTM ad loss.
115
one were to include the unquantifiable gains from foreclosure. In this regard, the
Commission considers that the evidence on switching in relation to foreclosure
outlined in the following sub-Section is sufficient to demonstrate that losses from
foreclosure would be limited relative to the gains.
(390) The Commission received further evidence from the Parties that corroborates the
magnitude of the critical switching ratio. In particular, in the reply to RFI 23, the
Parties provided the CTM ad spend that can be linked to business customers using a
CRM to connect to Meta (formerly Facebook)’s channels. These data seek to more
directly estimate the CTM ad revenues attributed to the approximately [Customer
Information] business customers most addressable by Kustomer. And, using these
alternate figures results in a critical switching ratio of [30-40]:1, which broadly
corroborates in terms of magnitude the critical switching ratio calculated in
paragraphs 387-389 above. These data indicate [Revenue Information], which
suggests that these revenues would not be lost as a result of foreclosure.
(391) In the reply to RFI 23, the Parties indicated that the “CTM revenues in 2020 that can
be linked to businesses using a CRM to connect to the Messenger API” was only
USD [Revenue Information].
590
The Parties also conceded that Meta (formerly
Facebook)’s Messenger channel accounted for [Revenue Information] of its total
CTM ad revenues.
591
Therefore, assuming that the proportion of CTM ads linked to a
CRM is the same for Instagram and WhatsApp as it is for Messenger, the CTM
revenues in 2020 that can be linked to businesses using a CRM to connect to Meta
(formerly Facebook)’s APIs across all of it channels would be approximately USD
[Revenue Information]. The Commission then assumes (conservatively) that all of
these CTM revenues were generated from the [Customer Information] businesses
most addressable by Kustomer,
592
which implies that CTM revenues in 2020 per
business that may consider Kustomer an appropriate CRM is USD [Revenue
Information]. Assuming (conservatively) that these CTM revenues per business in
2020 will scale in line with Meta (formerly Facebook)’s forecast for its total CTM
revenues in 2022, CTM revenues in 2022 per business that may consider Kustomer
an appropriate CRM is USD [Revenue Information].
593
As such, following the
calculations made in paragraph 389 above, the merged entity still gains substantially
more per business customer switching to Kustomer than it would lose per business
customer using a third party CRM and switching away from Meta (formerly
Facebook)’s B2C channels, USD [Revenue Information] and USD [Revenue
Information] respectively.
594
This is mainly driven by [Revenue Information].
595
This implies that, for foreclosure to be unprofitable, for each business customer
590
Parties’ reply to RFI 23, question 1, Table 1.
591
Parties’ reply to RFI 22, question 1c, Table 2.
592
This is a conservative assumption because the Parties indicate these revenues were earned from over
[…] businesses. In this regard, it is not necessarily unconservative to use this figure to calculate the
average CTM ad loss per business, even if the figure does not include all of Meta’s CTM revenues
linked to a CRM. It would only be unconservative to use this figure if the more than […] businesses
captured by this figure on average had a smaller (by a factor of […]) CTM ad spend than those […]
businesses most addressable by Kustomer. See LoF Response, paragraph 3.7(D).
593
Parties’ reply to RFI 9, Annex RFI 9.002. The Parties state that Meta (formerly Facebook)’s revenues
from CTM ads was USD [Revenue Information] in 2020 and will be USD [Revenue Information]
in 2022.
594
Note that the loss per business customer includes the USD [Revenue Information] and USD [Revenue
Information] losses per business with respect to messaging-related revenues and API access/usage
revenues respectively.
595
See paragraphs 366, and 387-389 above.
116
switching to Kustomer there would need to be more than [30-40] business customers
switching away from Meta (formerly Facebook)’s B2C channels, i.e., a critical
switching ratio of [30-40]:1. This figure calculated with alternative data is largely in
line in terms of magnitude with the critical switching ratio calculated in paragraphs
387-389 above.
(392) Further, Meta (formerly Facebook)’s internal documents similarly indicate that
[Description of expected uptake of and spending on CTM ads].
596
A Meta (formerly
Facebook) internal document estimates that [Description of expected uptake of and
spending on CTM ads].
597
A further Meta (formerly Facebook) internal document
considers [Description of expected uptake of and spending on CTM ads].
598
Another
Meta (formerly Facebook) internal document suggests that [Description of expected
uptake of and spending on CTM ads].
599
The following Meta (formerly Facebook)
internal document suggests [Description of expected uptake of and spending on CTM
ads].
600
[Description of expected uptake of and spending on CTM ads].
601
The
following Meta (formerly Facebook) internal document also shows that internally
Meta (formerly Facebook) [Description of expected uptake of and spending on CTM
ads]
602
[Description of expected uptake of and spending on CTM ads].
B. Switching is likely to be sufficiently significant to limit losses
(393) In general, business customers are unlikely to switch away from using Meta
(formerly Facebook)’s channels for the following reasons. As laid out in
Section 7.3.2.2.(a.ii), Meta (formerly Facebook) holds market power on the upstream
market, access to its channels is an important input for the downstream market, and
CRM software providers and their customers do not view other B2C communication
channels as substitutable for Meta (formerly Facebook)’s channels. Indeed, Meta
(formerly Facebook)’s channels are often seen as a requirement, because of the
omnichannel approach to B2C communications and the high penetration rate of Meta
(formerly Facebook)’s channels among end-consumers.
603
Further, the trend of
increased use of B2C messaging will only aggravate these factors. Zendesk notes that
“a significant portion of the market strongly values access to Facebook’s B2C
messaging channels.” and that “Facebook’s messaging channels are becoming
increasingly important, in particular, for consumer brands.”
604
(394) In response to the Commission’s market investigation, a majority of CRM software
providers, a strong majority of CPaaS providers and just under half of business
customers that had a view indicated that business customers would likely or very
likely switch to another CRM software or CPaaS provider that had access (or better
access) to Meta (formerly Facebook)’s messaging channels, if (API) access to those
596
By contrast, Meta (formerly Facebook) expects business customers of Kustomer to increase their CTM
ad spend post-Transaction. See paragraphs 387-389 above.
597
Meta (formerly Facebook)’s internal document, […].
598
Meta (formerly Facebook)’s internal document, […].
599
Meta (formerly Facebook)’s internal document, […].
600
Meta (formerly Facebook)’s internal document, […].
601
Meta (formerly Facebook)’s internal document, […].
602
Meta (formerly Facebook)’s internal document, […].
603
Parties’ reply to RFI 6, question 22.
604
Agreed minutes of the call of 9 September 2021 with Zendesk, paragraph 6 (ID: 2085).
117
channels were restricted or degraded from their current CRM software or CPaaS
provider.
605
(395) Similarly, all CRM software providers and half of business customers that had a
view indicated that business customers are likely or very likely to consider switching
to Kustomer, if Kustomer’s integrations with Meta (formerly Facebook)’s messaging
channels were significantly superior (e.g., in terms of functionality) to the
integrations available to rival customer service and support CRM tools.
606
(396) The Notifying Party argues that businesses switch CRM provider infrequently as it
involves significant time and costs.
607
Even if this may be more so the case for large
‘enterprise’ businesses that may require a high degree of customisation in their CRM
tool, this is not necessarily as true of SMBs (and even to an even lesser extent for
businesses not currently using a CRM tool). In response to the Commission’s market
investigation, those businesses that indicated that they tend to switch customer
service and support provider infrequently were all large businesses with more than
[Regulatory Submission] employees, while those that indicated they switch relatively
often were predominantly SMBs ([Regulatory Submission] employees), in addition
to certain companies with more than [Regulatory Submission] employees.
608
That
switching CRM provider is not unduly costly or time consuming for SMBs was also
confirmed by Crisp, a customer service and support CRM software provider and
close competitor of Kustomer, which stated that “switching would be straightforward
since most CRM software tools offered to SMBs and mid-sized businesses are off-the-
s[h]elf/‘plug-and-play’ (e.g. the solutions offered by Hubspot, Crisp, etc.), and
require very little customisation”.
609
Twilio, a major CPaaS provider indicated that
“it considers that Kustomer’s product is more an “off the shelf product” [i.e., one
that did not involve significant customisation].”
610
which suggests Kustomer’s tool is
one that businesses, in particular SMBs and businesses not currently using a CRM
tool, could switch to with relative ease. This point was implicitly confirmed by the
Notifying Party in that it accepts that “[Regulatory Submission]”.
611
(397) Indeed, an internal document from Meta (formerly Facebook) states (in the context
of the Transaction) that “[Business Plans]”.
612
(398) Further evidence of ease of switching can be found in an email sent by […], to a
potential customer of Kustomer: “[Business Plans].”
613
Ease of switching is also
evidenced by the various testimonials from small businesses (i.e., with a turnover of
less than $50 million) on Gartner's website
614
, which often mention a deployment
period of less than 3 months.
615
Further evidence of ease of switching can be found
605
Replies to questionnaire 1 to CRM software providers, question 54; replies to questionnaire 2 to CPaaS
providers, question 42; replies to questionnaire 4 to business customers, question 63.
606
Replies to questionnaire 5 to business customers, question 14; replies to questionnaire 7 to CRM
software providers, question 43.
607
SO Response, paragraph 4.14.
608
Replies to questionnaire 5 to business customers, questions 1.1 and 10.
609
Agreed minutes of the call of 15 September 2021 with Crisp, paragraph 8 (ID: 2132).
610
Agreed minutes of the call of 14 June 2021 with Twilio, paragraph 13 (ID: 2029).
611
Notifying Party Submission on Foreclosure Issues dated 10 October 2021, paragraphs 1.18 and 2.18.
612
Meta (formerly Facebook)’s internal document, […].
613
Kustomer’s internal document, […].
614
See: https://www.gartner.com/reviews/market/crm-customer-engagement-center, last accessed on 29
November 2021.
615
For example: https://www.gartner.com/reviews/market/crm-customer-
engagementcenter/vendor/zendesk/product/zendesk-suite/review/view/1348935, last accessed on 29
118
in the case of prior multihoming between two CRM providers, for example in the
situation where one offers a freemium tier. For instance, one start-up commented on
Gartner's website that it had switched from Zendesk to Zoho after multihoming by
using the Zoho free/freemium tier prior to ceasing its use of Zendesk (i.e., which
would allow for training of agents on using the Zoho CRM tool and integrating Zoho
with the companies own IT infrastructure while the Zendesk CRM tool was still in
place), after which the business upgraded to the Zoho paid version: “Our company
has been using Zendesk for the last three years … Three years in though we kept
bumping up against the limitations of Zendesk put in place for our subscription
(Teams) but could not justify the high price (almost double the cost) to upgrade. We
had already started using CRM (migrated from Salesforce) and Sign (migrated from
Docusign) by Zoho... and found we could purchase the top plan for Zoho Desk for
less than we were paying Zendesk”.
616
(399) In its SO Response, the Notifying Party argues that businesses would be constrained
from switching away from a chosen CRM provider due to the integrations and
features other than OTT B2C messaging channels.
617
However, the statistic that
“[Regulatory Submission]”
618
rather suggests that having such a wide variety of
integrations (excluding OTT B2C messaging channels and other B2C
communication channels, which by definition are needed for communications
between customer service agents and end customers) are likely to not constitute a
particularly important factor for a business customer’s choice of customer service
and support CRM software provider in view [Regulatory Submission].
619
Moreover,
as made clear by Kustomer during the technical sessions in the week of 20
September 2021, Kustomer (and competing CRM software providers) [Regulatory
Submission], which against suggests that provided such an integration can be done
upon request, e.g., via a publicly accessible API, not having a particular integration
will not necessarily deter or prevent a business from switching.
(400) Therefore (and with specific reference to the market questionnaire evidence in
paragraphs 394, 395 and 396 above), it appears that in general a significant
proportion of business customers, and in particular SMBs (as well as businesses not
currently using a CRM tool), are likely to switch to Kustomer following foreclosure,
and therefore unlikely to switch away from Meta (formerly Facebook) following
foreclosure. Losses would therefore be limited.
November 2021 (ID: 2458); https://www.gartner.com/reviews/market/crm-customer-
engagementcenter/vendor/zendesk/product/zendesk-suite/review/view/1309433, last accessed on 29
November 2021 (ID: 2456); https://www.gartner.com/reviews/market/crm-customer-
engagementcenter/vendor/zendesk/product/zendesk-suite/review/view/1299728, last accessed on 29
November 2021 (ID: 2457); https://www.gartner.com/reviews/market/crm-customer-
engagementcenter/vendor/freshworks/product/freshdesk-omnichannel/review/view/1279519, last
accessed on 29 November 2021 (ID: 2464).
616
https://www.gartner.com/reviews/market/crm-customer-engagement-center/vendor/zoho/product/zoho-
desk/review/view/3710576, last accessed on 29 November 2021 (ID: 2459).
617
SO Response, paragraph 16.
618
SO Response, paragraph 4.21. In SO Response, Annex 3 paragraph 2.99 states that “Zendesk has 1,314
integrations compared to only 61 for Kustomer. This makes Zendesk businesses particularly unlikely to
switch away from Zendesk, as a substantial proportion of Zendesk’s business customers are likely to
have integration needs which are unmet by many of the potential alternative CSCRM providers (and in
particular, which are unmet by Kustomer).”
619
See, e.g., Form CO, paragraph 6.67 Table 6.13. Separately, at Form CO, paragraph 9.16, the Notifying
Party indicated that following the Transaction [Facebook internal strategy in relation to growth of
Kustomer’s customer base].
119
(401) Moreover, Meta (formerly Facebook) would have the ability to further limit its losses
from foreclosure by targeting foreclosure at Kustomer’s close competitors, who
themselves are more focussed on SMBs. As outlined in Section 7.3.2.2.(a) above,
Meta (formerly Facebook) exercises a significant degree of control over the API
access to its channels, which in turn would provide Meta (formerly Facebook) with
the ability to target foreclosure at not only specific API access-seekers but also
specific business customers of them. Meta (formerly Facebook)’s control over its
API access would also allow it to act with precision with regard to the avenues
through which it could restrict or degrade access. Meta (formerly Facebook) would
therefore have the incentive to target foreclosure at Kustomer’s close competitors
(and potential entrants who would become close competitors). This would, by
definition, increase the likelihood that business customers would switch towards
Kustomer (downstream), and decrease the likelihood that business customers would
switch away from Meta (formerly Facebook) (upstream). Meta (formerly Facebook)
would thereby limit its losses from foreclosure.
(402) For example, as outlined in Section 7.3.2.2.(a), Meta (formerly Facebook) has the
ability to target foreclosure at specific businesses of its strategic partners, e.g.,
business customers of Kustomer’s close competitor, Zendesk. Therefore, at the
extreme, Meta (formerly Facebook) could target foreclosure at specific business
customers that were most likely to switch to Kustomer. In such a case, Meta
(formerly Facebook) would make very little losses.
(403) Kustomer is a relatively small customer service and support CRM software provider
targeting mainly SMBs (as stated by the Parties themselves
620
). The following
evidence suggests that Kustomer’s close competitors are similarly sized customer
service and support CRM software providers also focussing on the same types of
businesses, SMBs. These include but are not limited to Zendesk, Intercom,
Freshworks, Gorgias, Gladly and Crisp.
(404) First, while the question whether there is a separate market for the supply of
customer service and support CRM software or whether it is part of the overall
market for the supply of CRM software is left open, Section 6.1.1.3 indicates that
customer service and support CRM software providers are closer competitors than
CRM software providers offering other functionalities. Similarly, while the question
whether the market for the supply of customer service and support CRM software (or
the broader market for CRM software overall) should be further segmented based on
business customer size is left open, Section 6.1.1.3 indicates that CRM software
providers targeting similar business customer sizes are closer competitors than those
targeting different business sizes.
(405) Second, this view of Kustomer’s close competitors is shared by the smaller customer
service CRM software providers that tend to target SMBs. Zendesk (a smaller
customer service CRM provider) considers that Kustomer is a “direct and significant
competitor of Zendesk. Both Kustomer and Zendesk target broadly the same
customers, with Zendesk serving a wider range of customers and Kustomer focusing
more on the larger customers within that range. Zendesk has lost customers to
Kustomer over the years”.
621
Crisp (a very small customer service CRM provider)
620
Notifying Party Submission on Foreclosure Issues dated 10 October 2021, paragraphs 1.18 and 2.18.
621
Agreed minutes of the call of 18 May 2021 with Zendesk, paragraph 5 (ID: 2090).
120
states that “it expects Facebook will compete, via Kustomer, with both the Crisp
freemium and paid solutions”.
622
(406) Third, this view of Kustomer’s close competitors is shared by the larger CRM
software providers such as SAP, Salesforce, Microsoft and Oracle. SAP states that
“SAP does not consider Kustomer as a close competitor, as they are only competitors
in the sense that they are both active in the customer service segment of CRM
software, but their respective offerings are rather differentiated, and have very
different customer bases.”
623
SAP also confirmed that “we don’t see Kustomer as a
competitor” in their response to the Phase I market questionnaire.
624
Salesforce
considers that it “serves all size of companies” while “Kustomer generally serves
only smaller businesses”.
625
In addition, Salesforce notes that “as businesses grow
they will tend to look for more sophisticated CRM tools”.
626
Similarly, Microsoft
indicated that while it has customers of different sizes it “primarily targets larger
business customers”
627
and it appears from public sources that the same is true of
Oracle’s CRM offering.
628
(407) Fourth, in the absence of diversion ratios from competitors, the Commission has
analysed Kustomer’s win/loss data, specifically the share of wins from
competitors.
629
Kustomer’s win/loss data indicates that in absolute terms the majority
of Kustomer’s wins from competitors came from the close competitors of Kustomer
listed above (i.e., Zendesk, Intercom, Freshworks, Gorgias, Gladly), over [Win-Loss
Data].
630
[Win-Loss Data].
(408) For example, [Win-Loss Data].
631
This example demonstrates that the switching rates
of business customers of Kustomer’s close competitors are magnitudes higher than
those of other CRM software providers. Indeed, similar observations could be made
for the competitors that account for Kustomer’s third, fourth, and sixth most wins,
[Win-Loss Data]; all of which are closer competitors to Kustomer than Salesforce.
Similar conclusions can be drawn from Kustomer’s loss data (i.e., the share of losses
to competitors); although the Commission notes that, in the context of foreclosure,
Kustomer’s win data is more relevant since one is looking at switching from
competitors to Kustomer.
(409) In light of the above, the Commission considers that the close competitors of
Kustomer are similarly sized customer service and support CRM software providers
also focussing on the same types of businesses, SMBs. In particular, the example in
the paragraph above (as well as the evidence on closeness of competition)
demonstrates that business customers of Kustomer’s close competitors are more
likely to switch to Kustomer than business customers of Salesforce, i.e., switching
622
Agreed minutes of the call of 15 September 2021 with Crisp, paragraph 6 (ID: 2132).
623
Agreed minutes of the call of 18 May 2021 with SAP, paragraph 7 (ID: 2008).
624
Reply to questionnaire 1 to CRM software providers, question 43.1 (ID: 800).
625
Agreed minutes of the call of 19 May 2021 with Salesforce, paragraphs 3 and 4 (ID: 2005).
626
Agreed minutes of the call of 19 May 2021 with Salesforce, paragraph 9 (ID: 2005).
627
Reply to questionnaire 1 to CRM software providers, question 7.1 (ID: 726).
628
See https://enlyft.com/tech/products/oracle-crm, last accessed on 15 December 2021 (ID: 2800).
629
Parties’ reply to RFI 21, Annex 3 and Form CO, paragraph 6.67 Table 6.13.
630
This figure and the rest of the analysis in this paragraph excludes […]% of wins from “home grown” or
“greenfield” projects.
631
See Salesforce. #1 in Sales, available at: https://www.salesforce.com/campaign/worlds-number-one-
SALES/, last accessed on 15 December 2021 (ID: 2803). See Meet Gorgias, the #1 rated helpdesk for
ecommerce merchants, available at: https://www.gorgias.com/about-us, last accessed on 15 December
2021 (ID: 2805).
121
rates to Kustomer of business customers of Kustomer’s close competitors are
magnitudes higher than those of other CRM software providers. In this regard, the
Commission considers that a foreclosure strategy targeted at Kustomer’s close
competitors would maximise the switching from the foreclosed CRM software
providers to Kustomer.
(410) In this regard, Zendesk (a close competitor of Kustomer) notes that “large businesses
typically need heavy customisation solutions, which creates a sort of lock-in or
disincentive to switching. Below that level, for mid-market and SMB business
customers, the CRM product is less customizable, and therefore for these segments
Kustomer has a real ability to influence switching. So for SMB and mid-market (i.e.
around 100-1000 employees – above 1000 there would be the need for more
customisation), Kustomer could scale rapidly and influence switching (without being
overwhelmed itself) as solutions targeted at these segments of the market are more
out of the box and need much less customization”.
632
Similarly, Crisp (a close
competitor of Kustomer) states that “businesses wishing to use CRM software
solutions, in particular small businesses, may have the incentive to switch to
Kustomer if Facebook provides a better integration of these channels with Kustomer
compared to third party CRM solutions”.
633
(411) Therefore, in light of the above conclusion that in general a significant proportion of
business customers, and in particular SMBs, are likely to switch to Kustomer
following non-targeted foreclosure, the Commission considers that a targeted
foreclosure strategy would encourage an even higher level of switching. As such, the
Commission expects that a sufficient proportion of business customers are likely to
switch to Kustomer so that losses from foreclosure would be limited (relative to the
gains). As noted in paragraph 389 and 391 above, it would require only a limited
proportion of the foreclosed business customers to switch to Kustomer for
foreclosure to be profitable.
(412) The Commission notes that this does not imply that a significant proportion or even
the majority of all business customers would not switch to non-foreclosed CRM
software providers other than Kustomer (e.g., SAP or Oracle) as a result of a targeted
foreclosure strategy. However, notwithstanding the conclusion of the previous
paragraph, the evidence on closeness of competition above suggests that the
foreclosed business customers of Kustomer’s close competitors would on balance be
more likely to switch to Kustomer than other CRM software competitors, such as
SAP or Oracle, precisely because (almost by definition) Kustomer’s offering is a
closer competitor/substitute for the offerings of these close competitors of Kustomer.
Moreover, even though such larger CRM players may have offerings for SMBs,
given their primary focus on larger and less price-sensitive enterprise clients with
highly complex needs (and which likely account for a material portion of their
revenues), it is far from certain that they would react in a timely manner to a
foreclosure strategy of the merged entity targeted at close competitors of Kustomer,
including start-ups and smaller customer service and support CRM players focused
on SMBs.
(413) In any case, Meta (formerly Facebook) would not lose any significant revenues if, as
a result of targeted foreclosure, business customers were to switch to CRM software
competitors that are not foreclosed. Therefore, to the extent that business customers
632
Agreed minutes of the call of 9 September 2021 with Zendesk, paragraph 14 (ID: 2085).
633
Agreed minutes of the call of 15 September 2021 with Crisp, paragraph 7 (ID: 2132).
122
were to switch to other non-foreclosed CRM software competitors, Meta (formerly
Facebook) would not experience any significant losses but would not get any gain
from the foreclosure either. Similarly, Meta (formerly Facebook) would not lose any
revenues if, as a result of targeted foreclosure, business customers were to choose to
multi-home with Kustomer and their foreclosed CRM software provider.
(414) Therefore, losses to Meta (formerly Facebook), as a result of a targeted foreclosure
strategy, would be limited to business customers that (i) decide to stay with the
foreclosed CRM software provider (which are close competitors with Kustomer), and
(ii) that also decide to limit their use of Meta (formerly Facebook)’s channels as a
result of the degradation to their CRM software provider’s API access.
(415) In its SO Response
634
and the paper submitted on 10 October 2021,
635
the Notifying
Party argues that Meta (formerly Facebook) would not have the incentive to execute
a targeted foreclosure strategy with respect to Kustomer’s close competitors. This is
essentially because, in this scenario: (i) the gains from foreclosing these close
competitors would still be minimal, because business customers would not switch
sufficiently to Kustomer, and (ii) the cost of foreclosing these close competitors
would still be substantial, again because business customers would not switch
sufficiently to Kustomer.
(416) This submission does not alter the Commission’s conclusions stated throughout
Section 7.3.2.2.(b) Contrary to what the Notifying Party argues, the Commission
considers that, by targeting the close competitors of Kustomer, the level of switching
to Kustomer will be sufficient to ensure that gains from foreclosure will be
significant and losses limited. The Commission arrives at this conclusion on the basis
of both the evidence outlined in Section 7.3.2.2.(b) and the evidence put forward by
the Parties. Indeed, the Notifying Party concludes business customers would not
switch sufficiently to Kustomer based on historical win/loss data. The Commission
considers that these data are not informative in this respect because there has not
been an example in the past where the owner of a leading B2C channel owns a CRM
(i.e., the pre-Transaction scenario is significantly different to the post-Transaction
scenario). As such, historical switching rates would not be indicative of switching
following a foreclosure strategy, i.e., in response to a deterioration of Meta (formerly
Facebook) access for some of Kustomer’s competitors but not for Kustomer. Indeed,
a foreclosure strategy in this case would inherently change the product of foreclosed
CRM software providers. As such, these historical switching rates can at best provide
a lower bound for the switching rates to Kustomer in response to such a
deterioration, with no indication that the actual switching rates would be close to that
lower bound. It is for this reason that, whilst in general historical switching rates are
helpful for assessing total market wide foreclosure strategies, there are shortcomings
for using them to assess partial or targeted foreclosure strategies.
(417) In addition, there are further shortcomings of these data. The diversion ratios are
based on Kustomer’s historical win data in 2020, which covers a sample of [Win-
Loss Data] wins. Although, Kustomer is a small player it has been growing
significantly over the past 3 years, which is when the win/loss data is from. This
would imply these historical data should be sufficiently discounted to account for the
fact that Kustomer’s business is rapidly growing through mainly brownfield
opportunities (i.e., business customers switching from competitors to Kustomer).
634
SO Response, Section 4.
635
Notifying Party Submission on Foreclosure Issues dated 10 October 2021.
123
This data does not include [Win-Loss Data]. In addition (and as noted by the Parties
themselves), these diversion ratios are not a good estimate for switching as a result of
targeted foreclosure (“if a foreclosure strategy targeted multiple CSCRM providers
at the same time, the resulting diversion ratios to Kustomer could be higher than
historic diversion ratios.”).
636
A targeted foreclosure strategy will include multiple
competitors of Kustomer. Historic switching rates would need to be adjusted to
disregard switching to other foreclosed CRM providers (as the Parties themselves
note
637
). This is likely to be significant because the Commission assesses a targeted
foreclosure strategy at Kustomer’s close competitors, who almost by definition will
have a significant amount of switching to each other.
(418) In this regard, the Commission considers that the evidence collected from business
customers directly on their willingness to switch following a foreclosure strategy, as
well as views from CRM software competitors, is more relevant for assessing the
level of switching following foreclosure (and therefore for assessing the incentives
for the merged entity to undertake a targeted foreclosure strategy).
(419) Notwithstanding the above, the Commission notes that, by accounting for some of
the shortcomings of the historical switching data provided by the Parties, the data in
fact indicates that switching to Kustomer will be sufficient to ensure that gains from
foreclosure will be significant and losses limited (despite these data constituting a
lower bound). In particular, accounting for the main shortcoming, namely that
historical switching data do not account for the fact that a business that switches
away from a foreclosed CRM provider in order to gain (superior) access to Meta
(formerly Facebook)’s B2C messaging channels would not switch to another
competitor of Kustomer if that player is also foreclosed), increases the diversion ratio
to the point where foreclosure would be profitable: an [10-20]:1 switching ratio
(which is less than the “critical switching ratio” of [40-50]:1 or [30-40]:1 estimated
by the Commission in paragraphs 389 and 391 above). The Commission notes that
even once correcting for these shortcomings, this historical switching data is still
unlikely to be a good indicator for switching post-foreclosure, mainly because it
includes switching for all kinds of reasons and not just switching to gain better
access to Meta (formerly Facebook)’s channels.
(420) First, the estimated diversion ratio to Kustomer from competitors increases
significantly when accounting for Kustomer’s growth as well as using a correct
sample (i.e., business customers who are likely targets of foreclosure). According to
Kustomer’s win data, Kustomer made […] wins from competitors in the first half of
2021, which would imply […] in total in 2021.
638
Assuming the same growth rate
between the number of wins in 2022 and 2021, Kustomer is expected to make […]
wins from competitors in 2022.
639
As stated by the Parties, Kustomer may be
considered an appropriate CRM by […] business customers.
640
This constitutes a
representative sample for the businesses that may be the targets of foreclosure. The
Parties assume an annual churn rate of […], which implies […] of these business
636
SO Response, Annex 3, paragraph 2.94.c.ii.
637
SO Response, Annex 3, paragraphs. 2.44 and 2.45.
638
Parties’ submission on Foreclosure Issues, dated 12 October 2021, Sheet “Win Data”. Note that
contrary to the Parties’ analysis, this includes wins from unknown competitors.
639
Idem. This is conservative as the growth rate is increasing over time ([…], […], and […] in 2019, 2020
and 2021 respectively).
640
SO Response, paragraph 4.9(B)(ii)(b). These […] businesses therefore constitute the universe of
business customers that would consider switching to Kustomer, and for this reason constitutes the
appropriate denominator.
124
customers that might consider Kustomer would have likely switched CRM software
provider.
641
Given that Kustomer is expected to make […] wins from competitors in
2022, […] or […] of business customers that might consider Kustomer are expected
to switch to Kustomer from competitors in 2022. This is already above the critical
switching ratio of [40-50]:1 or [30-40]:1 calculated by the Commission.
(421) Second, the estimated diversion ratio to Kustomer from competitors increases
significantly when accounting for the fact that business customers are unlikely to
switch to another foreclosed CRM provider if they are leaving their current CRM
software provider due to foreclosure. Proceeding from the calculations in the
paragraph above, […] of business customers are expected to switch to Kustomer
(based on first preferences), which means […] of business customers are expected to
switch to another CRM provider. However, “taking Kustomer’s data on which
CSCRM providers it lost new opportunities to (which is broadly indicative of, but not
the same as, diversion ratios): […] of opportunities were lost to [Kustomer’s 6 close
competitors: Zendesk, Intercom, Freshworks, Gorgias, Gladly and Crisp]”
642
(emphasis added). As stated in paragraphs 403-409 above and in paragraph 308 of
the SO, these six close competitors are likely to be the targets of foreclosure;
although this list is non-exhaustive. This figure implies that […] of these […]
business customers not switching to Kustomer would prefer to switch to another
foreclosed CRM competitor, i.e., […].
643
Conversely, […] of these […] business
customers not switching to Kustomer would prefer to switch to a non-foreclosed
CRM competitor, i.e., […]. Without further evidence, the Commission assumes that
the businesses wanting to switch to another foreclosed CRM provider would switch
to Kustomer and non-foreclosed CRM competitors in the same proportion as the
other businesses.
644
This results in an estimated diversion ratio to Kustomer from
close competitors of [Win-loss Data] or [Win-loss Data].
645
This is significantly
above the critical switching ratio of [40-50]:1 or [30-40]:1 calculated by the
Commission in paragraphs 389 and 391 above.
(422) In light of the evidence outlined above, the Commission considers it unlikely that
business customers would decide to stay with the foreclosed CRM software
providers that are close competitors of Kustomer (or switch to a foreclosed potential
entrant into the market who would be a close competitor of Kustomer). In this
regard, Meta (formerly Facebook) has the ability to limit its losses from foreclosure
by targeting at Kustomer’s close competitors.
C. Aggravating factors
(423) The Commission notes that there are potentially further aggravating factors that
would allow Meta (formerly Facebook) to further encourage switching to Kustomer
and thereby limit to a greater extent the losses from foreclosure (increasing the
incentive for Meta (formerly Facebook) to foreclose). It is worth noting that the
Commission does not view these aggravating strategies as inherently anti-
641
SO Response, Annex 3 Figure 1.
642
SO Response, Annex 3, paragraph 2.106.
643
In the absence of any further evidence, the Commission assumes that diversion ratios are symmetric
between Kustomer and its close competitors.
644
This is conservative given that Kustomer is a close competitor of the targets of foreclosure, and
therefore switching to Kustomer is more likely from these CRM competitors than other non-foreclosed
CRM competitors.
645
In other words, disregarding the […] who switch to other foreclosed CRMs, […] business customers
out of the […] business customers not switching to a foreclosed CRM will go to Kustomer, i.e., […].
125
competitive per se, but that in parallel with a foreclosure strategy they would
exacerbate the incentive to foreclose, by reducing the associated losses. The
Commission considers that Meta (formerly Facebook) could do so in two main ways:
(i) offering Kustomer for free or below market price, and (ii) bundling Kustomer
with Meta (formerly Facebook)’s ecosystem of products (see Section 7.3.2.2.(b.i)D
above regarding Meta (formerly Facebook)’s ecosystem of products).
(424) First, Meta (formerly Facebook) may offer Kustomer for free or below market price,
in order to encourage switching to Kustomer. Meta (formerly Facebook) has a
history of offering its products and services for free, and monetising them through
online advertising. In addition, [Business Plans].
646
Indeed, Zendesk (a close
competitor of Kustomer) notes that “price of the software is an important factor, if
not one of the most important factors, for business customers in the lower end of the
market, more so than the features offered by the CRM tool. For many businesses,
customer support is not a revenue centre, but more a cost centre. If it is possible to
reduce customer support costs, many businesses, especially SMBs and mid-sized
businesses, would be inclined to use a cheaper or even a free CRM”.
647
And, in the
context of switching by business customers between CRM software providers, Crisp
states that “[business customers] may switch to Kustomer if Facebook offers
Kustomer for free or based on a freemium model (e.g. by monetising it through
advertising)”.
648
(425) In parallel to a foreclosure strategy, a strategy of offering Kustomer for free or below
market price would further encourage switching to Kustomer from foreclosed CRM
providers (relative to the situation without such an aggravating strategy). This would
therefore reduce the number of business customers that decide to stay with the
foreclosed CRM software provider (which are close competitors of Kustomer), and
thereby limit further the losses from foreclosure. In particular, offering Kustomer for
free may encourage business customers to multi-home in the short-term with both
Kustomer and their foreclosed CRM software provider, which as outlined above
would not result in any significant losses for Meta (formerly Facebook).
(426) Second, Meta (formerly Facebook) may offer Kustomer as part of a bundle with its
other products, also to encourage switching to Kustomer (see Section 7.3.2.2.(b.i)D
above regarding Meta (formerly Facebook)’s ecosystem of products). Market
participants raised concerns that Meta (formerly Facebook) could leverage its
position in, for example, ecommerce in order to encourage business customers to
switch to Kustomer.
649
Strategies to bundle or leverage Meta (formerly Facebook)’s
position to encourage switching may even be more subtle. An example put forward
by Zendesk suggests that “Facebook would be able to identify businesses using the
Kustomer product and directly or indirectly promote their experiences as being
differentiated from others. For example, this could be done by adding a
“checkmark” or other icon to the business profile page on
WhatsApp/Messenger/Instagram that would appear only when using the full suite of
Facebook CX products.”.
650
646
Meta (formerly Facebook)’s internal document, Attachment E029 – […].
647
Agreed minutes of the call of 9 September 2021 with Zendesk, paragraph 7 (ID: 2085).
648
Agreed minutes of the call of 15 September 2021 with Crisp, paragraph 7 (ID: 2132).
649
See, for example, the agreed minutes of the call of 9 September 2021 with Zendesk, paragraph 12 (ID:
2085).
650
Agreed minutes of the call of 9 September 2021 with Zendesk, paragraph 3 (ID: 2085).
126
(427) In parallel to a foreclosure strategy, a strategy of bundling Kustomer with other Meta
(formerly Facebook) products would further encourage switching to Kustomer from
foreclosed CRM providers (relative to the situation without such an aggravating
strategy). Similarly to offering Kustomer for free, this would therefore reduce the
number of business customers that decide to stay with the foreclosed CRM software
provider (which are close competitors of Kustomer), and thereby limit further the
losses from foreclosure.
(428) Finally, it is worth noting that, whilst this Section has focussed on Meta (formerly
Facebook) limiting its losses by maximizing the switching to Kustomer, such a
foreclosure strategy (or any parallel aggravating strategies) would also increase the
gains to Meta (formerly Facebook). As discussed in Section 7.3.2.2.(b.i), Meta
(formerly Facebook) is able to make numerous, diverse and significant gains from
business customers that switch to Kustomer, as a result of foreclosure.
(b.iii) Conclusion
(429) On the basis of the assessment in Section 7.3.2.2.(b) above, the Commission
concludes that Meta (formerly Facebook) would have the incentive to engage in a
targeted input foreclosure of restricting or degrading API access of customer service
and support CRM software providers, both in the EEA and worldwide, to a large and
important part of the EEA-wide, if not worldwide, OTT B2C messaging market. This
conclusion also holds for the broader EEA-wide, if not worldwide, upstream market
for B2C communication services overall, given the importance of Meta (formerly
Facebook)’s channels for CRM software providers and their business customers, as
well as the complementarity of different channels in such a wider market, as well as
the broader EEA-wide, if not worldwide, downstream market for CRM software
overall, given the importance of customer service and support CRM software within
an overall CRM software market. As outlined in Section 7.3.2.2.(a.i) above, the
Commission’s conclusion would not change on the basis of the potential segments of
the customer service and support CRM software market, or the overall CRM
software market, based on (i) business customer size, (ii) mode of deployment or (iii)
the industry sector of the business customer because OTT B2C messaging channels
would remain an equally important input irrespective of such further segmentation.
This conclusion remains the same irrespective of whether the relevant upstream and
downstream markets, and potential segments thereof, are defined as EEA-wide or
worldwide in geographic scope.
(c) Impact on effective competition
(430) The Commission has reached the conclusion that it is likely, and there is even a
strong probability, that a targeted input foreclosure strategy would have significant
effects on competition by reducing the sales prospects of Kustomer’s close
competitors and deterring potential entrants.
(431) The effects of an input foreclosure strategy would, as a result of the restricted or
degraded API access and in light of the importance of Meta (formerly Facebook)’s
channels as an input into CRM software (see Section 7.3.2.2.(a)), reduce competition
in the CRM software market, and/or the customer service and support CRM software
market and possible segments thereof. This reduction in competition may result in
higher prices, lower quality and less innovation for business customers (in particular
SMBs), which may in turn be passed on to consumers. In addition, business
127
customers may be deprived of choice and no longer have the possibility to choose a
CRM provider without an ecosystem such as Meta (formerly Facebook).
651
(432) The Commission notes that even a partial targeted foreclosure strategy targeted at
small CRM providers would make it more difficult for these to compete, as Meta
(formerly Facebook)’s channels are considered an important input into CRM
software and foreclosed rivals would not be able to compete on an equal footing with
Kustomer which, combined with a potential loss of customers, could hinder
foreclosed rivals to innovate and effectively compete.
(433) The impact on effective competition would be immediate, in particular as regards
deterrence to new entrants in the CRM software market. As detailed in section
7.3.2.2.(a.iii): (i) Meta (formerly Facebook) has the (technical and contractual)
ability to restrict or degrade access to its API without much delay; and (ii) [Business
Plans].
652
(434) The impact may be particularly acute as (i) potentially foreclosed firms play a
sufficiently important role in the competitive process (Section 7.3.2.2.(c.i)); (ii) the
mere likelihood that the merged entity would carry out a foreclosure strategy post-
merger may already create a strong deterrent effect on potential entrants (Section
7.3.2.2.(c.ii)); there is currently no CRM software provider which is vertically
integrated (Section 7.3.2.2.(c.iii)); countervailing factors would not be sufficient to
maintain effective competition (Section 7.3.2.2.(c.iv)); and (v) claimed efficiencies
have not been substantiated (Section 7.3.2.2.(c.v)).
(c.i) Potentially foreclosed firms play a sufficiently important role in the competitive
process
(435) According to the Non-Horizontal Guidelines, significant harm to effective
competition normally requires that the foreclosed firms play a sufficiently important
role in the competitive process on the downstream market. The higher the proportion
of rivals which would be foreclosed on the downstream market, the more likely the
merger can be expected to result in a significant price increase in the downstream
market and, therefore, to significantly impede effective competition therein. Despite
a relatively small market share compared to other players, a specific firm may play a
significant competitive role compared to other players, for instance because it is a
close competitor of the vertically integrated firm or because it is a particularly
aggressive competitor.
653
(436) The Commission considers that new-entrants, start-ups and small CRM providers
with a focus on SMB customers are most likely the targets of a total or partial
foreclosure strategy by Meta (formerly Facebook) and these play a sufficiently
important role in the competitive process on the CRM market, as they are particular
drivers of innovation.
651
As set out in Section 7.3.2.2.(b.i.)D, Meta (formerly Facebook) may steer businesses into its ecosystem
of products with a view to cross-selling different products (e.g., Facebook Marketplace, Facebook
Shops and Workplace).
652
Parties’ reply to RFI 21, question 5(i).
653
See Non-Horizontal Guidelines, paragraph 48.
128
A. Small CRM providers focus on SMB customers (while large CRM providers focus on large
enterprises)
(437) The market investigation provided evidence to support that, by and large, smaller
CRM providers target SMB customers whereas many of the large CRM providers
focus on enterprise customers.
(438) First, this finding is supported by evidence from CRM providers’ responses to the
market investigation, which generally confirmed the focus on enterprise customers
by large CRM providers.
654
To this regard, SAP indicated it estimates 85% of its
clients are large business customers and confirmed that SMBs are “not a relevant
market for SAP”.
655
Moreover, Microsoft confirmed that it “primarily targets larger
business customers for D365 sales” (product viewed as targeting large enterprises, as
explained below).
656
By contrast, Crisp, a small CRM player, indicated around 60%
of its customers to be small business customers, 40% medium-sized business
customers and 1% large business customers.
657
(439) Second, the replies of large CRM players to the Commission’s RFIs provide
information that corroborates such distinction (between large and small CRM players
focusing on large enterprises and SMB customers respectively). Indeed, Oracle
explained in response to an RFI dated 16 November 2021 that it focuses on
enterprise customers and that this is reflected in its development efforts: “Oracle
focuses on enterprise size customers so development efforts is focused towards those
segments.”
658
(440) In fact, large CRM providers would need to change their business strategy to further
focus on SMB customers. In this regard, in response to RFIs dated 16 November
2021, large CRM providers stated the following:
• “If Microsoft were to adopt a strategy of more proactively targeting SMBs, it
would primarily increase focus in its marketing and sales channels to target
that specific segment. It might also create offerings with less customizability
and extensibility. As a general matter, enterprise-sized customers have
complex and varying needs, and may use a multitude of IT solutions that either
target CRM use cases or are adjacent to them. It is therefore important to
enterprises to have CRM solutions that can work with as many other different
IT offerings as possibility. Customizability and extensibility are therefore key –
hence the modularity, openness, and extensibility of D365. Smaller companies
with less complex IT systems may not require meaningful customizability and
extensibility. They may also not have the time or resources available to
implement complex customizations and extensions. They may therefore merely
require a simpler end-to-end solution that works out of the box without
requiring meaningful effort to set up and customize.”
659
654
Replies to questionnaire 7 to CRM providers, question 6.
655
Reply to questionnaire 7 to CRM providers, question 6.1 and 6.3 (ID: 1646).
656
Reply to questionnaire 7 to CRM providers, question 6.1 (ID: 1661).
657
Reply to questionnaire 7 to CRM providers, questions 6.1.-6.3 (ID: 2125).
658
Oracle’s reply to RFI dated 16 November 2021 (ID: 2315).
659
Microsoft’s reply to RFI dated 16 November 2021 (ID: 2301).
129
• “Oracle’s main customers are large business enterprises so we cannot
comment on specific needs of SMBs. In our experience, interests completely
depend on the customer’s use case as opposed to the customer size”.
660
• “To further target SMB customers, Salesforce must continue to innovate and
drive demand for its products. While the interests of SMB customers are
largely the same as enterprise customers, some SMB customers are more price
sensitive and do not demand the same breadth of features as enterprise
customers”.
661
• “A key aspect would be pricing and sizing - it is known than SMB customers
are quite price sensitive. [sic.] In essence, SAP would need to do three things:
1) create suitable price models that are attractive to SMB customers with
smaller infrastructure needs, 2) align features and functions with the
requirements of SMB customers, usually less complex than enterprise
customers and 3) resolve issues of processes and integration/implementation
that are/may be unique to the SMB market”.
662
(441) Third, the information submitted by the Notifying Party also confirms this
distinction: [Information about focus of CRMs].
663
(442) Finally, this distinction (between large and small CRM players focusing on large
enterprises and SMB customers respectively) is generally confirmed by the results of
the annual CRM awards by CRM Magazine. As regards large CRM providers, both
the 2020
664
and the 2021
665
CRM awards for best enterprise CRMs mention
Microsoft, Oracle, Salesforce and SAP in the top five best enterprise CRMs. The
CRM awards for best CRM for SMB customers mention however smaller CRM
providers alongside larger providers in 2020
666
(Creatio, HubSpot, Microsoft,
Salesforce and Zoho) and 2021
667
(Freshworks, HubSpot, Salesforce, SugarCRM and
Zoho).
(443) In light of the above, the Commission concludes that small CRM providers target
SMB customers whereas large CRM providers focus on enterprise customers.
668
660
Oracle’s reply to RFI dated 16 November 2021 (ID: 2315).
661
Salesforce’s reply to RFI dated 16 November 2021 (ID: 2304).
662
SAP’s reply to RFI dated 16 November 2021 (ID: 2310).
663
Notifying Party Submission on Foreclosure Issues dated 10 October 2021, page 28.
664
CRM Magazine, The best enterprise CRM software and solutions: the 2020 CRM industry leader
awards, 31 august 2021, available at: https://www.destinationcrm.com/Articles/Editorial/Magazine-
Features/The-Best-Enterprise-CRM-Software-and-Solutions-The-2020-CRM-Industry-Leader-Awards-
142561.aspx, last accessed on 29 November 2021 (ID: 2461).
665
CRM Magazine, The best enterprise CRM software and solutions: the 2021 CRM industry leader
awards, 26 August 2021, available at: https://www.destinationcrm.com/Articles/Editorial/Magazine-
Features/The-Best-Enterprise-CRM-Software-and-Solutions-The-2021-CRM-Industry-Leader-Awards-
148586.aspx, last accessed on 29 November 2021 (ID: 2453).
666
CRM Magazine, The best CRM for Midsize and Small Businesses: the 2020 CRM industry leader
awards, 31 August 2021, available at: https://www.destinationcrm.com/Articles/Editorial/Magazine-
Features/The-Best-CRM-for-Midsize-and-Small-Businesses-The-2020-CRM-Industry-Leader-Awards-
142562.aspx, last accessed on 29 November 2021 (ID: 2452).
667
CRM Magazine, The best CRM for Midsize and Small Businesses: the 2021 CRM industry leader
awards, 26 August 2021, available at: https://www.destinationcrm.com/Articles/Editorial/Magazine-
Features/The-Best-CRM-for-Midsize-and-Small-Businesses-The-2021-CRM-Industry-Leader-Awards-
148587.aspx, last accessed on 29 November 2021 (ID: 2463).
668
The Notifying Party argues that larger CRM players also serve SMB customers, which the Commission
does not dispute or ignore in its assessment. However, given all the evidence presented in this Section,
130
Despite the Notifying Party’s views that, in a scenario of targeted foreclosure of
small CRM providers, large CRM providers would remain non-foreclosed and exert
competitive pressure on Kustomer
669
, all the aforementioned elements indicate that
this would not be the case as large CRM players do not focus on SMB customers
(and would therefore not innovate for SMB customers, unlike small CRM players as
it will be demonstrated below).
B. Small CRM players that target SMB customers play a disproportionally important role in
the competitive process (as these are particular drivers of innovation)
(444) The CRM market is highly dynamic and is constantly evolving as a result of ongoing
innovation.
(445) In this regard, both the information submitted by the Notifying Party and the market
investigation support that innovation is an important parameter of competition in the
CRM market. The Notifying Party submits that CRM players “are actively
innovating” when arguing the CRM market is highly competitive
670
, and that
“companies continue to focus on the next level of innovation and the first to identify
the next “big thing” in Customer Service CRM, which is often quickly emulated”.
671
Moreover, a large portion of business customers that expressed a view considered the
benefit from the current level of competition in the CRM market was “High
innovation (e.g. more new features, further AI capabilities)” – more than for any
other response like “low prices” or “high quality” (in a multiple choice question).
672
One participant stated that the availability of innovative functionalities is one of the
most important considerations in its choice of CRM (even more than price): “Price is
an important aspect to consider when choosing a support CRM provider, but e.g.
innovative functionality and the seamless integration into the existing IT structures
are more important.”
673
(446) [Information about innovation], which was highlighted by Kustomer in one of the
site visit meetings.
674
Indeed this has been acknowledged by Meta (formerly
Facebook) employees assessing the Transaction in internal documents, where they
discuss the reaction to the Transaction by, amongst others, [Information about
innovation].
675
(447) The innovative role of small CRM players with a focus on SMB customers has also
been observed by industry reports, such as Gartner’s “Magic Quadrant”, which stated
that “Startups and new entrants from adjacent markets are typically found in the
Visionaries quadrant … Visionaries are ahead of many competitors in terms of
delivering innovative products and delivery models”
676
(emphasis added). Gartner
the Commission concludes larger CRM players focus on large enterprise customers and small CRM
players target SMB customers.
669
SO Response, paragraph 5.28.
670
Form CO, paragraph 6.70.
671
PowerPoint presentation session 2 “Facebook’s proposed acquisition of Kustomer - Kustomer and the
CRM Market” of the site visit held on 22 September 2021, slide 17.
672
Replies to questionnaire 5 to business customers, question 17.
673
Reply to questionnaire 5 to business customers, question 15.1 (ID: 1366).
674
This was explained by [Kustomer executive], during the site visit held on 22 September 2021
(session 2, “Kustomer and the CRM market”): [Kustomer executive opinion on innovation within the
CRM software market]– notes taken by the case team.
675
Meta (formerly Facebook)’s internal document, […].
676
Gartner’s “Magic Quadrant”, https://www.gartner.com/doc/reprints?id=1-
1Z70UGZF&ct=200608&st=sb, last accessed on 29 November 2021 (ID: 2516); see also Form CO,
footnote 88.
131
considered ServiceNow and Freshworks as Visionaries in 2019 and 2020
respectively.
677
Moreover, Gartner publishes a yearly list of “cool vendors” in CRM
customer service and support. Companies are selected for the “cool vendors” list
because Gartner considers them to be particularly innovative, impactful and
intriguing companies. The Gartner “cool vendors” list has previously included
Gladly and Helpshift (both small CRM providers), as well as several technologies
that have since been acquired by CRM providers.
678
As highlighted by one market
participant: “disruptive technological approaches are typically adopted by
newcomers to the industry whose approach and agility are difficult to replicate by
incumbents. A classic example here is Salesforce’s approach of delivering cloud-
based software as a service at a time [i.e., when it first entered the market in
1999
679
] when on-premise[s] software was the norm. Industry reports do not tend to
highlight the innovative nature of smaller players in the CRM market because it is
accepted that smaller companies must innovate in order to stay competitive in a
market historically dominated by large businesses. Gartner Research collects a list
of vendors they consider "cool" due to their "innovative ways of supporting better
customer service and operational outcomes" who are all smaller vendors, many of
whom focus on SMB customers”
680681
(emphasis added).
(448) Based on the market investigation and the assessment of the merged entities’
incentives set out in Section 7.3.2.2.(b), there is a particular concern that an input
foreclosure strategy would most likely target close competitors to Kustomer or new
entrants which are particularly innovative.
(449) One market participant stated that “there would likely be a negative impact on start-
ups, maverick players and new entrants at the lower end of the CRM market that are
pushing the envelope at present. There is a lot of innovation at this end of the
market.”
682
(emphasis added). This directly contradicts the Notifying Party’s
argument that the market investigation indicates that the CRM market is expected to
remain highly competitive post-Transaction.
683
677
Gartner’s “Magic Quadrant”, https://www.gartner.com/doc/reprints?id=1-
1Z70UGZF&ct=200608&st=sb, last accessed on 29 November 2021 (ID: 2516); see also Form CO,
footnote 88
678
An overview of cool vendors identified from 2012-2019 is included in Gartner, The Gartner Customer
Service Technology Vendor Guide, 2019, available at: https://productionkeywords.s3-us-west-
2.amazonaws.com/uploads/tokdm6bu/Gartner%20Customer%20Support.pdf, last accessed on 29
November 2021 (ID: 2460)
679
See https://freshservice.com/general/history-saas-blog/, last accessed on 15 December 2021 (ID: 2812).
680
Zendesk’s reply to RFI dated 12 October 2021 (ID: 2143); “The Cool Vendors 2021: Building Trust
With Digital Innovation, Competence and Ethics report” is generated for subscribers, but can be
generally found here: https://www.gartner.com/en/doc/749379-cool-vendors-must-build-trust-with-
digital-innovation, last accessed on 15 December 2021.
681
It should be noted that the Notifying Party argues that even though they are key innovators, larger CRM
players are necessarily excluded from Gartner’s “Cool Vendors” list which requires the vendor to be a
“small company” and from the “Visionary” category in Gartner’s “Magic Quadrant” report, which
requires the company to be one whose products and market presence are not yet complete or
established. The Commission considers the fact that these only concern smaller CRM players or new
entrants demonstrates that the innovation by these is very important and is particularly valued in the
market.
682
Agreed minutes of the call of 9 September 2021 with Zendesk, paragraph 15 (ID: 2085).
683
Moreover, the Commission considers the Notifying Party at times misrepresents the results of the
market investigation. For example, when drawing conclusions from the replies to questionnaire 7,
question 63.2 (about the impact on competitors if Kustomer was offered with superior integration), the
Notifying Party counts respondents that abstained from answering the cited question (Microsoft,
132
(450) As explained in Section 7.2.1. above, the CRM software market is highly fragmented
and split between a long tail of smaller competitors. Small players (like Zendesk,
Freshdesk or Gladly) represent up to 50% of the CRM software market (or the
customer service and support CRM market).
684
(451) The Commission notes that the Notifying Party relies on Microsoft/LinkedIn to
suggest that a foreclosure of less than 30% of the CRM software market was
considered by the Commission not to “affect a sufficiently important proportion of
Microsoft’s competitors to result in a price increase or a reduction of market
incentives to innovate.”
685
However, Microsoft/LinkedIn related to the potential
foreclosure of LinkedIn data for rivals of Microsoft. Such data would only, if at all,
be relevant for two sub-segments (CRM B2B marketing and B2B sales) that together
accounted for less than 30% of the entire CRM software market.
686
In the present
case, foreclosure could harm the entire market for customer service and support
CRM software (which accounts for 36% of the overall CRM market in 2020) due to
the disproportionately important role of small CRM providers in driving innovation
on the market. Moreover, in the referred decision, the Commission noted there were
many other possible sources of data
687
and considered that “there is uncertainty as to
whether in the near future LinkedIn full data would effectively become an important
input”.
688
In the present case however, the Commission concludes that Meta
(formerly Facebook)’s B2C messaging channels are an important input for the
market for customer service and support CRM software and that other B2C
messaging channels are largely complementary. Therefore, the Commission
considers the Notifying Party is misrepresenting the conclusions of such decision,
which are clearly not applicable to the present case.
(452) In addition, these small players have a disproportionate impact on competition
(greater than may be expected on the basis alone of their combined market share).
(453) Indeed, the long tail of smaller competitors is illustrated on the website
Beststartup.eu that published 2021 lists of CRM start-ups and companies in several
EEA countries which stand out because of their innovative character:
• 101 top CRM startups and companies in France (2021)
689
;
• 67 top CRM startups and companies in Ireland (2021)
690
;
• 29 top CRM startups and companies in Austria (2021)
691
;
Pegasystems, Salesforce, Cisco) as replying that the Transaction would have minimal or no impact on
competitors. Moreover, none of the respondents that mentioned the alleged efficiencies resulting from
the Transaction (cited by the Notifying Party) are CRM players except for Salesforce which is a large
CRM provider and therefore unlikely to be targeted by the foreclosure strategy.
684
The Notifying Party submits Zendesk, Intercom, Freshworks, Gorgias, Gladly and Crisp represent
only 6% of the market. However, as expressly stated above in Section 7.3.2.2.(b), those are examples of
close competitors and not an exhaustive list (as evidenced by Section 7.2.1.).
685
The Notifying Party cites Commission decision of 6 December 2016 in case M.8124 –
Microsoft/LinkedIn, paragraph 275.
686
Commission decision of 6 December 2016 in case M.8124 – Microsoft/LinkedIn, paragraphs 275
and 261.
687
Commission decision of 6 December 2016 in case M.8124 – Microsoft/LinkedIn, paragraph 262-263.
688
Commission decision of 6 December 2016 in case M.8124 – Microsoft/LinkedIn, paragraph 250.
689
Available at: https://beststartup.eu/101-top-crm-startups-and-companies-in-france-2021/, last accessed
on 29 November 2021 (ID: 2467).
690
Available at: https://beststartup.eu/67-top-crm-startups-and-companies-in-ireland-2021/, last accessed
on 29 November 2021 (ID: 2466).
133
• 53 top CRM startups and companies in Sweden (2021)
692
;
• 14 top CRM startups and companies in Estonia (2021)
693
;
• 51 top CRM startups and companies in Denmark (2021)
694
;
• 42 top CRM startups and companies in Finland (2021)
695
;
• 55 top CRM startups and companies in Belgium (2021)
696
;
• 65 top CRM startups and companies in Spain (2021)
697
;
• 101 top CRM startups and companies in the Netherlands (2021)
698
• 37 top CRM startups and companies in Portugal (2021)
699
;
• 57 top CRM startups and companies in Italy (2021)
700
;
• 10 top CRM startups and companies in Croatia (2021)
701
.
(454) Even if the Notifying Party submits that less than 20% of the mentioned companies
are CRM providers (and the rest no longer exist or are not a CRM - they are either
unrelated businesses or out-of-market actors that may integrate with a CRM)
702
, the
Commission considers these lists are still a significant indicator of the innovative
character of the many companies (132 companies) that are indeed small CRM
players.
(455) The disproportionate impact of small CRM providers on the market for customer
service and support CRM software is also supported by the growing importance of
the SMB segment in this market, which is confirmed by CRM Magazine: “Though
enterprise-scale companies still make up the bulk of CRM system users, demand is
growing fast from small and midsize businesses, analyst firms have noticed. In fact,
the SMB sector accounted for about 44 percent of the total worldwide CRM market,
which has been valued at $47.6 billion this year by Grand View Research. Industry
estimates also suggest that around 91 percent of all organizations with more than 10
employees currently use at least some very basic CRM systems, creating a huge
691
Available at: https://beststartup.eu/29-top-crm-startups-and-companies-in-austria-2021/, last accessed
on 29 November 2021 (ID: 2472).
692
Available at: https://beststartup.eu/53-top-crm-startups-and-companies-in-sweden-2021/, last accessed
on 29 November 2021 (ID: 2449)
693
Available at: https://beststartup.eu/14-top-crm-startups-and-companies-in-estonia-2021/, last accessed
on 29 November 2021 (ID: 2448).
694
Available at: https://beststartup.eu/51-top-crm-startups-and-companies-in-denmark-2021/, last accessed
on 29 November 2021 (ID: 2469).
695
Available at: https://beststartup.eu/42-top-crm-startups-and-companies-in-finland-2021/, last accessed
on 29 November 2021 (ID: 2471).
696
Available at: https://beststartup.eu/55-top-crm-startups-and-companies-in-belgium-2021/, last accessed
on 29 November 2021 (ID: 2470).
697
Available at: https://beststartup.eu/65-top-crm-startups-and-companies-in-spain-2021/, last accessed on
29 November 2021 (ID: 2450).
698
Available at: https://beststartup.eu/101-top-crm-startups-and-companies-in-the-netherlands-2021/, last
accessed on 29 November 2021 (ID: 2465).
699
Available at: https://beststartup.eu/37-top-crm-startups-and-companies-in-portugal-2021/, last accessed
on 29 November 2021 (ID: 2473).
700
Available at: https://beststartup.eu/57-top-crm-startups-and-companies-in-italy-2021/, last accessed on
29 November 2021 (ID: 2468).
701
Available at: https://beststartup.eu/10-top-crm-startups-and-companies-in-croatia-2021/, last accessed
on 29 November 2021 (ID: 2447).
702
LoF Response, paragraph 4.5(D)(ii).
134
opportunity for industry expansion.”
703
In this regard, the Notifying Party submits
that the growth of the SMB segment (in the CRM software market) has no bearing on
the impact of smaller CRM players on innovation. Nevertheless, the Commission
considers that if the demand from SMB customers is rising and small CRM players
are the ones focusing on serving and innovating for SMB customers, this constitutes
one further element to support that small CRM players have a growing importance in
the CRM software market. In particular, the foreclosure of these small CRM players
would mean the growing number of SMB customers looking for a CRM provider
would not benefit from the innovation that small CRM players create which is
focused on that segment of the market.
(456) Indeed, these small CRM providers are able to develop new innovative features that
the larger players cannot with their legacy systems and products. One market
participant explained that “bigger companies such as Salesforce could technically
innovate, but they have constraints given their size, such as a huge number of legacy
customers who are used to how their product works, etc”.
704
Moreover, these
innovations may have an important impact on competition in the (highly innovative
and growing
705
) CRM software market both today and in the future.
(457) A variety of internal documents from Kustomer [Information about innovation].
706
For instance, a Recommendations Report of Kustomer from April 2020 states,
[Information about innovation]
707
(emphasis added). The Report further reinforces
this by noting, [Information about innovation]
708
(emphasis added) and [Information
about innovation]
709
(emphasis added).
(458) In addition, Kustomer refers to this distinction in a document named [Information
about innovation]
710
. In a document named [Information about innovation],
703
CRM Magazine, The best CRM for Midsize and Small Businesses: the 2021 CRM industry leader
awards, 26 August 2021, available at: https://www.destinationcrm.com/Articles/Editorial/Magazine-
Features/The-Best-CRM-for-Midsize-and-Small-Businesses-The-2021-CRM-Industry-Leader-Awards-
148587.aspx, last accessed on 29 November 2021 (ID: 2463).
704
Agreed minutes of the call of 9 September 2021 with Zendesk, paragraph 15 (ID: 2085).
705
PowerPoint presentation session 2 […].
706
In its LoF Response, the Notifying Party adds that the internal documents cited by the Commission
which describe larger CRM software providers as “older legacy players” and not innovative are, with
the exception of one e-mail, all sales and marketing related documents which are by definition designed
to portray Kustomer in a positive light compared to other customer service and support CRM providers.
The Commission notes that these documents do not merely describe [Information from internal
documents]. Moreover, even if these are [Information from internal documents], Kustomer could be
described and promoted therein in a positive manner by highlighting other advantageous aspects other
than its innovative character by contrast with legacy CRM players. The fact that Kustomer chooses to
distinguish itself focusing primarily on the innovation aspect (as a key differentiator) is indicative that
innovation is one of the most important features that distinguishes Kustomer from larger CRM players
(as is further supported by all the other evidenced referred in Section 7.3.2.2.(c.i)). In fact, these internal
documents further demonstrate that, in order to grow in the CRM market, a small player such as
Kustomer must innovate (hence the reason why [Information from internal documents]. The
Commission therefore considers these internal documents are important evidence to indicate that there
is a distinction between new innovative smaller players and larger companies in the CRM market.
707
Kustomer’s internal document, […].
708
Kustomer’s internal document, […].
709
Kustomer’s internal document, […].
710
Kustomer’s internal document, […].
135
Kustomer states: [Information about innovation]
711
. Finally, in a presentation named
[Information about innovation], Kustomer states: [Information about innovation]
712
.
(459) Moreover, Kustomer refers [Information about innovation]
713714
.
715
(460) In fact, the very reason why Kustomer was founded was due to [Information about
innovation]
716
. For example, [Information about innovation]
717
, or in an email to a
potential customer, Kustomer states [Information about innovation]
718
(emphasis
added).
(461) Other CRM providers such as Zendesk confirm the distinction between new
innovative players and legacy players. For example, a marketing text promoting its
CRM platform “Zendesk Sunshine” states “To make matters worse, legacy CRM
platforms force us into their proprietary technology. … It’s time break free from
CRM systems that lock businesses into an outdated view of customers. We need a
CRM platform that supports a new reality”.
719
In a document named “Zendesk VS.
Freshservice (example of how Zendesk may position against Freshdesk)”, Zendesk is
described as an “innovative leader in areas like Machine Learning”.
720
Gartner
Reports confirm that Zendesk can be seen as a particularly innovative player. For
example, in the 2020 “Gartner’s Magic Quadrant for the CRM Engagement Center
Report”, it describes Zendesk as a CRM with innovation as one of its strengths, and
notes, [Gartner’s estimates on the size and growth of CRM market segments].
721
Zendesk’s CEO has talked publicly about “Zendesk Sunshine” as disruptive, “truly a
next-generation headless CRM, that can have the potential to change the landscape
again”.
722
(462) Finally, there are several concrete examples of innovations introduced in recent years
by customer service support CRM software market providers with a focus on SMB
customers, as well as examples of material innovations anticipated in the coming
years by such players in the market.
(463) Kustomer itself is one CRM provider which is highly innovative (notably for
creating a ticketless solution) with mostly SMB customers (in particular, medium-
sized business customers).
723
SAP highlighted: “the technology behind Kustomer is
exceptional. Kustomer changed the CRM business model by creating a model based
on reaching a resolution over chatbots or over the phone without documenting it
(i.e., it is a more efficient solution). That is what the customer wants whereas the
711
Kustomer’s internal document, […].
712
Kustomer’s internal document, […].
713
Kustomer’s internal document, […].
714
Kustomer’s internal document, […].
715
Kustomer’s internal document, […].
716
Kustomer’s internal document, […].
717
Kustomer’s internal document, […].
718
Kustomer’s internal document, […].
719
Kustomer’s internal document, […].
720
Kustomer’s internal document, […].
721
Kustomer’s internal document, […].
722
https://diginomica.com/zendesk-ceo-disruptive-player-crm, last accessed on 29 November 2021
(ID: 2455).
723
Form CO, paragraphs 2.2.9 and 3.25; PowerPoint presentation session 2 “Facebook’s proposed
acquisition of Kustomer - Kustomer and the CRM Market” of the site visit held on 22 September 2021,
slide 4; agreed minutes of the call with Salesforce on 19 May 2021, paragraph 4 (ID: 2005); agreed
minutes of the call of 9 September 2021 with Zendesk, paragraph 8 (ID: 2085).
136
traditional approach, and what businesses wanted, involved a preference for
documenting the resolution of problems via tickets”.
724
(464) The Notifying Party submitted numerous examples of key innovations in the
customer service support CRM software market which were introduced in the
customer service and support CRM software market by small CRM players with a
focus on SMB, such as Zendesk and Freshworks:
725
• Zendesk Chatbots (in the area of Self-Service);
• Freshworks Chatbots (in the area of Self-Service);
• Kustomer KIQ Conversational Assistant (in the area of Self-Service);
• Helpshift Chatbot (in the area of Self-Service);
• Kustomer KIQ Language Detection (in the area of Routing & Workflows);
• Freshworks Freddy AI Actions (in the area of Agent Assistance;
• Helpshift Agent Desktop (in the area of Agent Assistance); and
• Freshworks Freddy Insights (in the area of “Business Intelligence”).
(465) In addition, Zendesk, a customer service and support CRM software provider with a
focus on SMB customers provided concrete evidence of an innovation introduced in
recent years to its customer service and support CRM software offering, i.e., “early
innovation in AI for customer service. Answer Bot - AI powered responses to
customer questions which suggest relevant help center articles that are likely to
answer the customer’s question”.
726
(466) The Notifying Party argues this innovation was introduced by large CRM players
first. However, the market investigation supports that there was indeed an innovative
aspect to “answer bot”.
(467) According to Zendesk, “Answer Bot product offerings have been innovative in that
businesses have not had to “train” a machine learning model to accurately predict
appropriate responses or resolutions to customer inquiries. Leading solutions that
predate Answer Bot (and frankly still most leading solutions today) require extensive
AI model development and training on a customer-by-customer basis, limiting these
solutions’ addressability to businesses with sufficient time, resources, and training
data. Additionally, Answer Bot introduced the concept of “zero configuration” bots.
When a customer enables Answer Bot on their account, it immediately starts
providing answers automatically in response to customer inquiries by using content
that already exists within the knowledge base of the business. With Answer Bot
capabilities, Zendesk’s products will offer predictions for customers without any
customization or training data required”.
727
Although the Notifying Party submits “it
724
Agreed minutes of the call of 18 May 2021 with SAP, paragraph 16 (ID: 2008).
725
PowerPoint presentation session 2 “Facebook’s proposed acquisition of Kustomer - Kustomer and the
CRM Market” of the site visit held on 22 September 2021, slide 17. It should be noted that this
presentation was submitted to the Commission prior to the SO where the Commission presented its
arguments regarding an input foreclosure targeted at Kustomer’s close competitors. In its LoF
Response, Annex 2, the Notifying Party then submitted that neither of these examples (which it had
pointed out itself as “key innovations” in the CRM market) were: (i) innovations; or (ii) by the small
CRM players it had previously named.
726
Zendesk’s reply to RFI dated 12 October 2021 (ID: 2143).
727
Zendesk’s reply to RFI dated 16 November 2021 (ID: 2311).
137
is debatable” whether Zendesk’s chatbot was “the first that was easier to set up”
728
in the terms just described, it should be noted that Zendesk was nominated for an AI
award in the category of “Best Intelligent Assistant Innovation” in the year it
introduced “answer bot” in the CRM market (2017).
729
(468) Zendesk anticipates innovating in the coming years, in the areas of “business
messaging experiences” (focused on the end-customer and in “building tools and
technologies that allow businesses using [our CRM tool] to deliver rich experiences
to their customers directly within the conversation”
730
) and “multi-party messaging”
(focused on bringing conversations happening outside of a CRM tool, into the CRM
tool).
731
Zendesk explained why it considers it is innovating in such areas:
“Zendesk’s business messaging is unique in the CRM market as it’s built on a
powerful CPaaS platform (Sunshine Conversations) that allows businesses open
access to all of the conversational data in real time. This enabled a number of
innovative features, many of which have gone on to inspire the competition: 1.
Channel Transfers - People interacting with a business can start a conversation on a
public messaging channel, such as Facebook Messenger, WeChat, or LINE and be
redirected to a secure and private messaging channel such as one found on the
company’s mobile app or website. When the redirection happens, all context about
the user and conversation is maintained and the user does not have to reauthenticate
or repeat previous parts of the conversation. The inverse is also possible: a
conversation that starts on a website can be transferred over to a public messaging
app (…). 2. Zendesk was granted two patents for allowing systems from multiple
vendors to share and participate in conversations with end users. These unique
mechanisms allow businesses using Zendesk to integrate any number of third party
systems through a documented standard, ensuring that a business has complete
freedom of choice over its technology partners. Zendesk’s own products integrate
with one another for the purposes of sharing conversational data using these same
technologies. (…) 3. Zendesk’s multi-party messaging allows businesses to connect
various participants in their business (ie, customers, employees, suppliers,
contractors) together in a single conversation, even if they aren’t all using Zendesk
software. For example, a food delivery company could connect customer support
agents using Zendesk, food delivery contractors using a mobile app and a customer
using any consumer messaging channel together in a conversation. Because they are
powered by the underlying CPaaS, these conversations can be archived and stored
within Zendesk, monitored for compliance and trust and safety issues, and used as
context when the customer requires further assistance. To our knowledge no other
vendor in the CRM space is able to provide this multi-party flexibility in terms of
conversation channels or level of integration.”
732
In its SO Response, the Notifying
Party confirmed “there will undoubtedly be further innovation in these areas”.
733
(469) Zendesk also anticipates future material innovations by other small CRM players that
also focus in particular on SMB customers, in the area of “automation and chatbots
728
LoF response, Annex 2.
729
Available at https://aibusiness.com/document.asp?doc_id=760435, last accessed on 29 November 2021
(ID: 2451).
730
The CRM software provider explained: “for example, an airline using [our CRM tool] could deliver a
visual seat selection tool to their customer during a flight booking, on any communication channel, so
that their customer can stay in the conversation and does not need to switch contexts” (ID: 2143).
731
Zendesk’s reply to RFI dated 12 October 2021 (ID: 2143).
732
Zendesk’s reply to RFI dated 16 November 2021 (ID: 2313).
733
SO Response, Annex 1, line 15.
138
for customer service” (leveraging machine learning to automatically deploy
conversational automation for a business using the history of past interactions).
734
(470) Another small customer service and support CRM software provider which focuses
in particular on SMB customers, Crisp,
735
also provided concrete evidence of
innovations introduced in recent years to its customer service and support CRM
software offering, which include “automated/realtime chat translations”, “co-
browsing”, “embedded audio/video calls”, and “integrations” with different
software applications (“Slack, Messenger, Whatsapp, Twitter, etc”) and with other
CRMs
736
(e.g., with “Hubspot, Zendesk, Salesforce”), “status page” (which allows
for monitoring a company’s infrastructure and lets users know when something is not
working without having to contact customer service).
737
(471) The Notifying Party argues all these innovations were introduced by large CRM
players first.
738
Nevertheless, the market investigation supports that there was an
innovative aspect to each of these examples.
(472) In what concerns “embedded audio/video calls”, Crisp clarified that it is an
innovation as “we are one if [sic.] the first companies providing embedded video
calls in a customer support chat widget, without any external third party plugin. The
main advantage is companies don’t need to install Zoom/Skype/Google Meets. The
video call solution works as is, in the browser, without requiring any third-party
software. A such feature already existed before as a niche market: Some customer
support players were offering chat widgets only focused on video/audio. Crisp was
the first company releasing audio-video calls as a core feature on a
SMB/generalistic customer support software”.
739
(473) As for “automated chat translation”, Crisp clarified it was the first CRM player to
make it available for SMB customers: “the goal of such feature is to let customers
write using their native language (for instance the customer says “bonjour” in
French). The customer support agent sees “hello” on his end. Then the customer
agent replies “how can I help?”. The software submits the message and auto-
translates the message to the end customer “comment je peux aider?”. Crisp was the
first company introducing this on a SMB solution, with a fair price (it is included in
our 95 euros / month plan). Some other companies started doing then same after us,
like Zendesk, Gosquared, Freshdesh.”
740
(474) In what regards “co-browsing” Crisp explained “the main innovation is it doesn’t
requires a third-party plugin. The end-customers doesn’t require to install anything
on their browser. For instance, Oracle Co-Browse solution require the user to install
734
Zendesk’s reply to RFI dated 12 October 2021 (ID: 2143).
735
Agreed minutes of the call of 15 September 2021 with Crisp, paragraph 2 (ID: 2132) “Crisp started
almost five years ago in France as a chat software for smaller companies to get in touch with their
customers. Most users are SMBs and start-ups, but bigger companies are starting to become
interested”.
736
According to the same market participant, “integrations for CRM are just like the App-Store for iPhone.
It expends functionalities. You are not forced to use the default browser, mail app provided by your OS.
So it is with CRM as well. You can use a different E-Mailing software, also connect with other CRM”.
Business customers use different software for different purposes (e.g., one software for e-mailing and
another for tasks) and “integrations” allow business customers to connect these different software using
their CRM software.
737
Crisp’s reply to RFI dated 12 October 2021 (ID: 2034).
738
SO Response, paragraph 5.17(A)(i).
739
Crisp’s reply to RFI dated 16 November 2021 (ID: 2446).
740
Crisp’s reply to RFI dated 16 November 2021 (ID: 2446).
139
a software on the computer. Crisp works in the browser directly. This way, when a
customer reports a problem, or requires help, you can directly see the problem, in a
few seconds.”
741
(475) Crisp further clarified that while “integrations” with different software applications
and with other CRMs “is not something specific to Crisp and as most of companies
on the market are doing this”, Crisp created an innovative feature within this area:
“we also have integrations with Team messaging software such as Slack and
Microsoft Teams. The main innovation is it lets company reply directly from
Slack/Microsoft Teams, so they do need to use the Crisp interface.”
742
(476) As for “status page”, Crisp explained it was the first company to introduce such
solution in customer service and support: “this feature allows companies to monitor
all their servers (a such system already existed before). This way, once a server goes
down, the companies is notified. Crisp is the first company doing this with in the
customer support level. In situations where an online service is affected, at the same
time, thousands of clients can reach the customer service, creating a crisis situation
for customer service agents, having to explain the situation to many different clients.
With the Crisp Status Page system, if something goes down, customers are notified
automatically in the chat widget, so they don’t have to reach the customer service to
know what is going on.”
743
(477) Crisp also provided further examples of innovations by other small CRM players
focused on SMB customers, stating it “took some good ideas from competitors, like
Campaigns Systems (Intercom), Ticketing Center in 2021 (Zendesk / Salesforce)”.
744
Crisp explained that “Campaigns systems” is a solution that “allows companies to
send emails to all their customers. For instance, send a newsletter when a new
product is released. The main innovation is it lets customers, using only one product
(Crisp), with a fair price. They had to use multiple expensive products previously to
achieve the same tasks.”
745
As for “Ticketing center”, it “allows companies to create
a ticker [sic.] center. It’s basically a website page they can embed on their
dashboard, allowing end users to create tickets.” Zendesk confirmed many of its
innovations are somehow related to its core workflow functionality, which is based
on a ticket (set of interactions between an organization and end-user customers):
“this is the primary workflow enabled by Zendesk customer service products and has
been a source of significant innovation around our ability to manage interactions
across channels, apply business rules and automation, and report on activity. A
substantial majority of the product development that Zendesk has been engaged in
since its founding in some way or another is related to the management of tickets as
our core data object.”
746
(478) Crisp predicted that “innovation in the future” would focus, amongst other things, on
“integrations” (e.g., with B2C messaging channels), as well as “AI and Bots.”
747
In
its SO Response, the Notifying Party confirmed “there will undoubtedly be further
innovation in these areas”.
748
741
Crisp’s reply to RFI dated 16 November 2021 (ID: 2446).
742
Crisp’s reply to RFI dated 16 November 2021 (ID: 2446).
743
Crisp’s reply to RFI dated 16 November 2021 (ID: 2446).
744
Crisp’s reply to RFI dated 12 October 2021 (ID: 2034).
745
Crisp’s reply to RFI dated 16 November 2021 (ID: 2446).
746
Zendesk’s reply to RFI dated 16 November 2021 (ID: 2311).
747
Crisp’s reply to RFI dated 12 October 2021 (ID: 2034).
748
SO Response, Annex 1, line 16.
140
(479) It should also be noted that Zendesk explained that its “answer bot” innovation has
been adopted by “all of the major support CRM providers”.
749
Crisp likewise echoed
the view that innovations in the past have frequently been copied or taken up by
CRM competitors. Just as Crisp took some good ideas from Intercom,
Zendesk/Salesforce (as mentioned above in relation to “Campaign Systems” and
“Ticketing Center”), many of Crisp’s innovations, including “Automated / realtime
chat translations”, “co-browsing” and integrations with different kinds of software
applications were copied by larger “competitors like Freshworks”. The same was
true of innovations introduced by many other SMB-focused players. For example,
“Drift offer is a messaging system … It was then copied by Hubspot”, “Intercom
came with an all-in-one approach …[and]… released some features in past years
like Embedded chat widgets (2018/2019), Onboarding system (2018). … Most of
their features are then copied Drift, Freshworks and other players.”
750
(480) In light of the above, it is reasonable to conclude that future innovations introduced
by such smaller CRM software providers focused in particular on SMB customers
([Business Plans]
751
) would to a large extent be copied or taken up by other players
in the market, including larger CRM software providers. As such, a targeted
foreclosure of new-entrants, start-ups and close competitors of Kustomer, i.e.,
smaller customer service and support CRM software providers that focus in
particular on SMB customers, would reduce competition in the customer service and
support market to an even greater extent than may be expected on the basis alone of
combined market share of the targets of foreclosure.
(481) The Commission therefore considers that the firms that could potentially be the
target of total or partial foreclosure strategies play a sufficiently important role in the
competitive process.
(482) The Notifying Party argues that large CRM players develop innovations for both
large and SMB customers
752
. In this regard, the Commission does not dispute that
larger CRM players also innovate in the market.
(483) Nevertheless, contrary to what the Notifying Party argues, the Commission considers
that smaller CRM providers play a specific innovative role in the market, in
particular for SMB customers, as evidenced by all the arguments put forward in the
present Section and in particular the several examples of innovations specifically
directed at SMB customers referenced above
753
, which can be differentiated from the
innovations by major CRM providers that are not specific for SMB customers.
(484) In addition, contrary to what the Notifying Party argues
754
, the Commission
considers that if these small players were foreclosed from the CRM market, larger
CRM players would not react by reinforcing their innovation efforts. Not only are
those larger CRM players less adaptable (and therefore do not have the same
incentive to bring innovation to this segment in the market), but they would also have
749
Zendesk’s reply to RFI dated 12 October 2021 (ID: 2143).
750
Crisp’s reply to RFI dated 12 October 2021 (ID: 2034).
751
Meta (formerly Facebook)’s internal document, […].
752
SO Response, paragraph 5.33.
753
The Notifying Party argues the Commission solely relies on the responses of two small CRM players.
In fact, in its market investigation, the Commission reached out to several other small CRM players
(which were not responsive to the Commission’s RFIs), as well as large CRM players, and it also took
into consideration in its assessment all the elements mentioned above including industry reports,
internal documents and public sources.
754
SO Response, paragraph 1.10.
141
less competitive pressure to do so once smaller (innovative) rivals have been
foreclosed and, therefore, the impact would be a significant lessening of innovation
in the CRM market (where innovation is a key characteristic).
(c.ii) The mere likelihood that the merged entity would carry out a foreclosure strategy
will likely already have anticompetitive effects
(485) According to the Non-Horizontal Guidelines, the mere likelihood that the merged
entity would carry out a foreclosure strategy post-merger may already create a strong
deterrent effect on potential entrants.
755
(486) There has been a large number of entrants in the CRM software market, and/or the
customer service and support CRM software market over the past years.
(487) However, potential new entrants in the market may be deterred if they consider it
likely that Meta (formerly Facebook) will restrict or degrade API access to its
channels since these are an important input into CRM software and such foreclosure
would have a direct impact on the quality of their product offerings and ability to
compete in the market. Even if these potential new entrants would try to innovate
through creating new additional features to differentiate their offerings (and compete
without having the full range of communication channels), they would be at a
disadvantage in relation to CRM software providers with access to Meta (formerly
Facebook)’s messaging channels (including Kustomer) and, given the importance of
this input (see Section 7.3.2.2.(a.i)), they would not be able to overcome this obstacle
and compete on the merits.
(488) The Commission therefore considers that the mere likelihood that the merged entity
would carry out foreclosure strategies would likely already create a strong deterrent
effect on potential entrants.
(c.iii) Competition from vertically integrated downstream competitors or alternative input
providers would not constrain the merged entity
(489) The Non-Horizontal Guidelines set out that if there remain sufficient credible
downstream competitors whose costs are not likely to be raised, for example because
they are themselves vertically integrated or they are capable of switching to adequate
alternative inputs, competition from those firms may constitute a sufficient constraint
on the merged entity and therefore prevent output prices from rising above pre-
merger levels.
756
(490) In this regard, it should be noted there is currently no CRM software provider which
is vertically integrated.
(491) As for the possibility of CRM software providers switching to adequate alternatives,
the Commission considers this would be unlikely.
(492) First, according to the information provided in market calls held by the Commission,
several communication channels are complementary (for instance email, phone and
SMS) and business customers in particular consider it important for CRM providers
to have an omnichannel offering.
757
755
See Non-Horizontal Guidelines, paragraph 49.
756
See Non-Horizontal Guidelines, paragraph 50.
757
Agreed minutes of the call of 19 May 2021 with Salesforce, paragraph 7 (ID: 2005) and agreed minutes
of the call of 18 May 2021 with SAP, paragraph 15 (ID: 2008).
142
(493) Second, as demonstrated in Section 7.3.2.2.(a.ii) above, there are no other channels
with high penetration rates similar to Meta (formerly Facebook)’s messaging
channels.
(494) The Commission therefore considers there would be no competition from vertically
integrated downstream competitors and CRM providers would not be capable of
switching to adequate alternative input providers so as to sufficiently constrain the
merged entity.
(c.iv) Countervailing factors would not be sufficient to maintain effective competition
(495) As set out in the Non-Horizontal Guidelines, the effect on competition on the
downstream market must also be assessed in light of countervailing factors such as
the presence of buyer power or the likelihood that entry upstream would maintain
effective competition.
758
(496) As regards countervailing buyer power, as mentioned above, a targeted foreclosure
strategy aimed at Kustomer’s close competitors and new entrants would impact small
niche innovative players in the customer service and support CRM software.
(497) Such players do not have the size, resources or bargaining strength to exert
competitive pressure on Meta (formerly Facebook), nor the ability to switch to other
B2C messaging channels within a reasonable timeframe as a reaction to Meta
(formerly Facebook)’s foreclosure strategy. For the same reasons, it is also unlikely
they could sponsor upstream entry for instance by persuading a potential entrant to
enter by committing to execute a substantial agreement with this company.
(498) As regards the likelihood that entry upstream would maintain effective competition,
it should be noted that Google RCS is only in an early stage of development of its
offering in the B2C messaging market and, according to Meta (formerly Facebook)’s
internal documents, [Business Plans]. On 3 February 2019, [Name] (part of the team
that “evaluated the Kustomer transaction from a business or strategic
perspective”
759
) sent an e-mail on the state of the B2P market indicating that:
[Business Plans]
760
The uncertainty regarding the future growth of alternative
channels upstream is corroborated by Google itself (regarding Google RCS) and by
Apple (regarding Apple Business Chat) as detailed in Section 7.3.2.2.(a.i).
(499) In addition, based on the information provided in market calls, although having a
large user base in P2P messaging can indeed facilitate entry in B2C messaging
761
,
achieving high P2P penetration in a short timeframe is unlikely.
(500) The Commission therefore considers that eventual countervailing factors would not
be sufficient to maintain effective competition.
(c.v) Claimed efficiencies have not been substantiated
(501) According to the Non-Horizontal Guidelines, the effects on competition need to be
assessed in light of efficiencies substantiated by the merging parties.
762
(502) The Notifying Party submits the Transaction will create the following efficiencies
and customer benefits: (i) increase take-up of messaging for B2C communications,
758
See Non-Horizontal Guidelines, paragraph 51.
759
Parties’ reply to RFI 7, question 3.
760
Meta (formerly Facebook)’s internal document, […].
761
Agreed minutes of the call of 20 September 2021 with Google, paragraph 6 (ID: 2127).
762
See Non-Horizontal Guidelines, paragraph 52.
143
(ii) drive innovation and competition in the B2C messaging space; (iii) drive
innovation and competition in the CRM market, and (iv) facilitate access to higher
quality ads services for businesses.
763
(503) In assessing any claims regarding efficiencies, the Commission applies the
Horizontal Merger Guidelines which establish a cumulative set of requirements to
take efficiencies into consideration.
764
(504) First, the "relevant benchmark" in the assessment of efficiency claims is that
consumers should be no worse off as a result of the merger. For that purpose,
efficiencies have to be substantial and timely, and should, in principle, benefit
consumers in those relevant markets where it is otherwise likely that competition
concerns would occur.
765
(505) Any efficiency should be passed on to consumers. The scope for pass-on is often
related to the existence of competitive pressure from the remaining firms in the
market and from potential entry. The greater the possible negative effects on
competition, the more the Commission has to be sure that the claimed efficiencies
are substantial, likely to be realized, and to be passed on, to a sufficient degree, to the
consumer.
766
(506) Second, efficiencies should be merger specific and it should not be possible for them
to be achieved to a similar extent by less anticompetitive alternatives.
767
(507) Finally, the efficiencies should be verifiable so that the Commission can be
reasonably certain that the efficiencies are likely to materialize, and be substantial
enough to counteract a merger's potential harm to consumers.
768
(508) In the Form CO, the Notifying Parties have submitted very brief and generalised
claims regarding cost savings (for businesses that service consumers through
messaging) and how the Transaction will drive innovation and competition in the
relevant markets.
769
(509) It is incumbent upon the Notifying Party to provide the Commission in due time with
all the relevant information necessary to demonstrate that the claimed efficiencies are
merger-specific and likely to be realised and that the efficiencies are likely to
counteract any adverse effects on competition that might otherwise result from the
merger, and that the claimed efficiencies therefore benefit consumers.
770
(510) The Commission therefore considers the Notifying Party has not shown that any
efficiencies are verifiable, merger specific and would be passed on to consumers.
(c.vi) Conclusion
(511) On the basis of the assessment in Section 7.3.2.2.(c) above, the Commission
concludes that, were the merged entity to pursue a targeted input foreclosure strategy
of restricting or degrading API access, it is likely, and there is even a strong
probability that the Transaction would have significant negative effects on
763
Form CO, paragraph 9.1.
764
See Non-Horizontal Guidelines, paragraph 53; Horizontal Merger Guidelines, paragraph 78.
765
See Horizontal Merger Guidelines, paragraph 79.
766
See Horizontal Merger Guidelines, paragraph 84.
767
See Horizontal Merger Guidelines, paragraph 85.
768
See Horizontal Merger Guidelines, paragraph 86.
769
Form CO, Section 9.
770
See Horizontal Merger Guidelines, paragraph 87.
144
competition in the EEA-wide, if not worldwide, market for customer service and
support CRM software. This conclusion also holds for the broader downstream
market for CRM software, given the importance of customer service and support
CRM software within an overall market for CRM software. As outlined in
Section 7.3.2.2.(a.i) above, the Commission’s conclusion would not change if the
customer service and support CRM software market, or the overall CRM software
market, were to be segmented based on (i) business customer size, (ii) mode of
deployment or (iii) the industry sector of the business customer, because OTT B2C
messaging channels would remain an equally important input irrespective of such
further segmentation.
7.3.2.3. Conclusion
(512) Based on the assessment in Section 7.3.2.2, and in light of the results of the market
investigation and of all the evidence available to it, the Commission has reached the
conclusion that it is likely, and there is even a strong probability, that the Transaction
would significantly impede effective competition as a result of vertical non-
coordinated effects arising from vertical links between the EEA-wide, if not
worldwide, upstream market for OTT B2C messaging services and the EEA-wide, if
not worldwide, downstream market for customer service and support CRM software.
This conclusion also holds for the potential broader upstream market for B2C
communication services overall, given the importance of Meta (formerly
Facebook)’s channels for CRM software providers and their business customers, as
well as the complementarity of different channels in such a potential wider market,
as well as the broader potential downstream market for CRM software overall, given
the importance of customer service and support CRM software within a potential
overall CRM software market. As outlined in Section 7.3.2.2.(a.i) above, the
Commission’s conclusion would not change on the basis of the potential segments of
the customer service and support CRM software market, or the overall CRM
software market, based on (i) business customer size, (ii) mode of deployment or
(iii) the industry sector of the business customer because OTT B2C messaging
channels would remain an equally important input irrespective of such further
segmentation.
(513) In view of the above considerations and in light of the results of the market
investigation and the evidence and information available to it, the Commission
concludes that post-Transaction, Meta (formerly Facebook) would have the ability
and the incentive to target foreclose (partially or totally) at certain third-party access
to Meta (formerly Facebook)’s B2C messaging channels. The Commission
concludes that such targeted foreclosure (be it partial or total) would significantly
impede effective competition on the market for customer service and support CRM
software, as well as the broader market for CRM software overall or any potential
segments thereof.
7.4. Assessment of horizontal effects
(514) In Section 7.3.2.2.(b), the Commission took into account the potential accumulation
of data by Meta (formerly Facebook) through the ownership of Kustomer as an
incentive to engage in an input foreclosure strategy on the market for customer
service and support CRM software. The Commission has also separately considered
whether the accumulation of data might lead to raised barriers to entry and expansion
on the market for online display advertising or any segment thereof.
(515) Although the Transaction does not give rise to any horizontally affected markets in a
traditional sense, the Commission considers that, after the Transaction, the
145
availability to Meta (formerly Facebook) of commercially exploitable data, which
can be obtained from Kustomer’s activities, would be augmented.
(516) In particular, the Transaction may result in an increase of Meta (formerly
Facebook)’s data collection capabilities through the acquisition of Kustomer if Meta
(formerly Facebook) incentivises Kustomer’s customers to give access to their data.
Meta (formerly Facebook)’s databases are not made available today to third parties
and Kustomer, as a processor under the GDPR, cannot on its own take a decision to
give access to a database of its business customers without agreement from the
business customer. [Business Plans in relation to Data Sharing].
771
(517) The Horizontal Merger Guidelines set out a number of factors which may influence
whether significant non-coordinated effects are likely to result from a merger.
According to paragraph 36 of the Horizontal Merger Guidelines, a merger can
significantly impede effective competition if the merged entity gains such a degree of
control over an asset that expansion or entry by rival firms may be more difficult.
(518) In the present case, the Transaction would allow Meta (formerly Facebook) to
combine its already very prominent datasets with data it could potentially obtain
through the acquisition of Kustomer by incentivising Kustomer’s customers to share
data, thus strengthening the Parties’ ability to further target their supply in the online
display advertising market, and any sub-segment thereof, and creating or reinforcing
barriers to entry and expansion in such market. Such a concern would arise to the
extent the merged entity has the ability to combine Meta (formerly Facebook)’s
datasets with data it could obtain through the acquisition of Kustomer and is
therefore merger specific.
(519) The Commission notes in this respect that there are certain regulatory limitations to
prevent the illegal combination of datasets, as previously set out in Apple/Shazam
772
and Google/Fitbit
773
.
(520) First, the Commission recalls that the processing of personal data is subject to the
applicable EU rules dealing with personal data protection, and most notably to
Regulation (EU) 2016/679 of the European Parliament and of the Council
774
(the
“GDPR”). Such rules apply to personal data, that is "any information relating to an
identified or identifiable natural person (‘data subject’); an identifiable natural person
is one who can be identified, directly or indirectly, in particular by reference to an
identifier such as a name, an identification number, location data, an online identifier
or to one or more factors specific to the physical, physiological, genetic, mental,
economic, cultural or social identity of that natural person"
775
.
(521) Pursuant to Article 4 GDPR, a controller is the natural or legal person which
determines the purposes and means of the processing of personal data, while a
processor is a natural or legal person which processes personal data on behalf of the
771
Form CO, paragraph 3.2.12.
772
Commission decision of 6 September 2018 in case M.8788 – Apple/Shazam, paragraphs 225-235.
773
Commission decision of 17 December 2020 in case M.9660 – Google/Fitbit, paragraphs 403-413.
774
Regulation (EU) 2016/679 of the European Parliament and of the Council of 27 April 2016 on the
protection of natural persons with regard to the processing of personal data and on the free movement of
such data, and repealing Directive 95/46/EC (General Data Protection Regulation),OJ L 119, 4.5.2016,
page 1. In this Decision, the Commission discusses the rules under that Regulation only for the
purposes of the assessment the Transaction under the Merger Regulation. The analysis in this Section is
therefore entirely without prejudice to the relevant administrative or legal procedures where the Parties'
compliance with those rules may be assessed.
775
GDPR, Article 4(1).
146
controller.
776
A processor may not process data except on instructions from the
controller, unless required to do so by EU or Member State law.
777
(522) Pursuant to Article 5(1)(b) GDPR, personal data which has been collected for
specified, explicit and legitimate purposes may not be further processed in a manner
that is incompatible with those purposes. Data which qualifies as personal data under
the GDPR can be transmitted to and processed by a third party only to the extent that
there exists a legal basis for the transmission to the third party and a legal basis for
the processing by that third party. Further, the GDPR requires that individuals
concerned by the processing must be informed in a transparent manner on all
relevant circumstances of the processing, including on the identity of each controller
and the purposes of the processing.
778
(523) Second, the Commission also recalls that Union rules dealing with privacy and the
protection of the confidentiality of communications, notably Directive 2002/58/EC
of the European Parliament and of the Council
779
(the “e-Privacy Directive”) may
also pose some limitations to data combinations.
(524) Article 5(3) of the e-Privacy Directive requires that Member States ensure that the
storing of information or gaining access to information already stored in the terminal
equipment of a subscriber or user is only allowed on condition that the subscriber or
user concerned has given his or her agreement, having been provided with clear and
comprehensive information, in accordance with the GDPR, inter alia, about the
purposes of the processing. This does not prevent any technical storage or access for
the sole purpose of carrying out the transmission of a communication over an
electronic communications network, or as strictly necessary for the provider of an
information society service explicitly requested by the subscriber or user to provide
the service.
(525) Meta (formerly Facebook) and Kustomer are obliged to implement appropriate
technical and organizational measures to ensure and to demonstrate that processing is
performed in accordance with the GDPR and the e-Privacy Directive, as transposed
in Member States’ laws. In particular, they must ensure the lawfulness of their data
collection from the users’ devices and of the processing of personal data they collect
and have to comply with the principles relating to the processing of personal data,
including the principles of purpose limitation, fairness, legality (in particular having
chosen appropriate legal basis) and transparency.
(526) Without prejudice to the assessment of the matter by the competent data protection
authorities, the assessment of the effects of the Transaction under the Merger
Regulation in this Decision is predicated on the assumption that Meta (formerly
Facebook) could gain access to additional datasets through the acquisition of
Kustomer. Even if any possible future access to additional datasets would not be in
compliance with the GDPR or the e-Privacy Directive, as transposed in Member
776
GDPR, Articles 4(7) and 4(8).
777
GDPR, Article 29.
778
GDPR, Articles 5, 13 and 14.
779
Directive 2002/58/EC of the European Parliament and of the Council of 12 July 2002 concerning the
processing of personal data and the protection of privacy in the electronic communications sector
("Directive on privacy and electronic communications" or "e-Privacy Directive"), OJ L 201, 31.7.2002,
pages 37-47. In this Decision, the Commission discusses these rules only for the purposes of the
assessment the Transaction under the Merger Regulation. The analysis in this Section is therefore
without any prejudice to the relevant administrative or legal procedures where the Parties' compliance
with those rules may be assessed.
147
States’ laws, the assessment of the effects of the Transaction under the Merger
Regulation would be the same, unless the consequences of a breach would
effectively prevent the Parties from getting access to additional datasets.
(527) While there are EU rules dealing with data protection, privacy, and the protection of
the confidentiality of communications that have the aim to prevent the illegal
combination of datasets, these regulations do not eliminate the risk that the Parties’
ability to get access to such data by incentivizing Kustomer’s customers to give
access to their data could render the expansion or entry by rival firms more difficult
if not impossible.
(528) Therefore, the Commission examines whether the combination of Kustomer’s data,
or data that could be collected from customers of Kustomer, and Meta (formerly
Facebook)’s data and data collection capabilities could give rise to anticompetitive
horizontal non-coordinated effects by strengthening Meta (formerly Facebook)’s
market position in the supply of online display advertising services or any sub-
segment thereof, thereby significantly impeding effective competition in this market.
7.4.1. Legal Framework
(529) The Horizontal Merger Guidelines describe two main ways in which horizontal
mergers may significantly impede effective competition, in particular by creating or
strengthening a dominant position: (i) by eliminating important competitive
constraints on one or more firms, which consequently would have increased market
power, without resorting to coordinated behaviour (non-coordinated effects); and (ii)
by changing the nature of competition in such a way that firms that previously were
not coordinating their behaviour, are significantly more likely to coordinate and raise
prices or otherwise harm effective competition (coordinated effects) as a result of the
proposed concentration.
780
For the purpose of this Decision, only the potential non-
coordinated effects are assessed.
(530) A merger giving rise to horizontal non-coordinated effects might significantly
impede effective competition by creating or strengthening the dominant position of a
single firm, one which, typically, would have an appreciably larger market share than
the next competitor post-merger. Moreover, also mergers that do not lead to the
creation of or the strengthening of a single firm’s dominant position may create
competition concerns under the substantive test set out in Article 2(2) and Article
2(3) of the Merger Regulation. Regarding mergers in oligopolistic markets, the
Merger Regulation clarifies that “under certain circumstances, concentrations
involving the elimination of important competitive constraints that the merging
parties exerted upon each other, as well as a reduction of competitive pressure on
the remaining competitors, may, even in the absence of a likelihood of coordination
between the members of the oligopoly, result in a significant impediment to effective
competition”.
781
(531) The Horizontal Merger Guidelines list a number of factors which may influence
whether or not significant horizontal non-coordinated effects are likely to result from
a merger
782
, such as the large market shares of the merging firms or the fact that the
merged entity would be able to hinder the expansion of competitors. Not all those
780
Horizontal Merger Guidelines, paragraph 22.
781
Merger Regulation, recital 25. Similar wording is also found in Horizontal Merger Guidelines,
paragraph 25.
782
Horizontal Merger Guidelines, paragraphs 26 and further.
148
factors need to be present to make significant non-coordinated effects likely and it is
not an exhaustive list. Further, the Horizontal Merger Guidelines state that,
“according to well-established case law” very large market shares of 50% or more
may in themselves be evidence of the existence of a dominant market position.
783
(532) Specifically as regards the merged entity’s ability to hinder expansion by
competitors, the Horizontal Merger Guidelines state that some proposed mergers
would significantly impede effective competition by leaving the merged firm in a
position where it would have the ability and incentive to make the expansion of
smaller firms and potential competitors more difficult or otherwise restrict the ability
of rival firms to compete. This could be the case, for example, where the merged
entity exercises control over certain types of inputs, or intellectual property making
the expansion or entry of rival firms more costly.
784
7.4.2. Data held by Kustomer
(533) CRM providers, including Kustomer, generally do not own / control the data that is
stored on their systems. Indeed, all CRM providers expressing an opinion indicated
during the market investigation that their business customers own the data they
store.
785
Any CRM provider would therefore need to obtain agreement / instruction
from its business customers before it could use any of its data.
(534) In paragraphs 330-337 above, the Commission set out the types of data that
Kustomer stores on their systems and that Meta (formerly Facebook) might therefore
get access to following the Transaction, were it to obtain agreement from business
customers to access and use the data. Meta (formerly Facebook) makes a distinction
in this regard between on the one hand Customer Transaction Data and on the other
hand Other Event Data.
(535) “Customer Transaction Data” (or conversions data) relates to data on completed
transactions that businesses may share with Kustomer. More specifically, this data
could include, for example, data such as customer contact information, their gender
or date of birth, as well as their order and purchase history, payment information,
return data and customer interaction data.
(536) “Other Event Data” relates to other data on customer actions falling short of
purchases (sometimes referred to as “mid-funnel” events). This includes website
views, adds to cart, adds to wish-list, store visits, app downloads, or other “custom
events” that a business may decide to record and share.
(537) As regards these other “custom events” that a business may decide to share with
Kustomer, the Notifying Party clarified that Kustomer’s CRM software allows
businesses to upload data in relation to their end customers, including data in relation
to incomplete transactions so that it appears in their Kustomer CRM software
interface. This is achieved through incorporating a snippet of code on the business’
website.
786
The data will be integrated into the relevant business’ account as custom
object data. Moreover, according to a training video demonstrating Kustomer’s CRM
software, what makes Kustomer "unique” is that it is able to bring into its CRM
software data from other systems, such as, for example, e-commerce platforms (such
783
Horizontal Merger Guidelines, paragraphs 17.
784
Horizontal Merger Guidelines, paragraph 36.
785
Replies to questionnaire 1 to CRM software providers, question 58.
786
Form CO, footnote 226. According to the Notifying Party, currently only around [a small percentage] of
customers of Kustomer actually made use of this function.
149
as Shopify), a business’ back-end system, and through widgets that are included in
for example e-commerce platforms.
787
(538) Finally, as set out previously, Kustomer’s CRM software may also generate some of
its own data. For example, Kustomer’s CRM software calculates a “customer
sentiment” score.
7.4.3. Raised barriers to entry and expansion as a result of data accumulation
(539) Some respondents to the market investigation raised concerns that the Transaction
would have a negative impact on the market for online display advertising services if
Meta (formerly Facebook) would obtain access to the data held by Kustomer, as this
might raise the barriers to entry and expansion on the market.
788
7.4.3.1. Notifying Party’s views
789
(540) In the Form CO and Article 6(1)(c) Response, the Notifying Party argues that, while
the Transaction may provide Meta (formerly Facebook) with additional data “signal”
by virtue of improved access to data collected and processed by Kustomer on behalf
of its business customers, this is an efficiency that generates material benefits for
business and end users.
790
(541) First, the Notifying Party submits that it would not be able to reduce ads service
rivals’ access to that data because: (i) any sharing of Event Data via Kustomer does
not involve Meta (formerly Facebook) accessing any new types of data that it could
not access before the Transaction; (ii) business customers control and store all of
their Event Data and can share it with Meta (formerly Facebook) and third parties
(including Meta (formerly Facebook)’s ads service rivals) independently of
Kustomer
791
; and (iii) currently, any such data stored by Kustomer can only be used
for ads purposes if the business specifically agrees and Meta (formerly Facebook)
intends to continue this Kustomer’s policy.
792
(542) Second, the Notifying Party considers that Meta (formerly Facebook) would not have
access to significant incremental Event Data because Kustomer is currently a very
small player in the CRM market (holding a fraction of Event Data held by competing
providers) and any improved access to data would therefore not materially enhance
Meta (formerly Facebook)’s position in ads services (or online ads services
considered separately).
793
(543) Third, the Notifying Party considers that any foreclosure strategy would not have
effects on competition in ads services given that Meta (formerly Facebook) faces
787
Kustomer live demo dated 24 April 2019, available at: https://kustomer.wistia.com/medias/aj05oeenqj,
last accessed on 21 July 2021
788
Replies to questionnaire 3 to B2C communication service providers and online ads service providers,
questions 72 and 75; replies to questionnaire 4 to business customers, question 75; replies to
questionnaire 6 to B2C communication service providers and online ads service providers, question 33.
789
In the Form CO, the Notifying Party submits its views on data concerns on ads services under the
Section for conglomerate effects (Form CO, paragraphs 6.118–6.125).
790
Form CO, paragraph 6.119; Article 6(1)(c) Response, paragraph 4.3.
791
The Notifying Party further clarified that businesses usually do provide this data other than via CRM
providers and even when these are used as a channel, nothing prevents ad providers from working with
their advertising customers and CRMs to enable data to be sent via CRMs. (Meta (formerly Facebook)
Comments on State of Play Meeting, paragraphs 3.2.-3.4.).
792
Form CO, paragraph 6.122.
793
Form CO, paragraph 6.123.
150
significant competition from ads service rivals (including ads services in the offline
channel).
794
(544) In the Article 6(1)(c) Response, the Notifying Party reiterated that Kustomer is and
will remain a tiny player in the CRM market, whose acquisition is not capable of
resulting in merger-specific harm as a result of an increase of barriers to entry and
expansion.
795
(545) To substantiate this point, the Notifying Party argues that in terms of advertising
numbers, the projected number of businesses using Kustomer in 2025 that are
expected to share data with Meta (formerly Facebook) equates to [Customer
Information] % of the projected number of all businesses worldwide sharing such
data with Meta (formerly Facebook) in 2025. In terms of revenue numbers,
compared to the revenue Meta (formerly Facebook) is expected to make from online
display advertising services on social networks in 2025, the revenue generated from
the data that Meta (formerly Facebook) expects will be shared by Kustomer’s
customers in 2025 would amount to less than [Customer Information]%.
796
7.4.3.2. Commission’s Assessment
(a) Meta (formerly Facebook)’s market position in online display advertising
(546) The Commission considers that Meta (formerly Facebook) holds significant market
power in the market for online display advertising and in certain segments thereof.
Moreover, Meta (formerly Facebook)’s current data collection capabilities present a
significant advantage on the market for online display advertising services.
(a.i) Meta (formerly Facebook) holds significant market power in online display
advertising
(547) As set out above, the Commission concluded that a separate market for online
display advertising services exists. For the purpose of this Decision, it can however
be left open whether the market for the supply of online display advertising services
should be further segmented by (i) on-/off-social networks, (ii) video/non-video ads,
and (iii) mobile/desktop.
(548) In relation to online display advertising and sub-markets/segments thereof, for the
purpose of this Decision, and in keeping with the Horizontal Merger Guidelines and
case law, the Commission considers that Meta (formerly Facebook) holds significant
market power in the following markets (where its market shares are above 50%,
which according to well-established case law, may in themselves be evidence of the
existence of a dominant position):
797
– In the supply of online display advertising services in Austria, Belgium,
Croatia, Czechia, Denmark, Finland, France, Germany, Greece, Hungary,
Iceland, Italy, Norway, Poland, Portugal, Romania, Slovakia, Slovenia, Spain,
and Sweden;
798
– In the supply of online display advertising services on social networks in
Austria, Belgium, Bulgaria, Croatia, Czechia, Denmark, Estonia, Finland,
794
Form CO, paragraphs 6.87 and 6.125.
795
Article 6(1)(c) Response, paragraph 4.3.
796
Article 6(1)(c) Response, paragraph 4.14.
797
Horizontal Merger Guidelines, paragraph 17.
798
Following linguistic borders within the EEA, this may also include Cyprus, Liechtenstein, Luxembourg,
and Netherlands.
151
France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Latvia,
Liechtenstein, Malta, Netherlands, Norway, Poland, Portugal, Romania,
Slovakia, Slovenia, Spain, and Sweden;
799
– In the supply of online display video advertising services in Austria, Belgium,
Croatia, Czechia, Denmark, France, Germany, Greece, Hungary, Iceland, Italy,
Norway, Poland, Portugal, Romania, Slovakia, Spain, and Sweden;
800
– In the supply of online display non-video advertising services in Austria,
Belgium, Croatia, Czechia, Denmark, Finland, France, Greece, Hungary,
Iceland, Italy, Norway, Poland, Portugal, Romania, Slovakia, Slovenia, Spain,
and Sweden;
801
– In the supply of online display advertising services on mobile in Austria,
Belgium, Croatia, Czechia, Denmark, Finland, France, Germany, Greece,
Hungary, Iceland, Italy, Netherlands, Norway, Poland, Portugal, Romania,
Slovakia, Slovenia, Spain, and Sweden;
802
and
– In the supply of online display advertising service on desktop in Slovenia.
(549) The Commission bases this conclusion on the market shares of Meta (formerly
Facebook), as outlined in Section 7.2.3.
803
(a.ii) Meta (formerly Facebook) has significant data collection capabilities
(550) In order to be successful on the market for online display advertising services, Meta
(formerly Facebook) as well as other market participants rely on data access and
collection of data in order to improve their ads services. In the Form CO, Meta
(formerly Facebook) describes that it “collects data from its users to serve relevant
ads on behalf of advertisers. It may also combine that with data that advertisers opt
to share with Facebook, in order to maximise the effectiveness of those ads”.
804
(551) The data Meta (formerly Facebook) collects could give it an advantage on the market
for online display advertising services in three ways. First, data could be used to
improve ad targeting, i.e., by using data to target an ads campaign to a specific
audience. Second, data could be used to measure “conversions” (i.e., the
effectiveness of particular ads campaigns in leading to more sales).
805
Third, Meta
(formerly Facebook) also uses data shared by advertisers for general system
improvement. This involves the improvement to predictions around which users are
likely to take the actions that are important to the relevant advertiser. If an advertiser
would, for example, choose to optimise its data use for Link Clicks, its ads will be
shown to people who may be more likely to click on those links according to
799
Following linguistic borders within the EEA, this may also include Cyprus and Luxembourg.
800
Following linguistic borders within the EEA, this may also include Cyprus, Liechtenstein, Luxembourg,
and Netherlands.
801
Following linguistic borders within the EEA, this may also include Cyprus, Liechtenstein, Luxembourg,
Germany and Netherlands.
802
Following linguistic borders within the EEA, this may also include Cyprus, Luxembourg and
Liechtenstein.
803
This is without prejudice to the possibility that Google could be found to be holding an equally
significant degree of market power in the markets at stake, in particular in the overall national markets
for the supply of online display advertising, as well as in the national markets for the supply of online
display advertising off social networks.
804
Form CO, paragraph 6.32.
805
Form CO, paragraph 6.120.
152
predictions of Meta (formerly Facebook)’s machine learning models.
806
Meta
(formerly Facebook)’s website identifies many other objectives that an advertiser
might have to share its data for general system improvement.
807
(552) Meta (formerly Facebook) collects / receives two types of data: (i) on-site data; and
(ii) off-site data. “On-site data” are gathered as a result of user behaviour on Meta
(formerly Facebook)’s services and may include a range data including, e.g., users’
connections, preferences and activities, interactions between a user and Meta
(formerly Facebook)’s services, etc. “Off-site data” are data that Meta (formerly
Facebook) collects or receives from third parties and relating to users’ behaviour on
those third parties’ products and services. Off-site data can include for example
information on the ways someone might interact with a third party’s website/app
(e.g., visits, purchases and ads seen/interacted with).
808
(553) Businesses can share off-site data with Meta (formerly Facebook) directly, or via
Meta (formerly Facebook)’s partners, through Meta (formerly Facebook) Business
Tools, which include Facebook Pixel, Conversions API, App Events via Facebook
SDKs, App Events API and Offline Conversions API.
809
• Facebook Pixel is a small piece of code that third parties can choose to add to
their websites to enable them to measure, optimise and build audiences for
advertising campaigns;
• Conversions API enables third parties to share web events from the third
party’s server with Meta (formerly Facebook). This is similar to Facebook
Pixel, but it does not rely on cookies;
• Meta (formerly Facebook) SDKs are packages of code around one or more
APIs. Third parties can choose to integrate these SDKs to their mobile apps or
websites to enable Meta (formerly Facebook) to receive information about the
actions of users on the apps/websites;
• App Events API is a tool that allows advertisers to track actions that occur in
their mobile apps or websites. This helps an advertiser to measure ad
performance and build audiences for ad targeting; and
• Offline Conversions API allows third parties to upload existing offline events
such as in-store purchases, over-the-phone bookings etc. and to match this data
to that third party’s Meta (formerly Facebook) ads campaigns.
(554) Because of its collection of on-site and off-site data, the Commission considers that
Meta (formerly Facebook) has data collection capabilities that give it a significant
advantage on the market for online display advertising services.
(555) Meta (formerly Facebook) itself confirms its data collection capabilities in the
Article 6(1)(c) Response where it states that Meta (formerly Facebook) currently
receives similar ads data as the data that Kustomer stores from [Customer
information] million businesses.
810
As set out above, this data is collected both
through Meta (formerly Facebook)’s platforms Facebook and Instagram (on-site
806
Parties’ reply to RFI 21, question 11a.
807
See: https://www facebook.com/business/help/416997652473726, last accessed on 7 December 2021
(ID: 2811).
808
Parties’ reply to RFI 17, question 13b.
809
Parties’ reply to RFI 16, question 12.
810
Article 6(1)(c) Response, para. 4.14.
153
data) as well as through Meta (formerly Facebook) Business Tools (off-site data).
The potential on-site data collection in itself is significant as Meta (formerly
Facebook) announced in its most recent quarterly report that it has 2.91 billion
monthly active users.
811
Moreover, given its significant market position in online
display advertising, the Commission considers that Meta (formerly Facebook) also
has a significant capability to attract off-site data. Similarly, the CMA concluded in
its Market Study “Online platforms and digital advertising” of 1 July 2020 that the
data capabilities of other platforms than Google and Meta (formerly Facebook) are
relatively limited, indicating that Meta (formerly Facebook) has exclusive access to a
large amount of data, giving it a significant competitive advantage over other market
participants.
812
(556) Third parties confirm Meta (formerly Facebook)’s current position on the market for
online display advertising services and its significant data advantage. One party notes
that “many SMBs use Facebook as their sole medium of communication as the
platform offers them an entire service package that is not available elsewhere. This
makes those SMBs very reliant on Facebook, which in turn controls how those
businesses make contacts, how they acquire visibility, and how Facebook manages
their data”.
813
Another company confirms that Meta (formerly Facebook) can do
much more with the data it owns: “Facebook has strong algorithms, and can infer
much more than anything Viber is capable of. Viber indicates that in the advertising
world there are not many companies capable of offering advertising at the level of
Facebook”.
814
(557) A respondent to the market investigation that is itself active on the market for online
display advertising services considers that “There are likely to be existing barriers to
entry and expansion for any provider who would like to offer online display
advertising”. As advertisers generally have limited advertising budgets, they may
seek an online ads service provider “who offers exposure to a large number of
consumers, or the ability to target particular categories of consumers. This will often
rely on a large and diverse data set, which comes both from economies of scale and
from having a large network of users. These considerations will often favour a large
incumbent”. In this regard, the respondent considers that “A more established player,
or a player with access to particularly valuable forms of data, may therefore have an
immediate advantage over new entrants”.
815
(558) Specifically in relation to the Customer Transaction Data, or the type of data that is
stored in CRM software, one respondent to the market information states that this
type of data “is the cornerstone of the economy of Facebook and Google – both
optimise to generate sales for their clients based on the behaviour of people who
have already purchased”.
816
The same respondent indicates “It has been Facebook’s
811
Meta (formerly Facebook) quarterly report, Q3 2021, press release. Available at:
https://investor fb.com/investor-news/press-release-details/2021/Facebook-Reports-Third-Quarter-
2021-Results/default.aspx, last accessed on 19 December 2021 (ID: 2802).
812
CMA, Market Study “Online platforms and digital advertising” of 1 July 2020, Appendix F,
paragraph 58.
813
Agreed minutes of the call of 3 June 2021 with Snapchat, paragraph 26 (ID: 2104).
814
Agreed minutes of the call of 16 September 2021 with Viber, paragraph 9 (ID: 2033).
815
Reply to questionnaire 3 to B2C communication service providers and online ads service providers,
question 72.1 (ID: 864).
816
Reply to questionnaire 4 to business customers, question 66.1 (ID: 768).
154
strategy to request companies to share as much information as possible with them –
be it via the Pixel or Facebook connect”.
817
(a.iii) Conclusion
(559) Based on the evidence presented in Section 7.5.3.2.(a) above, the Commission
considers that Meta (formerly Facebook) holds at least significant market power in
the market for online display advertising services and that already prior to the
Transaction Meta (formerly Facebook) has data collecting capabilities that provide a
significant data advantage.
(b) Potential data accumulation through the acquisition of Kustomer
(560) Before assessing the potential data accumulation by Meta (formerly Facebook) as a
result of the acquisition of Kustomer, it should be specified that the acquisition will
not lead directly to an increased market share of Meta (formerly Facebook) on the
market for online display advertising services or any segment thereof. Indeed,
Kustomer itself is not active on this market or any of its segments.
(561) In relation to the potential accumulation of data it is further important to note that
Kustomer itself does not own / control any of the data that its business customers
store on its systems. As a consequence, any data increment by Meta (formerly
Facebook) depends on the agreement of the business customers of Kustomer as well
as any required consent by the end customer.
818
(b.1) Meta (formerly Facebook) could encourage business customers to agree to the use of
data
(562) As set out above in paragraphs 346-348 above, the Commission considers that Meta
(formerly Facebook) will have the ability to encourage business customers to agree
to share data. Internal documents reveal that [Business Plans]
819
. In another
document, Meta (formerly Facebook) discusses [Business Plans]. In this document,
Meta (formerly Facebook) [Business Plans].
820
In practice, Meta (formerly
Facebook) [Business Plans].
821
(563) Respondents to the market investigation confirm that, in their view, the need to
obtain agreement would not hinder Meta (formerly Facebook) in its ability to obtain
a significant volume of data. Indeed, all B2C communication channel providers and
online ads service providers expressing an opinion indicate that Meta (formerly
Facebook) could encourage businesses using Kustomer to share more data through
its ownership of Kustomer.
822
A majority of the customers of Meta (formerly
Facebook) and Kustomer responding to the market investigation state that, following
the Transaction, they would be willing as a customer of CRM software to share data
with Meta (formerly Facebook) for online display advertising purposes.
823
817
Reply to questionnaire 4 to business customers, question 69.1 (ID: 768).
818
Form CO, paragraph 6.122.
819
Form CO, Attachment E2344, page 2.
820
Form CO, Attachment E027, page 4.
821
Form CO, paragraph 6.122.
822
Replies to questionnaire 3 to B2C communication service providers and online ads service providers,
question 70.
823
Replies to questionnaire 4 to business customers, question 64.
155
(b.ii) Data stored by Kustomer that is accessible to Meta (formerly Facebook) today
(564) In paragraphs 338-362 above the Commission assessed the volume, value, variety
and velocity of the data that Meta (formerly Facebook) could obtain following the
Transaction. On the basis of these findings, the Commission considers that the data
Meta (formerly Facebook) could obtain following the acquisition of Kustomer is
highly relevant for improving Meta (formerly Facebook)’s targeting capabilities for
the online display advertising services it provides. In addition to these findings, the
Commission has further assessed to what extent data that is stored by Kustomer is
accessible to Meta (formerly Facebook) today.
(565) The Notifying Party submits that the Transaction will not give Meta (formerly
Facebook) access to any types of data that it did not have access to before the
Transaction.
824
In relation to the first category of data that is stored on Kustomer’s
systems (i.e., Customer Transaction Data), Meta (formerly Facebook) submits that,
for example, customer lists are regularly shared through Meta (formerly Facebook)’s
Custom Audiences or through Meta (formerly Facebook) Business Tools. Other
Customer Transaction Data, such as actual transaction data or order history are
regularly shared through for example Facebook Pixel, Web Events API, App Events
API, SDKs, Conversions API and Offline Conversions API. Meta (formerly
Facebook) receives this information either directly from an advertiser, or indirectly
through one of its partners. As regards Other Event Data, the second category of
information that is stored in Kustomer’s systems, Meta (formerly Facebook) submits
that this is also often shared as off-site data through Meta (formerly Facebook)
Business Tools. Examples of this type of information include website visits or items
placed in a shopping cart.
825
(566) The Commission assessed during the market investigation what types of data
customers usually store in their CRM software. A majority of the responding CRM
software providers indicated the following types of data as the main categories that
are usually stored in their systems: customer identity, customer location, customer
contact information, purchasing history/conversions data, the number of customer
interactions and unstructured conversations data with customers.
826
While most of
this data is available to Meta (formerly Facebook), it does not receive unstructured
conversations data on channels other than its own B2C communications channels.
This is confirmed by a competitor in online display advertising services, which states
that “at a minimum, Facebook would have direct visibility into interactions between
a consumer and a business and correlate that consumer with profile information that
Facebook already has about the consumer”.
827
(567) Internal documents suggest that [Business Plans].
828
Kustomer also [Business Plans].
In a presentation to Meta (formerly Facebook), it [Business Data].
829
(568) On the basis of the above, the Commission considers that, while Meta (formerly
Facebook) might already have access to many if not most of the types of data that it
might obtain through its acquisition of Kustomer, it might still get access to certain
824
Form CO, paragraph 6.122i.
825
Parties’ reply to RFI 21, question 12.
826
Replies to questionnaire 1 to CRM software providers, question 60.
827
Reply to questionnaire 3 to B2C communication service providers and online ads service providers,
question 64 (ID: 855).
828
Form CO, Attachment E14223, page 2.
829
Form CO, Attachment E0010, slide 10.
156
new types of information. This could include in particular unstructured conversation
data. While Meta (formerly Facebook) submits that unstructured data is [Regulatory
Submission]
830
, the Commission cannot rule out that consumer interaction data from
B2C communication channels other than those of Meta (formerly Facebook) could
be potentially rich sources of information despite it being unstructured as is
suggested in public sources.
831
(569) The Commission finally took into account to what extent Meta (formerly Facebook)
might have access to data that is held by Kustomer, even if it does not represent ‘new
types’ of data. As regards the volume of new data that Meta (formerly Facebook)
might receive, the Notifying Party presented an estimate of the number of
Kustomer’s business customers that already share data with Meta (formerly
Facebook) for advertising purposes through some of Meta (formerly Facebook)
Business Tools (Facebook Pixel, SDKs and/or Conversions API). Of the [Customer
Information] business customers of Kustomer (as of 31 December 2020), [Customer
Information] customers have a positive ad spend with Meta (formerly Facebook) in
2020. This amounts to approximately [Customer Information]%.
832
Moreover, in its
Valuation Model of Kustomer, Meta (formerly Facebook) estimated that [estimates
in relation to ads data].
833
(570) The Commission considers that while some business customers may already share
similar data with Meta (formerly Facebook) through other means than Kustomer’s
CRM software, it cannot exclude that business customers may be more willing to
share such data following the Transaction. In this regard, a majority of respondents
expressing an opinion indicated that they would be willing to share Customer
Transaction Data with Meta (formerly Facebook) for online display advertising
purposes following the Transaction.
834
Moreover, the Commission is not able to
verify Meta (formerly Facebook)’s estimation that [estimates in relation to ads data].
(b.iii) Conclusion
(571) On the basis of the assessment in Section 7.4.3.2.(b) above, the Commission
concludes that Meta (formerly Facebook) may be able to gain some additional data
that could potentially be used to improve Meta (formerly Facebook)’s online display
advertising services granting it a further data advantage on the market for online
display advertising services or any segment thereof.
(c) Effects of data accumulation on the market for online display advertising services
(572) In order to assess the effects of the data accumulation on the market for the supply of
online display advertising services or any sub-segment thereof, the Commission
assessed to what extent competitors in the market would face a significant increase in
830
Article 6(1)(c) Response, paragraph 4.16.
831
See for example: Hubspot, Unstructured Data Vs. Structured Data: A 2-Minute Rundown, 9 April 2020,
available at: https://blog.hubspot.com/marketing/unstructured-data, last accessed on 20 December 2021
(ID: 2816); Forbes, What is unstructured data and why is it so important to businesses? An easy
explanation for anyone, 16 October 2019, available at:
https://www.forbes.com/sites/bernardmarr/2019/10/16/what-is-unstructured-data-and-why-is-it-so-
important-to-businesses-an-easy-explanation-for-anyone/?sh=3758697615f6, last accessed on 20
December 2021 (ID: 2807); and Simpli.fy, Winning with unstructured data in programmatic, 13
September 2018, available at: https://simpli.fi/winning-with-unstructured-data-in-programmatic/, last
accessed on 20 December 2021 (ID: 2806).
832
Parties’ reply to RFI 17, question 10a.
833
Parties’ reply to RFI 17, question 1a.
834
Replies to questionnaire 4 to business customers, question 64.
157
the barriers to entry and expansion. Further, the Commission assessed whether
similar datasets would be available to competitors and whether competitors have
counterstrategies available to counter the potential data accumulation by Meta
(formerly Facebook).
(c.i) Barriers to entry and expansion
(573) Above, the Commission considered that despite the fact that Kustomer itself does not
own / control the data of its business customers and despite the fact that Meta
(formerly Facebook) might already receive similar data through other means, the
Transaction may lead to further data accumulation by Meta (formerly Facebook).
(574) The Notifying Party submits that any accumulation of data will not have a substantial
impact on competition in the market for online display advertising services or any
segment thereof. First, it argues that because of the tiny scale of Kustomer, the
Transaction will not provide a sufficient competitive advantage to Meta (formerly
Facebook) to substantially impact competition. In the success case where Meta
(formerly Facebook) [estimates in relation to ads data]. Second, Meta (formerly
Facebook) argues that also by revenue numbers, the impact of the data accumulation
through the acquisition of Kustomer would be limited. Meta (formerly Facebook)
forecasts that [estimates in relation to ads data].
835
(575) Contrary to Meta (formerly Facebook)’s statements, a majority of B2C
communication service providers and online ads service providers believe that Meta
(formerly Facebook)’s access to data through the acquisition of Kustomer would
increase barriers to entry and expansion on the market for the supply of online
display advertising services.
836
Respondents state that:
(1) “Naturally, the more data FB has the more barrier it puts on its competition
for targeting”
837
;
(2) “As a general matter, any increase in scale, as well as access and control over
user data, by market leaders increases the already high barriers to entry and
expansion.”
838
;
(3) “There are likely to be existing barriers to entry and expansion for any
provider who would like to offer online display advertising. These barriers are
often related to establishing a provider’s reach and audience, rather than pure
technical barriers. In particular, advertisers generally have limited advertising
budgets so may seek out a provider who offers exposure to a large number of
consumers, or the ability to target particular categories of consumers. This will
often rely on a large and diverse data set, which comes both from economies of
scale and from having a large network of users. These considerations will often
favour a large incumbent. Growing this network of users can also be a
challenge for social networks in particular, as users are attracted to networks
where they can find a significant number of friends/professional
contacts/people with similar interests. A more established player, or a player
835
Article 6(1)(c) Response, paragraph 4.14.
836
Replies to questionnaire 3 to B2C communication service providers and online ads service providers,
question 72.
837
Reply to questionnaire 3 to B2C communication service providers and online ads service providers,
question 72 (ID: 741).
838
Reply to questionnaire 3 to B2C communication service providers and online ads service providers,
question 72 (ID: 873).
158
with access to particularly valuable forms of data, may therefore have an
immediate advantage over new entrants in this respect.”
839
(576) All B2C communication service providers and online ads service providers
expressing an opinion indicate that the data Meta (formerly Facebook) could get
access to would provide it with an increment of data that materially improves its
ability to personalise its online display advertising services, in particular with regard
to ad targeting.
840
Moreover, all those B2C communication service providers and
online ads service providers expressing an opinion consider that Meta (formerly
Facebook)’s ownership of Kustomer would allow it to attract a larger portion of an
advertiser’s online ad spend.
841
One respondent states in this regard: “the transaction
clearly presents an opportunity to provide the advertiser with a greater
understanding of the interactions between the consumer and business and how to
correlate those interactions with advertising profiles and behaviour on Facebook’s
platform.”
842
(577) Finally, several respondents to the market investigation expect the Transaction to
have a negative impact both on their company
843
, on the market for the supply of
online display advertising services or sub-segments thereof
844
and on businesses
purchasing online display advertising services
845
. Respondents state, among others:
(1) “With its acquisition of Kustomer, Facebook has the opportunity to create
“super-profiles”, which could confer a competitive advantage to Kustomer and
Facebook and may raise consumer data privacy concerns.”
846
; and
(2) “Depending on the data shared and added, Facebook which is already a
dominant player, could become an even more dominant player. In our view,
there are already markets where they are essentially the only effective channel
for targeted ads, which makes it very easy for them to increase prices.”
847
(578) The Commission considers it unlikely that the Transaction will lead to a significant
increase in the barriers to entry and expansion on the market for online display
advertising services or any segments thereof. While Meta (formerly Facebook) holds
a significant position in the market for online display advertising services and has
significant data collection capabilities prior to the Transaction, the size of Kustomer
would make a significant data increase unlikely, even taking into account Meta
(formerly Facebook)’s ambitious growth targets. Moreover, the Commission
839
Reply to questionnaire 3 to B2C communication service providers and online ads service providers,
question 72 (ID: 864).
840
Replies to questionnaire 3 to B2C communication service providers and online ads service providers,
question 68.
841
Replies to questionnaire 3 to B2C communication service providers and online ads service providers,
question 69.
842
Reply to questionnaire 3 to B2C communication service providers and online ads service providers,
question 69.1 (ID: 855).
843
Replies to questionnaire 3 to B2C communications services providers and online ads service providers,
question 73; replies to questionnaire 4 to business customers, question 70.
844
Replies to questionnaire 3 to B2C communications services providers and online ads service providers,
question 75; replies to questionnaire 4 to business customers, question 76.
845
Replies to questionnaire 6 to B2C communication service providers and online ads service providers,
question 33.
846
Reply to questionnaire 3 to B2C communication service providers and online ads service providers,
question 75.1 (ID: 886).
847
Reply to questionnaire 3 to B2C communication service providers and online ads service providers,
question 75.1 (ID: 741).
159
considered already above that Kustomer itself does not own any data and that at
present Meta (formerly Facebook) already has access to the same type of data and
data from a significant percentage of Kustomer’s customers. The Commission
assesses below whether similar datasets are available to competitors of Meta
(formerly Facebook) and whether its competitors have any potential
counterstrategies.
(c.ii) Availability of similar datasets and counterstrategies
(579) The Notifying Party submits that datasets similar to the one that Meta (formerly
Facebook) could obtain through its ownership of Kustomer would also be available
to its competitors. First, businesses have the ability to share customer transaction
data and other events data either directly or indirectly through other platforms rather
than through their CRM software. Second, even if CRM software would be
important for sharing this type of data, the Notifying Party argues that online ads
service providers could obtain this data through other CRM providers. For example,
Google would currently have integrations with software of Salesforce and Hubspot
through which business may decide to share data with Google.
848
(580) The Commission observes that businesses can share data with online ads service
providers in several ways depending on how they store data. A business can have a
stack of software that it uses in the ordinary course of business to record sales and
purchases, monitor inventory, for accounting and financial functions, customer
service, marketing, shipping and logistics. All these types of data could for example
be stored in CRM software, but also on e-commerce platforms and data management
platforms, in business analytics software etc. As such, a business is not dependent on
its CRM software to share data and business customers of Kustomer will not and
cannot be prevented to share data with competing online ads service providers. This
is also reflected in the results from the market investigation. Almost all respondents
to the market investigation stated that they make use of online display advertising
services of companies other than Meta (formerly Facebook).
849
In doing so, the vast
majority indicates that share data for ads purposes.
850
(581) For the reasons above, the Commission considers that competitors of Meta (formerly
Facebook) on the market for online display advertising services will still be able to
get access to data from customers of Kustomer. The Commission further assessed
whether competitors of Meta (formerly Facebook) are able to get access to data
through CRM software.
(582) The Commission notes first that Meta (formerly Facebook) has internally considered
[Business Plans]. One internal document states in this regard that [Business Plans].
851
(583) The results from the market investigation indicate that similar datasets could are
available. While all CRM providers and CPaaS providers expressing an opinion
indicate that they do not sell data from their customers
852
, the majority of responding
CRM providers allow their customers to share data directly with online ads service
848
Meta (formerly Facebook) comments on the State of Play Meeting, 22 July 2021, paragraphs 3.2 and
3.5; and Article 6(1)(c) Response, paragraph 4.22.
849
Replies to questionnaire 4 to business customers, question 6.
850
Replies to questionnaire 4 to business customers, question 6.2.
851
Form CO, Attachment E14233, page 1.
852
Replies to questionnaire 1 to CRM software providers, question 62; replies to questionnaire 2 to CPaaS
providers, question 50.
160
providers.
853
Moreover, every CRM provider expressing an opinion indicates that
business customers do make use of the ability to share data through their CRM
software.
854
While only a minority of the responding customers indicate that they
currently share data through their CRM software
855
, a majority of the responding
online ads service providers confirm that businesses currently share data through
their CRM software.
856
(584) Finally, two responding online ads service providers indicate that they would have
strategies available to counter the potential accumulation of data by Meta (formerly
Facebook) as a result of the Transaction. One respondent states that
“Counterstrategies include multiple partnerships with other CRM providers, and
working with ecommerce, messaging channels outside of Facebook”.
857
Another
respondent states that “rival online advertising companies would have a range of
options available to them in order to gather similar data, including through
partnerships with Kustomer itself and with other CRM tools”.
858
(c.iii) Conclusion
(585) On the basis of all of the evidence set out in the above, the Commission considers
that the potential data accumulation as a result of the Transaction will not give rise to
a significant increase in the barriers to entry and expansion on the market for online
display advertising services or any segment thereof. Moreover, the Commission
concludes that competitors on this market would still have access to data from
Kustomer’s business customers, or alternatively could get access to similar datasets
through for example partnerships with other CRM software providers.
7.4.3.3. Conclusion
(586) In view of the above considerations and in light of the results of the market
investigation and the evidence and information available to it, the Commission
considers that the Transaction would not significantly impede effective competition
in respect of the supply of online display advertising services, and or any sub-
segment thereof.
7.5. Assessment of conglomerate effects
7.5.1. Legal Framework
(587) According to the Non-Horizontal Guidelines, in the majority of circumstances,
conglomerate mergers will not lead to any competition problems.
859
(588) However, foreclosure effects may arise when the combination of products in related
markets may confer on the merged entity the ability and incentive to leverage a
strong market position from one market to another closely related market by means
of tying or bundling or other exclusionary practices. While tying and bundling have
often no anticompetitive consequences, in certain circumstances such practices may
853
Replies to questionnaire 7 to CRM software providers, question 51.
854
Replies to questionnaire 7 to CRM software providers, question 52.
855
Replies to questionnaire 5 to business customers, question 16.2.
856
Replies to questionnaire 6 to B2C communication service providers and online ads service providers,
question 18.
857
Reply to questionnaire 6 to B2C communication service providers and online ads service providers,
question 24 (ID: 1658).
858
Reply to questionnaire 6 to B2C communication service providers and online ads service providers,
question 24 (ID: 1667).
859
Non-Horizontal Guidelines, paragraph 92.
161
lead to a reduction in actual or potential competitors' ability or incentive to compete.
This may reduce the competitive pressure on the merged entity allowing it to
increase prices.
860
(589) In assessing the likelihood of such a scenario, the Commission examines, first,
whether the merged firm would have the ability to foreclose its rivals
861
, second,
whether it would have the economic incentive to do so
862
and, third, whether a
foreclosure strategy would have a significant detrimental effect on competition, thus
causing harm to consumers.
863
In practice, these factors are often examined together
as they are closely intertwined.
(590) In order to be able to foreclose competitors, the merged entity must have a significant
degree of market power, which does not necessarily amount to dominance, in one of
the markets concerned. The effects of bundling or tying can only be expected to be
substantial when at least one of the merging parties’ products is viewed by many
customers as particularly important and there are few relevant alternatives for that
product.
864
Further, for foreclosure to be a potential concern, it must be the case that
there is a large common pool of customers, which is more likely to be the case when
the products are complementary.
865
Finally, bundling is less likely to lead to
foreclosure if rival firms are able to deploy effective and timely counter-strategies,
such as single-product companies combining their offers.
866
(591) The incentive to foreclose rivals through bundling or tying depends on the degree to
which this strategy is profitable.
867
Bundling and tying may entail losses or foregone
revenues for the merged entity.
868
However, they may also allow the merged entity to
increase profits by gaining market power in the tied goods market, protecting market
power in the tying good market, or a combination of the two.
869
(592) It is only when a sufficiently large fraction of market output is affected by
foreclosure resulting from the concentration that the concentration may significantly
impede effective competition. If there remain effective single-product players in
either market, competition is unlikely to deteriorate following a conglomerate
concentration.
870
The effect on competition needs to be assessed in light of
countervailing factors such as the presence of countervailing buyer power or the
likelihood that entry would maintain effective competition in the upstream or
downstream markets.
871
7.5.2. Leveraging of Meta (formerly Facebook)’s position in the online display advertising
market into the CRM market
(593) The Notifying Party considers that businesses that buy CRM software may also buy
online advertising services, and in this sense these two markets are related.
872
860
Non-Horizontal Guidelines, paragraphs 91 and 93.
861
Non-Horizontal Merger Guidelines, paragraphs 95-104.
862
Non-Horizontal Merger Guidelines, paragraphs 105-110.
863
Non-Horizontal Merger Guidelines, paragraphs 111-118.
864
Non-Horizontal Merger Guidelines, paragraph 99.
865
Non-Horizontal Merger Guidelines, paragraph 100.
866
Non-Horizontal Merger Guidelines, paragraph 103.
867
Non-Horizontal Merger Guidelines, paragraph 105.
868
Non-Horizontal Merger Guidelines, paragraph 106.
869
Non-Horizontal Merger Guidelines, paragraph 108.
870
Non-Horizontal Merger Guidelines, paragraph 113.
871
Non-Horizontal Merger Guidelines, paragraph 114.
872
Form CO, paragraph 6.8.
162
(594) During the market investigation, the concern has been raised that the Transaction
may allow Meta (formerly Facebook) to leverage its position in the market for online
display advertising to strengthen Kustomer’s position in the CRM software
market.
873
(595) Such practices could be related to Meta (formerly Facebook)’s potential bundling of
online display advertising services and customer service and support CRM services.
(596) Several respondents (including CRM and CPaaS providers, as well as business
customers) also considered that Meta (formerly Facebook) would have incentives to
offer Kustomer’s CRM software for free or at a discount following the
Transaction.
874
(597) As regards the offering of Kustomer for free or at a discount, the Commission
considers this would, in the short term, allow business customers to benefit from the
lower or reduced price of its CRM software. In the longer term, this strategy (in
parallel with an input foreclosure strategy) could result in a reduction of competition
in the customer service and support CRM market.
875
The Commission addressed
such a concern in Section 7.3.2.2.(b.ii)C where it concluded that offering Kustomer
for free or below market price would not be an inherently anti-competitive strategy
per se but would exacerbate Meta (formerly Facebook)’s incentive to pursue a
targeted input foreclosure strategy of restricting or degrading API access, by
reducing the associated losses (i.e., by further encouraging switching to Kustomer,
thereby foreclosing CRM software providers and increasing revenues).
(598) As for the potential bundling of online display advertising with CRM software, the
Commission has further investigated whether post-Transaction Meta (formerly
Facebook) would have the ability and incentive to engage in such strategy, as well as
the potential resulting detrimental effect on competition.
7.5.2.1. Notifying Party’s views
(a) Ability
(599) The Notifying Party submits that Meta (formerly Facebook) would not have the
ability to foreclose CRM rivals by engaging in a conglomerate strategy related to its
ads services because it has an estimated market share in the market for online ads
services of only [30-40%] in the EEA and [20-30%] globally and faces competition
from many rivals (such as Google, Apple, YouTube or Amazon). The Notifying
Party considers Meta (formerly Facebook) also faces competition from offline
channels such as print, radio and TV, including Video On Demand.
876
(b) Incentive
(600) The Notifying Party also submits that Meta (formerly Facebook) would not have the
incentive to foreclose CRM rivals by engaging in such strategy because it contradicts
873
Agreed minutes of the call of 9 September 2021 with Zendesk, paragraph 12 (ID: 2085).
874
Replies to questionnaire 1 to CRM software providers, question 57; replies to questionnaire 2 to CPaaS
providers, question 45; replies to questionnaire 4 to business customers, question 75.1.
875
In order for the strategy of offering Kustomer for free to result in a reduction of competition, the ability
of rival CRM software providers (either existing or new entrants) to compete would have to be
significantly reduced so that the long-term impact on price, quality and innovation in the market more
than offset the short-term benefits to customers from the lower price of Kustomer’s software.
876
Form CO, paragraph 6.115.
163
Meta (formerly Facebook)’s strategy to increase the use of messaging for B2C
communications.
877
(601) According to the Notifying Party, such strategy would also require Meta (formerly
Facebook) to [Business Plans].
878
(602) Moreover, the Notifying Party submits it would not be able to recover those losses as
such strategy would not lead to material gains in the CRM software market. That is
because (i) Kustomer is a small player in the CRM market, (ii) in case of a price
increase in the standalone price of Meta (formerly Facebook)’s ads services (relative
to the bundle), customers of rival CRM providers are unlikely to switch to Kustomer
due to its size, scale and limited product offering, (iii) many Meta (formerly
Facebook) ads service customers do not use CRM software, (iv) switching CRM
software providers is costly and complex, and (v) CRM services and ads services are
not natural complements so there would be limited customer demand for a combined
product.
879
(603) The Notifying Party also submits that the valuation model submitted by Meta
(formerly Facebook) assumes that [Business Plans].
880
(c) Effects
(604) Finally, given the low level of probable switching to Kustomer due to the
aforementioned reasons, the Notifying Party considers that rival CRM providers
would not suffer material losses as a result of such foreclosure strategy which could
not therefore have a significant detrimental effect on competition.
881
7.5.2.2. Commission’s Assessment
(605) For the reasons set out below and based on the results of the market investigation, the
Commission considers, for the purposes of the present Decision, that Meta (formerly
Facebook) would not have the ability to bundle its online display advertising services
with CRM software. Therefore, for the purposes of the present Decision, the
Commission considers that the question whether Meta (formerly Facebook) would
have the incentive to bundle its online display advertising services with CRM
software and whether there would likely be a significant detrimental effect on
competition in the CRM software market as a result, are not relevant and can be left
open.
(a) Ability to foreclose
(606) For the reasons set out below, the Commission considers that Meta (formerly
Facebook) would not have the ability to engage in a bundling strategy of online
display advertising services with CRM software.
(607) As previously mentioned
882
, for the purposes of this Decision, the Commission
considers Meta (formerly Facebook) has a significant degree of market power in
online display advertising where its services are viewed by many customers as
particularly important and there are few relevant alternatives for that product.
877
Form CO, paragraph 6.116.
878
Form CO, paragraph 6.116.
879
Form CO, paragraph 6.116.
880
Article 6(1)(c) Response, paragraph 5.3.
881
Form CO, paragraph 6.117.
882
Sections 7.2.3. and 7.4.3.2.(a.i) above.
164
(608) Moreover, in the Commission’s market investigation, the majority of market
respondents that expressed a view considered that CRM software services and online
display advertising are generally purchased by the same set of business customers
883
and are complementary products.
884
(609) In this regard, several providers of online advertising services explained that the data
in CRM software services may be useful to track the effectiveness and improve the
targeting of online advertising campaigns.
885
One market respondent highlighted that
“online display advertising services can be effective forms of customer acquisition
and, at times, customer re-engagement. CRM software is best for organizing your
existing customer information and enabling activities with that data. Many
companies use both types of services to grow their business”.
886
The same
respondent explained why it is unclear whether CRM software services and online
display advertising services are closely related markets while confirming the
complementarity of the two products: “a significant amount of business’s online
display advertising spend is focused on building brand awareness and acquiring new
customers that the business could not otherwise reach directly (e.g. such as via an
email). CRM software is used to manage information, interactions, and workflows
with customers that are definitionally “known” by the business. As stated above, this
makes these services complementary -- customers are acquired via display
advertising and from that point most engagement with these customers is mediated
through a CRM. Some CRM providers offer product capabilities for using CRM data
to inform the purchase of display advertising, and others offer products that also
facilitate this purchase. However, especially at larger customers, these services are
purchased and used by different teams, and the ecosystems and jobs to be done
around both of these markets tend to be distinct. As such, it is difficult to determine
just how closely related these markets are”.
887
(610) Nevertheless, the majority of market respondents that expressed a view also
indicated that CRM software services and online display advertising services tend to
be purchased separately (i.e., at different points in time)
888
. One respondent
explained “this would be highly customer specific. We would expect that there are
many companies that purchase both online display advertising services and CRM
services, as well as many that only purchase one or the other. Even if a company
were to purchase both services, we would expect most to source these services
independently of each other.”
889
(611) Although respondents were not entirely clear on whether such purchases are made at
very different intervals (e.g., yearly purchase/renewal of a CRM software license and
883
Replies to questionnaire 7 to CRM software providers, question 54; replies to questionnaire 6 to B2C
communication service providers and online ads service providers, question 26.
884
Replies to questionnaire 7 to CRM software providers, question 53; replies to questionnaire 6 to B2C
communication service providers and online ads service providers, question 25.
885
Replies to questionnaire 6 to B2C communication service providers and online ads service providers,
question 25.1.
886
Reply to questionnaire 7 to CRM software providers, question 53.1 (ID: 1713).
887
Reply to questionnaire 7 to CRM software providers, question 56.1 (ID: 1713).
888
Replies to questionnaire 7 to CRM software providers, question 55; replies to questionnaire 6 to B2C
communication service and online ads service providers, question 27.
889
Reply to questionnaire 6 to B2C communication service providers and online ads service providers,
question 26.1 (ID: 1568).
165
a daily purchase of online display advertising services)
890
, the Commission considers
that this is likely the case as many respondents pointed out their purchase involves
separate and very different procedures that are usually handled by different
departments in the same business customer (usually the purchase of CRM software is
handled by the department in charge of sales whereas the marketing department is in
charge of the purchase of advertising). In this regard, one respondent highlighted that
“these are two distinct teams in a business - CRM is usually governed by the Sales
department, both for new sales, and for existing account managers to manage the
ongoing commercial relationship, while advertising is done through the marketing
department, which can either be an internal team to the business, or more usually, a
3rd party Agency dedicated in managing online advertising for their customers”.
891
Another respondent explained that “the purchase of enterprise software is a major
procurement exercise (often involving RFPs, etc.) that involves complex integrations
with other business systems. Advertising can be purchased and redirected on a much
simpler basis (often through self-serve auction features)”.
892
(612) The different purchasing patterns of online display advertising and CRM may also be
related to a company’s longevity. Indeed, one respondent explained that “although
the same customers may purchase online display advertising services and CRM
software services, the purchases may occur at different stages in the company’s
lifecycle. When a company is first launched, the founders may purchase online
display advertising to advertise the company. Once the company achieves some
success, the company may then purchase CRM software.”
893
Another respondent
confirmed this, stating “of course it depends on the business, but when you are
growing, you need online advertising immediately while a CRM becomes beneficial
only after a certain scale. Generally, learning about your customers first gives you
more information to pick the right tool for the job”.
894
(613) There was no clear consensus among CRM providers on whether they would be able
to replicate Meta (formerly Facebook)’s combined offer of its online advertising
services and Kustomer’s CRM software together as part of a bundle.
895
While one
respondent considered such replication is “near to impossible”
896
, another
highlighted it would need to “partner with an online media company”
897
whereas
another indicated it has “signup offers with multiple ad networks at any given
time”.
898
Given the results of the market investigation are mixed, it is unclear
whether rival firms are likely unable to deploy effective and timely counter-
strategies, such as single-product companies combining their offers.
890
Replies to questionnaire 7 to CRM software providers, question 55.2; replies to questionnaire 6 to B2C
communication service providers and online ads service providers, question 27.
891
Reply to questionnaire 6 to B2C communication service providers and online ads service providers,
question 26.1 (ID: 1658).
892
Reply to questionnaire 6 to B2C communication service providers and online ads service providers,
question 27.1 (ID: 1568).
893
Reply to questionnaire 6 to B2C communication service providers and online ads service providers,
question 27.1 (ID: 1693)
894
Reply to questionnaire 6 to B2C communication service providers and online ads service providers,
question 27.1 (ID: 2552)
895
Replies to questionnaire 7 to CRM software providers, question 58.
896
Reply to questionnaire 7 to CRM software providers, question 58.1 (ID: 2125).
897
Reply to questionnaire 7 to CRM software providers, question 58.1 (ID: 1655).
898
Replies to questionnaire 7 to CRM software providers, question 58.1 (ID: 1707).
166
(614) In light of the above, taking into consideration the overall market results and, in
particular, given the different purchasing patterns of online display advertising
services and CRM software, for the purposes of the present Decision, the
Commission considers Meta (formerly Facebook) would not have the ability to
engage in a strategy of bundling its online display advertising services with
Kustomer’s CRM software.
(b) Incentive to foreclose
(615) Since the Commission has concluded above in Section 7.5.2.2.(a) that Meta
(formerly Facebook) would not have the ability to bundle online display advertising
with CRM software, the question whether Meta (formerly Facebook) would have the
incentive to engage in such a bundling strategy (of online display advertising
services with CRM software) can be left open for the purposes of the present
Decision.
(c) Impact on effective competition
(616) Since the Commission has concluded above in Section 7.5.2.2.(a) that Meta
(formerly Facebook) would not have the ability to bundle online display advertising
with CRM software, the question whether such a bundling strategy (of online display
advertising services with CRM software) would have a detrimental effect on
competition can be left open for the purposes of the present Decision.
7.5.2.3. Conclusion
(617) In view of the above considerations and in light of the results of the market
investigation and the evidence and information available to it, for the purposes of the
present Decision, the Commission concludes the Transaction would not significantly
impede effective competition as a result of bundling of online display advertising and
CRM software, considering that Meta (formerly Facebook) would not have the
ability to engage in such a strategy.
8. COMMITMENTS
8.1. Introduction
(618) In order to remove the competition concerns arising from the Transaction described
in Section 7, the Notifying Party submitted commitments in Phase II.
(619) On 24 November 2021, the Notifying Party submitted commitments pursuant to
Articles 8(2) and 10(2) of the Merger Regulation (the “Initial Commitments”).
(620) The Commission launched a market test of the Initial Commitments on 26 November
2021 (“the market test”), seeking responses from the Parties’ customers and
competitors on the affected markets. The Commission informed the Notifying Party
of the results of the market test on 10 December 2021.
(621) Based on the Commission’s feedback, the Notifying Party submitted a revised final
set of commitments on 20 December 2020 (the “Final Commitments”).
167
8.2. Analytical Framework
(622) Where the Commission considers that a concentration raises competition concerns
parties may seek to modify the concentration in order to resolve such competition
concerns and thereby gain clearance of their merger.
899
(623) Under the Merger Regulation, the Commission only has the power to accept
commitments that are deemed capable of rendering the concentration compatible
with the internal market. The commitments must eliminate the competition concerns
entirely and must be comprehensive and effective from all points of view.
900
The
commitments must also be proportionate to the competition concerns
identified.
901
Furthermore, the commitments must be capable of being implemented
effectively within a short period of time as the conditions of competition on the
market will not be maintained until the commitments have been fulfilled.
902
(624) In assessing whether the proposed commitments will likely eliminate the competition
concerns identified, the Commission considers all relevant factors including inter alia
the type, scale and scope of the proposed commitments, assessed by reference to the
structure and particular characteristics of the market in which the competition
concerns arise, including the position of the parties and other participants on the
market.
903
(625) When assessing the commitments proposed by the merging parties, the Commission
has the legal duty to ensure that such commitments are effective. In order for the
commitments to remove the competition concerns entirely and be comprehensive and
effective, there has to be an effective implementation and ability to monitor the
commitments. Whereas divestitures once implemented do not require any further
monitoring measures, other types of commitments require effective monitoring
mechanisms in order to ensure that their effect is not reduced or even eliminated by
the parties. Otherwise such commitments would have to be considered as mere
declarations of intentions by the parties and would not amount to any binding
obligations as, due to the lack of effective monitoring mechanisms, any breach of
them could not result in the revocation of the decision in accordance with the
provisions of the Merger Regulation.
904
(626) Where the parties submit remedies proposals that are so extensive and complex that
it is not possible for the Commission to determine with the requisite degree of
certainty, at the time of its decision, that they will be fully implemented and that they
are likely to maintain effective competition in the market, an authorisation decision
cannot be granted. The Commission may reject such remedies in particular on the
grounds that the implementation of the remedies cannot be effectively monitored and
that the lack of effective monitoring diminishes, or even eliminates, the effect of the
commitments proposed.
905
(627) However, while divestiture commitments may be the best means of resolving
problems resulting from horizontal, vertical or conglomerate concerns, other
899
Commission notice on remedies acceptable under Council Regulation (EC) No 139/2004 and under
Commission Regulation (EC) No 802/2004 (the "Remedies Notice"), OJ 2008/C 267/01, paragraph 5.
900
Remedies Notice, paragraph 9.
901
Merger Regulation, recital 30.
902
Remedies Notice, paragraph 9.
903
Remedies Notice, paragraph 12.
904
Remedies Notice, paragraph 13.
905
Remedies Notice, paragraph 14.
168
structural commitments may be suitable to resolve all types of concerns if those
remedies are equivalent to divestitures in their effects,
906
such as access remedies
foreseeing the granting of access to key infrastructure, key technology, or essential
inputs, and which are normally granted on a non-discriminatory and transparent
basis.
907
Such access commitments may be submitted in order to ensure that
competition is not significantly impeded as a result of foreclosure and/or in order to
facilitate market entry by competitors.
908
Where applicable, it has to be further
ensured that the terms and conditions, including the pricing, under which the access
is granted (e.g., via a licence) do not impede the effective implementation of such a
remedy. However, the Commission will only accept such access commitments if it
can be concluded that these commitments will be effective and competitors will
likely use them so that foreclosure concerns will be eliminated.
909
(628) Access commitments are often complex in nature and necessarily include general
terms for determining the terms and conditions under which access is granted. In
order to render them effective, those commitments have to contain the procedural
requirements necessary for monitoring them, which normally has to be done by the
market participants themselves, i.e., by those undertakings wishing to benefit from
the commitments. Measures allowing third parties themselves to enforce the
commitments may in particular involve access to a fast dispute resolution mechanism
via arbitration proceedings (together with trustees). If the Commission can conclude
that the mechanisms foreseen in the commitments will allow the market participants
themselves to effectively enforce them in a timely manner, no permanent monitoring
of the commitments by the Commission may be required. However, the Commission
will only be able to accept such commitments where the complexity does not lead to
a risk of their effectiveness from the outset and where the monitoring devices
proposed ensure that those commitments will be effectively implemented and the
enforcement mechanism will lead to timely results.
910
(629) It is against this background that the Commission reviews the proposed commitments
in this case.
8.3. The Initial Commitments
8.3.1. Description of the Initial Commitments
(630) The Initial Commitments comprise two main elements: (A) a public API access
commitment, and (B) a core API functionality-parity commitment.
8.3.1.1. The public API access commitment
(631) By the public API access commitment, Meta (formerly Facebook) commits to
maintain access of Third Party CS CRM Providers (i.e., customer service and support
software providers other than Kustomer) with sales in the EEA to Publicly Available
B2C Messaging Channel APIs
911
(both existing and future ones) on a non-
discriminatory basis, meaning without differentiating in terms of (i) eligibility
criteria, (ii) API access pricing (to the extent any charge applies for access, which is
906
Remedies Notice, paragraph 17.
907
Remedies Notice, paragraph 62.
908
Remedies Notice, paragraphs 63-64.
909
Remedies Notice, paragraphs 64-65.
910
Remedies Notice, paragraphs 66.
911
This covers the B2C Messaging Channel APIs that are publicly available on Meta (formerly
Facebook)’s Developer Website as may be updated from time to time. Currently, this covers the
Messenger API and the Instagram messaging API.
169
not currently the case), (iii) functionality or (iv) performance compared to other
comparable users of those APIs.
(632) This commitment covers B2C Messaging Channel APIs that are publicly available
on Meta’s developer website as may be updated from time to time, which currently
extends to the Messenger API and the Instagram messaging API. It does not cover
the WhatsApp for Business API, as the latter is not currently publicly available on
Meta (formerly Facebook)’s developer website.
912
8.3.1.2. The core API functionality-parity commitment
(633) By the core API functionality-parity commitment, Meta (formerly Facebook)
commits to ensure that all Core B2C Messaging Channel API Functionalities (and
any future improvements on such functionalities) made available to the Kustomer
Service
913
(including any successor businesses or entities, irrespective of branding)
are also made available on an equivalent basis to Third Party CS CRM Providers.
(634) This commitment, through the definition of Core B2C Messaging Channel API
Functionalities, covers the following features of the Messenger, Instagram
Messaging and the WhatsApp Business Platforms: sending/receiving (i) text-based
messages, (ii) image attachments, and (iii) URL links.
8.3.1.3. Anti-circumvention, carve-outs and other aspects
(635) The Initial Commitments include anti-circumvention safeguards aimed at preventing
Meta (formerly Facebook) from discriminating against Third Party CS CRM
Providers through indirect means:
(a) General release and decisions affecting the availability of certain APIs or
functionalities: Meta (formerly Facebook) commits to not discriminate against
Third Party CS CRM Providers in relation to denying or unreasonably delaying
access to, changing, suspending, or deprecating Publicly Available B2C
Messaging Channel APIs (vis-à-vis other comparable users
914
) or Core B2C
Messaging Channel API Functionalities (vis-à-vis Kustomer); and
(b) The Initial Commitments extend to Third Party CS CRM Providers accessing
Meta (formerly Facebook)’s B2C Messaging channels indirectly (e.g., through
CPaaS providers) and to end business users: such users are provided access to
the Publicly Available B2C Messaging Channel APIs on a non-discriminatory
basis, and to Core B2C Messaging Channel API Functionalities on equivalent
terms irrespective of whether they are business customers of a Third Party CS
CRM Provider or business customers of Kustomer.
912
In its reply to RFI 24, Meta (formerly Facebook) indicated that it estimates that a cloud-hosted version
of its WhatsApp for Business API will become publicly available in 2022. Once this occurs, the
WhatsApp API would be captured by the non-discriminatory public API access commitment.
913
This is defined as “any CS CRM product or service provided by Kustomer at any time during the term
of the Commitments”.
914
Form RM, footnote 3, indicates that “whether an access seeker is comparable to another user would be
a factual question. As an illustrative example, business users of a Third Party CS CRM Provider that
are news organisations could be compared to news organisations accessing the API independently.
Comparable users to Third Party CS CRM Providers could include other business solution providers
that access similar messaging functionalities via the Publicly Available B2C Messaging Channel APIs
and have similar messaging needs (e.g. volumes).”
170
(636) The Initial Commitments provide for carve-outs to ensure that Meta (formerly
Facebook) will not be unduly restricted in its ability to undertake the following
measures:
(a) Develop and test new APIs or functionalities prior to the introduction of a
feature at scale. If this process results in the successful development and
release of new or improved versions of the B2C Messaging Channel APIs,
these will be made available to Third Party CS CRM Providers and their
customers as soon as Meta (formerly Facebook) makes them publicly available
via public APIs. Similarly, if this process results in the successful development
and release of improved Core Messaging Channel API Functionalities, these
will be made available to Third Party CS CRM Providers and their customers
in accordance with the Commitments as soon as they are made available to
50% or more of the business customers of Kustomer; and
(b) Develop the Kustomer Service through integration within the broader suite of
Meta Services or partnerships with third parties, including features that
complement (but do not in themselves constitute features of) the Publicly
Available B2C Messaging Channel APIs and/or Core B2C Messaging Channel
API Functionalities (to the extent, for the avoidance of doubt, that the
functioning of and access to Publicly Available B2C Messaging Channel APIs
and Core B2C Messaging Channel API Functionalities is not impeded for
Third Party CS CRM Providers).
(c) The Initial Commitments permit Meta (formerly Facebook) to take necessary
steps to, for example, comply with legal obligations, prevent fraud, promote
user privacy and security, enforce relevant terms and policies, preserve the
rights, safety, and integrity of the Meta Services users or Meta’s property as
required or permitted by law.
(637) A Fast Track Dispute Resolution is included, and will apply in cases where a Third
Party CS CRM Provider, showing a sufficient legitimate interest, claims that Meta is
failing to comply with its obligations arising from the Commitments.
(638) The Initial Commitments also provide that the Monitoring Trustee must be appointed
before the Notifying Party can close the Transaction. The Monitoring Trustee may in
turn appoint a technical expert (who may appoint other advisers) to assist in the
performance of his or her duties and obligations.
(639) The Initial Commitments would remain in effect for five years from the closing date
of the Transaction.
8.3.2. Results of Markets Tests of the Initial Commitments
(640) The Commission initiated a market test of the Initial Commitments on 26 November
2021 and received responses from competing CRM software providers, CPaaS
providers, B2C messaging service providers, business customers that use CRM
software and B2C messaging services, as well as from a Member State. The
Commission also received a response from the Parties to an RFI on the Initial
Commitments.
915
(641) The market test confirmed that several elements of the Initial Commitments were
broadly sufficient, including the principle of the public API access commitment and
915
Parties’ reply to RFI 24.
171
the fast-track dispute resolution mechanism. However, the market test also revealed
that some significant improvements are necessary, as set out below.
(642) Duration. A majority of respondents that expressed a view, including business
customers, considered that a five-year duration is insufficient.
916
Several reasons
were mentioned, including typical multi-year sales cycles in sourcing CRM software
tools, and the fact that Meta (formerly Facebook)’s OTT B2C messaging channels
are expected to remain important inputs in five years: “The Facebook B2C
communications channels are expected to continue to rule the market in 5 years and
the competitors are unlikely to create an alternative”.
917
One business customer
indicated that if they invested in using Meta (formerly Facebook)’s channels for
customer support, with the likely result that their end customers became accustomed
to using them, that business may need to switch CRM provider after five years if
those channels were no longer available, or not in the same way: “If a Customer
Service company builds it's channel strategy around Meta B2C channels, it would be
costly and difficult to change it (customers get used to it). If after 5 years the CS
CRM can not offer these services, it will be a relevant reason to change CS CRM
provider.”
918
(643) A range of alternative durations were proposed by market test respondents and a
Member State, with a minimum of 10 years being proposed in several cases,
including by competing CRM software providers: “We suggest … for example
10 years to address the EC’s competitive concerns and the likelihood of Meta’s
continued market strength and prominence”.
919
(644) Pricing. There was general consensus among respondents to the market test that
expressed a view that the public API access commitment was acceptable, except as
regards pricing. Currently there is no charge for API access (there is only a usage
charge for WhatsApp) and several substantiated responses expressed strong concerns
that Meta (formerly Facebook) could introduce a (so-called “non-discriminatory”)
price for all third parties but not for Kustomer, thereby favouring Kustomer. For
example, a CRM software provider that responded to the market test indicated that
“For now, all Facebook APIs are free to us. It means they could start making those
APIs paid for providers like us…”
920
while another CRM software provider
considered that “the API access should to be free.”
921
A competing provider of B2C
messaging services, despite not being a beneficiary of the commitments, echoed
these views, observing that “The pricing obligation is not to differentiate between
third parties... but they will obviously have the internal fee free advantage”
922
(645) While noting that it has [Business Plans], Meta (formerly Facebook) was equally not
able to identify or provide a specific example of a clear category of third party API
access seekers that could be used as a benchmark for monitoring of a commitment as
to non-discriminatory pricing vis-a-vis third party CS CRM providers.
923
916
Replies to questionnaire 8 (Commitments Market Test), question 19.
917
Reply to questionnaire 8 (Commitments Market Test), question 19.1 (ID: 2608).
918
Reply to questionnaire 8 (Commitments Market Test), question 19.1 (ID: 2436).
919
Reply to questionnaire 8 (Commitments Market Test), question 19.1 (ID: 2634).
920
Reply to questionnaire 8 (Commitments Market Test), question 20.1 (ID: 2415).
921
Reply to questionnaire 8 (Commitments Market Test), question 20.1 (ID: 2634).
922
Reply to questionnaire 8 (Commitments Market Test), question 6.1 (ID: 2608).
923
Parties’ reply to RFI 24, question 9(b): “Comparable users to Third Party CS CRM Providers could
include other business solution providers that access similar messaging functionalities via the Publicly
Available B2C Messaging Channel APIs and have similar messaging needs (e.g. volumes).”
172
(646) Core B2C Messaging Channel API Functionalities. Several of the more
substantiated responses expressed strong reservations about the narrow scope of the
definition of “core” functionalities, in particular the fact that:
(a) The definition is much narrower than the features currently covered by the
relevant public APIs; and
(b) The definition does not allow for new functionalities to be covered in the future
(only improvements to the existing “core” functionalities, which several
respondent pointed out are basic features
924
, and one CRM software provider
stated that they were “not sure if any additional improvements, visible or
otherwise, can be made against this basic feature set.”
925
).
(647) In parallel, Meta (formerly Facebook) indicated that overall there are currently […]
separate API features that make up (i) the public API for Messenger
(40 APIs/features) (ii) the public API for Instagram Messaging (16 APIs/features)
and (iii) the API for WhatsApp available through CPaaS providers ([…]
APIs/features). Meta (formerly Facebook) further confirmed that, out of this list, no
API/feature fell entirely within the definition of Core B2C Messaging Channel API
Functionalities, and only a small subset (overall […] APIs/features) are considered
“partially” “core”: Messenger (4 out of 40), Instagram (2 out of 16) and WhatsApp
([…]out of […]).
926
(648) In relation to the current definition, market participants, including many business
customers, indicated that access to such features is considered necessary to be able to
compete in the CS CRM market.
927
One such business customer expressed the view
that “all CS CRM systems should have the right to implement new features without
limitation.”
928
Another respondent to the market test indicated that they, i.e., business
“customers should be able to decide whether a function is relevant to the service they
offer [which] also may give ideas of improvement [i.e., to Facebook].”
929
By
implication, this means that CRM software provider would need to be able to offer a
full range of functionalities. Similarly, echoing these views, CPaaS providers, even
though they are not a direct beneficiary of the commitments, indicated that “all
available existing and future message types should be avaialble [sic] to [competing]
CS CRM”
930
software providers, including because the full range of functionalities
would be available to competitors (i.e., Kustomer) and it is moreover not possible to
anticipate all use cases of business customers: “most likely because competitors offer
such breadth of functionality and a provider cannot anticipate all customers’ use
cases”.
931
Finally, a close competitor of Kustomer indicated similarly that “as the
messaging channel functionalities increase in breadth, … A CRM … would need to
be able to offer support for the entirety of the functionality”
932
in order to effectively
compete in the customer service and support CRM software market.
(649) Further, respondents to the market test, including notably many business customers,
also indicated that new functionalities or features of Meta (formerly Facebook)’s
924
Replies to questionnaire 8 (Commitments Market Test), question 11.1.
925
Reply to questionnaire 8 (Commitments Market Test), question 11.1 (ID: 2543).
926
Parties’ reply to RFI 24, Annex 1.
927
Replies to questionnaire 8 (Commitments Market Test), question 9.
928
Reply to questionnaire 8 (Commitments Market Test), question 9 (ID: 2540).
929
Reply to questionnaire 8 (Commitments Market Test), question 9 (ID: 2540).
930
Reply to questionnaire 8 (Commitments Market Test), question 9 (ID: 2615).
931
Reply to questionnaire 8 (Commitments Market Test), question 9 (ID: 2593).
932
Reply to questionnaire 8 (Commitments Market Test), question 9 (ID: 2577).
173
B2C messaging channels that may be introduced in the future are likely to become
important for a customer service and support CRM software provider’s ability to
effectively compete, with several business customers indicating that they would
likely consider switching to a CRM software provider that could offer such
functionalities
933
: “If new functionalities that become standard appear in the next
coming 5 years and if they are not offered by our existing CS CRM provider, we
would consider to switch to a provider offering such functionality...”,
934
“it is very
likely that new functionalities or features of Meta’ B2C messaging channels will
become important to a CS CRM offering in the coming 5 years and that if a CS CRM
provider is not able to offer such feature, its customers consider changing CS CRM
provider”,
935
and “The most convenient feature will be preferred by the customer and
develop to an expected functionality over time. It can be estimated that companies
would consider switching to CS CRM provider that is able to offer these
functionalities or features to avoid unsatisfied customers.”
936
A close competitor of
Kustomer indicated that based on past trends, it is likely that new features and
functionalities will become important in the future and noted that business customers
evaluate the CRM offering not just based on availability of channels but also their
completeness in terms of functionalities: “There has been substantial new feature
and functionality introduction in Meta’s messaging channels over the past five years
and we believe it is likely that this product innovation will continue. Today,
businesses evaluate CS CRMs by not just the availability of the channels it offers but
also the completeness of each channel in terms of the functionality that is
supported.”
937
(650) Beta testing. A majority of market participants
938
, as well as a Member State that
provided comments, indicated that the provisions around beta testing require
improvement. The Parties provided confirmations that beta testing is normally done
with a small sub-set of a CRM provider’s customers
939
and the market test confirmed
that the limitations on the number of business customers of Kustomer that can be
involved in such testing was overly broad. In particular, the definition of “Broadly
Available” as 50% of Kustomer’s customers was considered to be too high, and that
this could be used to circumvent the commitments, for example since Kustomer
“could hold some dormant/very small customers to which they will not provide
access.”
940
Some respondents proposed instead using a small percentage, such as
“10%-15% of the total number of business customers”
941
(or lower), while others
considered it would be more appropriate to use a low absolute number, such as
“between 10 and 15 customers per CRM provider”
942
. The Commission further notes
that to date the highest number of Kustomer’s business customers were involved in a
beta test of a Meta (formerly Facebook) messaging channel was [Customer
933
Replies to questionnaire 8 (Commitments Market Test), question 10.1.
934
Reply to questionnaire 8 (Commitments Market Test), question 10.1 (ID: 2615).
935
Reply to questionnaire 8 (Commitments Market Test), question 10.1 (ID: 2565).
936
Reply to questionnaire 8 (Commitments Market Test), question 10.1 (ID: 2622).
937
Reply to questionnaire 8 (Commitments Market Test), question 10.1 (ID: 2577).
938
Replies to questionnaire 8 (Commitments Market Test), questions 12, 13 and 14.
939
See e.g., SO Response, Form RM and Parties’ reply to RFI 24.
940
Reply to questionnaire 8 (Commitments Market Test), question 11.1 (ID: 2608).
941
Reply to questionnaire 8 (Commitments Market Test), question 14 (ID: 2543).
942
Reply to questionnaire 8 (Commitments Market Test), question 14 (ID: 2540).
174
Information], which accounted for around [Customer Information] of Kustomer’s
[Customer Information] business customers at the time of notification.
943
(651) As regards reporting, one CRM software provider proposed that Kustomer should be
required to “report on all beta tests for each applicable API”
944
, rather than only
where such testing exceeds 12 months as provided for in the Initial Commitments.
Market test respondents moreover broadly confirmed that “12 months is a very long
period for testing”
945
and, [Business Plans], that “3-6 months is a generally accepted
timeframe to beta-test a feature with selected customers before making it generally
available.”
946
(652) Facebook and Instagram public comments. The market test sought respondents’
views the fact that Facebook and Instagram public comments are out of scope of the
Initial Commitments, which are limited to Meta (formerly Facebook)’s “private”
B2C messaging channels.
(653) A majority of respondents that expressed a view considered that Facebook and
Instagram public comments are OTT B2C messaging channels,
947
while the vast
majority of those with a view considered such public comments to be important
inputs to be able to compete in the customer service CRM software market.
948
This
was in particular the case because end customers may see Meta (formerly
Facebook)’s public and private messaging channels as an extension of one another,
and could for example leave a public comment expecting a follow-up exchange over
a private channel with customer service agents. As one major CRM provider
indicated “From a consumer standpoint, they may not see a difference between
Facebook and Facebook Messenger (i.e. they see it as “Facebook”). If the consumer
goes to the Facebook page of a business and leaves a comment that is later followed-
up privately by customer service agents on a messaging channel, then it is
reasonable to assume that the comments be an extension to OTT B2C messaging
channels”.
949
A business customer further indicated that it “uses both [public
comments and private messages] “such as on Facebook Messenger … for customer
relationship management as both are equally important and equally relevant for
CRM.”
950
Moreover, Facebook and Instagram public comments are integrated via
public APIs into Kustomer’s CRM software tool today.
951
(654) The results of the market test and responses provided by the Parties also revealed a
number of other areas where improvements to the Initial Commitments were
warranted.
952
943
Parties’ reply to RFI 24, Table 12.2.
944
Reply to questionnaire 8 (Commitments Market Test), question 15.1 (ID: 2634).
945
Reply to questionnaire 8 (Commitments Market Test), question 12.2 (ID: 2608).
946
Reply to questionnaire 8 (Commitments Market Test), question 12.2 (ID: 2543).
947
Replies to questionnaire 8 (Commitments Market Test), question 17.1.
948
Replies to questionnaire 8 (Commitments Market Test), question 17.2.
949
Reply to questionnaire 8 (Commitments Market Test), question 17.1.1 (ID: 2543).
950
Reply to questionnaire 8 (Commitments Market Test), question 17.1.1 (ID: 2565).
951
Parties’ reply to RFI 24, question 2(b).
952
While a small number of respondents to the market test raised a concern that Meta (formerly Facebook)
may use its Terms of Service or other eligibility criteria to obtain competitively sensitive information on
third party CS CRM providers, a greater number of respondents that expressed a view considered that
the information required was typically only technical information that was needed to be able to provide
and maintain API access. One major CRM software provider indicated that “by and large, [such
information tends to be] technical information [which] is generally most relevant to ensuring properly
functioning APIs.” See reply to questionnaire 8 (Commitments Market Test), question 7.1 (ID: 2509).
175
(655) First, new entrants, which the Commission identified as potential targets of
foreclosure, are not covered by the Initial Commitments, in view of the requirement
for Third Party CS CRM Provider to already have sales in the EEA.
953
(656) Second, while a majority of respondents that expressed a view considered that
broadly speaking the other definitions in the Initial Commitments were
appropriate
954
, market test respondents raised concerns in relation to certain specific
definitions. In particular, concerns were raised in relation to the definition of CS
CRM, which was defined in the Initial Commitments as “a software application used
by businesses for the primary purpose of collecting, storing, and organising
customer data for use by customer service agents to manage and respond to 1:1
customer service communications.” A major CRM software provider pointed out that
a customer service CRM “may not require a customer service agent”
955
while
another leading CRM software provider proposed “deleting […] “1:1” so that the
definition does not inadvertently exclude customer service communications that
reach more than one customer [and deleting] “primary” so that the definition does
not inadvertently exclude customer service CRM software that may be
multifunctional”.
956
(657) Third, as access to Meta (formerly Facebook)’s B2C messaging channel APIs is
subject to compliance with eligibility criteria and Terms of Service
957
and the Third
Party and App Assessment Process
958
, a number of market test respondents, in
particular competing customer service and support CRM software providers,
expressed concerns that Meta (formerly Facebook) may have the ability or incentive
to use such eligibility criteria or Terms of Service to place undue burdens on API
access seekers.
959
For example, one close competitor of Kustomer stated that “It is
possible that Meta could change the standard terms of service and assessment
processes such that it could inhibit or prohibit CS CRM providers from obtaining
access to these services. When ToS change (ie, when they change data access rights,
etc.) apps historically have needed to be re-reviewed and can lose access, and we do
not have significant visibility into that process”,
960
while a leading CRM software
provider pointed out that “Meta’s terms of use could hypothetically include (or could
This corroborates the Commission’s conclusion in Section 7.3.2.2.(a.iii)E above that Meta (formerly
Facebook) is unlikely to have the ability to obtain data on CRM competitors to provide a competitive
advantage to Kustomer.
953
Parties’ reply to RFI 24, question 4. Meta (formerly Facebook) clarified however that there no
minimum sales threshold, and that a competing customer service and support CRM software provider
“would be included in the definition of Third Party CS CRM Provider as soon as they generate a single
sale in the EEA, agree to the applicable Terms of Service and meet the Third Party and App Assessment
Process” (emphasis added).
954
Replies to questionnaire 8 (Commitments Market Test), question 18.
955
Reply to questionnaire 8 (Commitments Market Test), question 5.1 (ID: 2543).
956
Reply to questionnaire 8 (Commitments Market Test), question 7.2 (ID: 2577).
957
Defined in the Commitments as “the terms of service applicable to the use of the B2C Messaging
Channel APIs as may be updated from time to time, for instance to comply with legal requirements or
policy objectives as set out in paragraph 24 of these Commitments.”
958
Defined in the Commitments as “the review process by which Meta verifies (including on an ongoing
basis) that a third party’s use of its B2C Messaging channels complies with Meta’s applicable
policies.”
959
Replies to questionnaire 8 (Commitments Market Test), question 7.
960
Reply to questionnaire 8 (Commitments Market Test), question 7.2 (ID: 2577).
176
be later amended to include) the ability for Meta to selectively restrict or degrade
API access.”
961
(658) Regarding the fast track dispute resolution mechanism, the majority of the
respondents to the market test that expressed a view submitted that it will allow third
parties to enforce the Commitments. Only a limited number of respondents provided
a negative view on the mechanism
962
, indicating, for example, that it appeared to be
too complex or burdensome.
963
One respondent indicated that the procedure should
provide for an EEA-based arbitration venue, such as Paris.
964
(659) Subject to the point set out above in relation to reporting on beta testing, the majority
of the respondents to the market test submitted that the provisions regarding
monitoring are sufficient to ensure that the Initial Commitments are complied
with.
965
Several respondents indicated that there should also be reporting on any
“new API features and improvements”
966
which may be introduced, e.g., solely to
Kustomer.
(660) The majority of respondents to the market test that expressed a view also submitted
that the provisions of the Initial Commitments are capable of being implemented
effectively within a short period of time.
967
8.3.3. Commission’s assessment of the Initial Commitments
(661) The Commission assessed the appropriateness of the Initial Commitments in light of
the principles underlying its commitments policy and the results of the market test.
(662) The Commission concluded that, despite a number of positive elements, the Initial
Commitments fell short of completely removing the Commission’s competition
concerns as set out in the SO.
(663) In particular, the Commission identified the following shortcomings in the Initial
Commitments:
(664) First, the Commission considered that the five-year duration was insufficient, and
should be increased to at least 10 years, based on market test feedback in addition to
the overall policy consideration that access commitments should be “equivalent to
divestitures in their effects”.
968
(665) Second, on pricing, the Commission considered that there should be a clear
commitment to not introduce a charge for access to public APIs, in line with the
approach today and a confirmation that any usage or volume-based pricing be done
in a commercially reasonable way so as not to be applied with a view to
disadvantaging rivals of Kustomer.
(666) Third, the Commission considered that the definition of Core B2C Messaging
Channel API Functionalities should be substantially broadened, in particular since
the WhatsApp for Business API is not yet a public API (and therefore not covered by
the public API access commitment on day one following closing of the Transaction),
961
Reply to questionnaire 8 (Commitments Market Test), question 7.2 (ID: 2605).
962
Replies to questionnaire 8 (Commitments Market Test), question 16.
963
Reply to questionnaire 8 (Commitments Market Test), questions 16 and 16.1 (ID: 2634).
964
Reply to questionnaire 8 (Commitments Market Test), questions 16.1 (ID: 2534).
965
Replies to questionnaire 8 (Commitments Market Test), question 15.
966
See for example, reply to questionnaire 8 (Commitments Market Test), question 15.1 (ID: 2634).
967
Replies to questionnaire 8 (Commitments Market Test), question 21.
968
Remedies Notice, paragraph 17.
177
with a clear list of all functionalities that would be included from day one, and a fair
and workable mechanism to add new functionalities in the future. Effective
monitoring of the core API functionality-parity commitment would also require
timely reporting to the Monitoring Trustee on any new API, functionality or
improvement of a Meta (formerly Facebook) B2C messaging channel that is made
available in the future to Kustomer (e.g., via a private or non-public API) but not, or
not initially, to third party customer service and support CRM software providers.
(667) Fourth, on beta testing, the Commission considered that a clear upper limit should be
placed on any beta testing carried out with Kustomer, which should be substantially
lower than the proposal in the Initial Commitments of 50% of Kustomer’s customers.
Further, there should be periodic (e.g., quarterly) reporting to the monitoring trustee
on all beta testing done with Kustomer, and an explanation should be provided to the
monitoring trustee where any such testing exceeds 6 months.
(668) Fifth, the Commission considered that, in relation to the Facebook and Instagram
public comments, it would be necessary to either cover the relevant public APIs by
the commitments or, at a minimum, include an ability to read and import such public
comments into the conversation/interaction history in the customer service CRM
tool, and to respond to them over Meta (formerly Facebook)’s private OTT B2C
messaging channels.
(669) Sixth, the Commission considered that the definition of Third Party CS CRM
Providers should be broadened to include potential entrants in the customer service
and support CRM software market in the EEA.
(670) Seventh, the Commission considered that the definition of CS CRM should be
broadened to address the comments raised by competing customer service CRM
software providers, outlined in Section 8.3.2. above, that the definition was overly
restrictive (e.g., limited to 1-to-1 communications).
(671) Eighth, the Commission considered that the commitments should include language
that any changes to the applicable Terms of Service for accessing the relevant APIs
and API features would be consistent with the spirit of the commitments, i.e., would
not be used as a means to circumvent Meta (formerly Facebook)’s obligations under
the commitments.
(672) Ninth, the Commission considered that, as part of the fast-track dispute resolution
mechanism, an EEA-based arbitration forum should be added, at the option of the
API access seeker.
8.4. The Final Commitments
8.4.1. Description of the Final Commitments
(673) Following the feedback provided in the market test, the Notifying Party has
introduced a number of modifications in order to address the concerns raised by the
respondents.
(674) In particular, the improvements to the Initial Commitments contained in the Final
Commitment included the following:
(675) First, the duration has been significantly extended, from 5 to 10 years.
(676) Second, instead of a commitment not to discriminate between Third Party CS CRM
Providers and other comparable users in terms of charging for API access, as
included in the Initial Commitments, the Final Commitment include a clear
commitment that Meta (formerly Facebook) will not charge Third Party CS CRM
Providers for access to its Publicly Available B2C Messaging Channel APIs, in line
178
with the current situation. Meta (formerly Facebook) may apply usage-based charges
(e.g., by reference to volume or other relevant metric), as is currently the case for the
WhatsApp for Business API, provided any such charges are commercially
reasonable.
(677) Third, in place of the narrow list of Core B2C Messaging Channel API
Functionalities included in the Initial Commitments (i.e., the ability to send text-
based messages, images and URL links), Meta (formerly Facebook) has now
included a much broader list, in Annex 3 to the Final Commitments. Annex 3
includes all B2C Messaging Channel API Functionalities (i.e., all functionalities of
the Messenger, Instagram messaging and WhatsApp for business APIs) that are
currently integrated into Kustomer’s CRM software tool, irrespective of the extent to
which such functionalities are used by its business customers, in addition to features
that enable a business user of a CRM software tool to reply privately to a public
comment on the relevant business’s Facebook Page. The Final Commitments also
provide for a mechanism to add new Core B2C Messaging Channel API
Functionalities in the future, referred to as the “Usage Threshold”, whereby if a new
or improved functionality is integrated into the Kustomer Service and is used by
more than 25% of individual Kustomer Business Users via the Kustomer Service on
a weekly basis in the previous quarter,
969
the functionality will be designated as a
Core B2C Messaging Channel API Functionality.
970
In such a case, Meta (formerly
Facebook) will be required to make that functionality available to all Third Party CS
CRM Providers as soon as reasonably practicable, and in any case not later than
12 months after meeting the Usage Threshold. The Notifying Party explained that
such period is to allow that the relevant functionality can be scaled up from being
used only by Kustomer Business Users to a much wider set of business users, namely
those of all Third Party CS CRM Providers.
971
(678) Fourth, the Final Commitments provide that any API Development and Testing
conducted on the Kustomer Service can be done with up to 25% of Kustomer
Business Users, down from 50% in the Initial Commitments. Once a new API,
functionality or improvement is made available to more than 25% of Kustomer
Business Users, it then becomes subject to the Usage Threshold procedure, outlined
above, to determine if it is to be added to the list of add new Core B2C Messaging
Channel API Functionalities, and therefore made available to all Third Party CS
CRM Providers. Separately, while the Initial Commitments only required reporting
969
Meta (formerly Facebook) confirmed that the “Usage Threshold” may be met by a changeable subset of
Kustomer Business Users from one week to the next, i.e., that it does not need to be the same group of
Kustomer Business Users that use a new or improved functionality each week in a particular quarter.
See Parties’ reply to RFI 25 of 22 December 2021.
970
In justifying the weekly usage requirement, Meta (formerly Facebook) indicated that, in its view, it is
important that functionalities show consistency of usage to become Core. By contrast, a less frequent
usage requirement, e.g., “monthly” may bring in features that are only used once or twice in a month
and then not used again, which could therefore be unduly burdensome. Meta (formerly Facebook) also
pointed out that a weekly usage threshold is consistent with how Meta (formerly Facebook) often tracks
the performance of products/features internally to assess their popularity among users.
971
In the Form RM, footnote 7, Meta (formerly Facebook) explained that “Making those functionalities
available to all Third Party CS CRM Providers would result in exponentially greater messaging
volumes relative to Kustomer (especially given Kustomer’s small size in comparison to other Third
Party CS CRM Providers), and so a twelve month period is necessary to ensure that Meta can provide
the requisite support for the relevant APIs. Releasing an API will also require Meta to build and
implement new integrity and programmatic controls to ensure, for example, that the APIs could not be
accessed by bad actors. This may include hiring and training new employees.”
179
to the Monitoring Trustee of any beta tests that exceeded 12 months in duration, the
final Commitments provide for quarterly reporting to the monitoring trustee on any
beta tests, including the percentage and number of Kustomer Business Users using
the particular functionality being tested, and an explanation must also be provided if
any such testing exceeded six months (down from 12 months). In addition, there will
also be quarterly reporting on any B2C Messaging Channel API functionalities or
improvements of such functionalities made available to the Kustomer Service that
are not provided to Third Party CS CRM Providers (e.g., non-core functionalities).
(679) Fifth, the Final Commitments covers functionalities of the Meta (formerly
Facebook’s) B2C Messaging channels that enable businesses to read public
comments posted by customers on Facebook and Instagram pages and to respond to
them over Meta (formerly Facebook)’s private OTT B2C messaging channels. These
are covered not only under the public API access commitment, but also under the
core functionality parity commitment, through the addition of the “Private Reply”
features of Messenger and Instagram Messaging to the list of “core” functionalities
in Annex 3.
(680) Sixth, the Final Commitments cover not only third party CS CRM providers with
sales in the EEA (i.e., existing competitors of Kustomer) but also new entrants, that
actively target customers in the EEA (e.g., through entering into a contract to provide
CS CRM services in the EEA or establishing a local sales team in the EEA).
972
(681) Seventh, the Final Commitments address the concerns, outlined in Section 8.3.2.
above, that the initial definition of CS CRM was overly restrictive. For example, the
revised definition is no longer limited to 1-to-1 communications. Similarly, it is no
longer limited to CS CRM software applications used by “customer service agents”,
but also covers CS CRM software applications used by “customer service …
functions” of a business more generally.
973
(682) Eighth, in order to address concerns raised in the market test that Meta (formerly
Facebook) may seek to circumvent the commitments through changes to its Terms of
Service, the Final Commitments now clarify that it will not make any such changes
that would be equivalent to a breach of paragraph 1 (i.e., the public API access
commitment) or paragraph 2 (i.e., the core functionality parity commitment) of the
commitments.
(683) Ninth, and finally, as regards the fast-track dispute resolution, an arbitration venue
based in the EEA, namely Dublin, was added at the option of the relevant third party
CRM software provider.
8.4.2. Commission’s assessment of the Final Commitments
(684) The Commission concludes that the Final Commitments are capable of eliminating
the Commission’s competition concerns.
972
Similarly to the absence of any minimum sales threshold (where the Parties clarified that a “single sale
in the EEA” is sufficient, see Parties’ reply to RFI 24, question 4) the Final Commitments make no
mention of a minimum threshold as regards actively targeting customers in the EEA.
973
Moreover, Meta (formerly Facebook) specifically confirmed, in a comment to a draft of the
Commitments submitted on 16 December 2021 that the CS CRM software providers that responded to
the market test would be covered by the definition (and therefore by the Final Commitments): “We
understand that the following CS CRMs participated in the market test and would be included in the
definition - each of these refer to "agents" on their websites: Microsoft; Zendesk; Oracle; SAP;
Salesforce; Hubspot; Pegasystems; Crisp; Sprinklr.”
180
(685) First, the 10-year duration addressed the concerns raised in the market test and by a
Member State and, in line with the Remedies Notice, means that the commitments
may be considered, in this case, to be “equivalent to divestitures in their effects”.
974
(686) Second, the commitment that there will be no charge for access to public APIs
addresses clear concerns raised in the market test, and avoid a means of
circumvention whereby Meta (formerly Facebook) could have favoured Kustomer by
introducing API access fees for rivals of Kustomer.
(687) Third, the fact that (i) the list of Core B2C Messaging Channel API Functionalities
includes all functionalities integrated by Kustomer today
975
, and (ii) there is a clear
yet balanced mechanism to add new core functionalities in the future addresses the
concerns raised in the market test while being cognisant of the potential technical and
resource-related constraints that may be involved in rolling out a new API or API
feature at scale.
(688) Fourth, the reduction of the maximum percentage of Kustomer’s customers that may
be involved in beta testing to 25% (from 50%), together with enhanced reporting to
the Monitoring Trustee, sufficiently reduces the risk that Meta (formerly) Facebook
may circumvent the Commitments by rolling out important new features exclusively
to Kustomer, and making them available to a large number of Kustomer’s customers
for an extended period, under the guise of beta testing those features.
(689) Fifth, the finding from the market test that the ability to respond to public comments
on Facebook or Instagram pages is important for customer support has now also been
addressed. While the public APIs for Facebook and Instagram public comments have
not been included in the Final Commitments, features of the Messenger and
Instagram messaging APIs that allow for private replies to such public comments are
now covered by both the public API access commitment and the core functionality
parity commitment, which the Commission considers is sufficient to address the
concerns raised in the market test.
(690) Sixth, the definition of Third Party CS CRM Provider now includes new entrants that
target customers in the EEA, which ensures that the Final Commitments address the
concern identified by the Commission that such new entrants could be targeted by an
input foreclosure strategy. The definition of CS CRM has also been broadened
sufficiently to address the main concerns raised by the market test that it may have
been narrowly defined, which provides further comfort that that those customer
service and support CRM software providers that responded to the market test will be
covered by the Final Commitments.
(691) Seventh, the language in the Final Commitments that Meta (formerly Facebook) will
not change the applicable Terms of Service in a way that would breach the
commitments addresses, in the Commission’s view, the concerns raised by certain
respondents to the market test about potential circumvention of the Commitments in
this manner.
(692) Eighth, the addition of an arbitration venue based in the EEA, namely Dublin,
addresses a concern raised by one respondent to the market test about enforceability
974
Remedies Notice, paragraph 17.
975
This list is likely to be sufficient, at least on day one following closing of the Transaction to enable
rivals of Kustomer to effectively compete, in particular because one of the key reasons why Kustomer
was “attractive to Facebook [as an acquisition target in the first place] was [Meta strategic
considerations for acquiring Kustomer]. Meta (formerly Facebook)’s internal document, […].
181
in the EEA, as well as a broader policy consideration for having an EEA-based
dispute resolution forum in any commitments that may be accepted by the
Commission.
(693) Moreover, the Final Commitments are capable of being implemented effectively
within a short period of time and are proportionate to the competition concerns
identified by the Commission.
9. CONCLUSION
(694) In the light of the above, the Commission considers the Final Commitments capable
of rendering the Transaction compatible with the internal market and the EEA
Agreement as it will not create a significant impediment to effective competition in
all relevant markets in which competition concerns were identified, in particular on
the EEA-wide, if not worldwide, downstream market for customer service and
support CRM software. Although the Commission only identified competition
concerns on the narrower market for customer service and support CRM software,
the Final Commitments also address any concerns on the broader potential
downstream market for CRM software overall. Similarly, and for the same reasons as
outlined in Section 7.3.2.2.(a.i) above, the Commission’s conclusion that the Final
Commitments are capable of eliminating the competition concerns identified would
not change on the basis of the potential segments of the customer service and support
CRM software market, or the overall CRM software market, based on (i) business
customer size, (ii) mode of deployment or (iii) the industry sector of the business
customer.
10. CONDITIONS AND OBLIGATIONS
(695) Pursuant to the second subparagraph of Article 8(2) of the Merger Regulation, the
Commission may attach to its decision conditions and obligations intended to ensure
that the undertakings concerned comply with the commitments they have entered
into vis-à-vis the Commission with a view to rendering the concentration compatible
with the internal market.
(696) The fulfilment of the measure that gives rise to the structural change of the market is
a condition, whereas the implementing steps which are necessary to achieve this
result are generally obligations on the parties. Where a condition is not fulfilled, the
Commission’s decision declaring the concentration compatible with the internal
market is no longer applicable. Where the undertakings concerned commit a breach
of an obligation, the Commission may revoke the clearance decision in accordance
with Article 8(6) of the Merger Regulation. The undertakings concerned may also be
subject to fines and periodic penalty payments under Articles 14(2) and 15(1) of the
Merger Regulation.
(697) In accordance with the basic distinction described in the preceding Recital as regards
conditions and obligations, all Sections (including Annexes 1 to 4) of the Final
Commitments should be obligations within the meaning of Article 8(2) of the Merger
Regulation. The full text of the commitments is attached as a Schedule to this
Decision and forms an integral part thereof.
HAS ADOPTED THIS DECISION:
182
Article 1
The notified operation whereby Meta Platforms, Inc. acquires sole control of Kustomer, Inc.
within the meaning of Article 3(1)(b) of the Merger Regulation is hereby declared compatible
with the internal market and the EEA Agreement.
Article 2
Meta Platforms, Inc. shall comply with the obligations set out in the Schedule to this Decision.
Article 3
This Decision is addressed to:
Meta Platforms, Inc.
601 Willow Road, Menlo Park
California 94025
United States of America
Done at Brussels, 27.1.2022
For the Commission
(Signed)
Margrethe VESTAGER
Member of the Commission
Case No. M.10262 – META / KUSTOMER
COMMITMENTS TO THE EUROPEAN COMMISSION
Pursuant to Article 8(2) and 10(2) of Council Regulation (EC) No 139/2004 (the “Merger
Regulation”), Meta Platforms, Inc. hereby enters into the following Commitments (the
“Commitments”) vis-à-vis the European Commission (the “Commission”) with a view to
rendering its proposed acquisition of Kustomer, Inc. (the “Concentration”) compatible with the
internal market and the functioning of the EEA Agreement.
This text shall be interpreted in light of the Commission’s decision pursuant to Article 8(2) of the
Merger Regulation to declare the Concentration compatible with the internal market and the
functioning of the EEA Agreement (the “Decision”), in the general framework of European Union
law, in particular in light of the Merger Regulation, and by reference to the Commission Notice on
remedies acceptable under the Merger Regulation and under Commission Regulation (EC)
No 802/2004.
Section A. Commitments
1. Meta commits from the Closing Date to make the Publicly Available B2C Messaging
Channel APIs available to Third Party CS CRM Providers without charging for access and
on a non-discriminatory basis, meaning without differentiating between Third Party CS
CRM Providers and other comparable users in terms of eligibility criteria, functionality, and
performance. Meta also commits from the Closing Date not to discriminate against Third
Party CS CRM Providers in relation to denying or unreasonably delaying access to,
changing, suspending or deprecating Publicly Available B2C Messaging Channel APIs.
For the avoidance of doubt, Meta may apply commercially reasonable charges based on
usage of the Publicly Available B2C Messaging Channel APIs (by reference to volume or
any other relevant metric).
2. Meta further commits from the Closing Date to make equivalent Core B2C Messaging
Channel API Functionalities available to Third Party CS CRM Providers as to the
Kustomer Service on a non-discriminatory basis, meaning without materially differentiating
in terms of functionality and performance (including rate limits) depending on whether the
Core B2C Messaging Channel API Functionalities are accessed by a Third Party CS CRM
Provider or the Kustomer Service. For the avoidance of doubt, Meta also commits from
the Closing Date not to discriminate against Third Party CS CRM Providers relative to the
Kustomer Service in relation to denying or unreasonably delaying access to, changing,
suspending or deprecating Core B2C Messaging Channel API Functionalities.
3. Meta shall publish on Meta’s Developer Website, and update on a quarterly basis, a list
of:
a. Publicly Available B2C Messaging Channel API(s); and
b. Core B2C Messaging Channel API Functionalities made available to the Kustomer
Service (other than those made available to the Kustomer Service only through
Publicly Available B2C Messaging Channel APIs).
4. Meta shall be permitted under these Commitments to:
a. Conduct API Development and Testing solely on the Kustomer Service, with
Kustomer Business Users, and/or with certain third parties.
b. Develop features associated with Meta Services for use solely with the Kustomer
Service (or any other Meta Service, or as part of a partnership with a third party),
including features that complement (but do not in themselves constitute features
of) the Publicly Available B2C Messaging Channel APIs and/or Core B2C
Messaging Channel API Functionalities, provided that the development and
implementation of such features shall not impede the functioning of and access to
functionalities of Publicly Available B2C Messaging Channel APIs or Core B2C
Messaging Channel API Functionalities. Nothing in these Commitments shall
require Meta to license versions of Meta Services other than the Publicly Available
B2C Messaging Channel APIs and Core B2C Messaging Channel API
Functionalities (in each case in accordance with the terms of these Commitments).
For the avoidance of doubt, any Third Party CS CRM Provider shall remain free
similarly to develop and make available features associated with their own first-
party products, apps, or services.
5. Meta shall not be in breach of these Commitments in the event of any degradation or non-
availability of the Publicly Available B2C Messaging Channel APIs or Core B2C
Messaging Channel API Functionalities that is solely or primarily attributable to any acts
or omissions of a third party or other events outside of Meta’s reasonable control (including
unplanned service outages and downtime).
Section B. Monitoring Trustee
B. 1 Appointment procedure
6. No later than the Closing Date, Meta shall appoint a Monitoring Trustee to carry out the
functions specified in these Commitments for a Monitoring Trustee.
7. The Monitoring Trustee shall:
a. At the time of appointment, be independent of the Parties;
b. Possess the necessary experience, competence, and qualifications to carry out its
mandate, including via any technical expert appointed pursuant to paragraph 17
below; and
c. Neither have, nor become exposed to, a Conflict of Interest.
8. The Monitoring Trustee shall be remunerated by Meta in a way that does not impede the
independent and effective fulfillment of its mandate.
9. Proposal by Meta. No later than four weeks after the Effective Date, Meta shall submit
the name or names of one or more natural or legal persons whom Meta proposes to
appoint as the Monitoring Trustee to the Commission for approval. The proposal shall
contain sufficient information for the Commission to verify that the person or persons
proposed as Monitoring Trustee fulfil the requirements set out in paragraph 7 and shall
include:
a. The full terms of the proposed mandate, which shall include all provisions
necessary to enable the Monitoring Trustee to fulfil its duties under these
Commitments; and
b. The outline of a work plan that describes how the Monitoring Trustee would carry
out its duties under these Commitments.
10. Approval or rejection by the Commission. The Commission shall have the discretion to
approve or reject the proposed Monitoring Trustee(s) and to approve the proposed
mandate subject to any modifications it deems necessary for the Monitoring Trustee to
fulfil its obligations. If only one name is approved, Meta shall appoint or cause to be
appointed the person or persons concerned as Monitoring Trustee, in accordance with the
mandate approved by the Commission. If more than one name is approved, Meta shall be
free to choose the Monitoring Trustee to be appointed from among the names approved.
The Monitoring Trustee shall be appointed within one week of the Commission's approval,
in accordance with the mandate approved by the Commission.
11. New proposal by Meta. If all the proposed Monitoring Trustees are rejected, Meta shall
submit the names of at least two more natural or legal persons within one week of being
informed of the rejection, in accordance with paragraphs 9 and 10 of these Commitments.
12. Monitoring Trustee nominated by the Commission. If all further proposed Monitoring
Trustees are rejected by the Commission, the Commission shall nominate a Monitoring
Trustee, whom Meta shall appoint, or cause to be appointed, in accordance with a
Monitoring Trustee mandate approved by the Commission.
B. 2 Functions of the Monitoring Trustee
13. The Monitoring Trustee shall assume its specified duties and obligations in order to ensure
compliance with the Commitments. The Commission may, on its own initiative or at the
request of the Monitoring Trustee or Meta, give any orders or instructions to the Monitoring
Trustee in order to ensure compliance with the conditions and obligations attached to the
Decision.
14. The Monitoring Trustee shall:
a. Monitor the performance of the Commitments by Meta;
b. Propose in its first report to the Commission a detailed work plan describing how
it intends to monitor compliance with the obligations and conditions attached to the
Decision;
c. Provide to the Commission a written report, sending Meta a non-confidential copy
at the same time, within fifteen working days after the end of each six-month period
so that the Commission can assess whether the Commitments are being complied
with;
d. Propose, as applicable, to Meta such measures as the Monitoring Trustee
considers necessary to ensure Meta’s compliance with the Commitments;
e. Promptly report in writing to the Commission, sending Meta a non-confidential copy
at the same time, if it concludes on reasonable grounds that Meta has failed to
comply with the Commitments;
f. Act as a contact point for questions from third parties about the nature and scope
of the Commitments; and
g. Assume the other functions assigned to the Monitoring Trustee under the
conditions and obligations attached to the Decision.
B. 3 Duties and obligations of the Parties
15. Meta shall provide and shall cause its advisors to provide the Monitoring Trustee with all
such cooperation, assistance and information as the Monitoring Trustee may reasonably
require to perform its tasks (including the information set out in Annex 2). The Monitoring
Trustee shall have full and complete access to any of Meta’s books, records, documents,
management or other personnel, facilities, sites and technical information reasonably
necessary for fulfilling its duties under the Commitments and Meta shall provide the
Monitoring Trustee upon request with copies of any documents.
16. Meta shall indemnify the Monitoring Trustee and its employees and agents (each an
“Indemnified Party”) and hold each Indemnified Party harmless against, and hereby
agrees that an Indemnified Party shall have no liability to Meta for, any liabilities arising
out of the performance of the Monitoring Trustee’s duties under the Commitments, except
to the extent that such liabilities result from the willful default, recklessness, gross
negligence or bad faith of the Monitoring Trustee, its employees, agents or advisors.
17. At the expense of Meta, the Monitoring Trustee shall appoint a technical expert and may
appoint other advisors, subject to Meta’s approval (this approval not to be unreasonably
withheld or delayed) if the Monitoring Trustee considers the appointment of such advisors
necessary or appropriate for the performance of its duties and obligations under the
mandate, provided that any fees and other expenses incurred by the Monitoring Trustee
are reasonable. Should Meta refuse to approve the technical expert and/or advisors
proposed by the Monitoring Trustee the Commission may approve the appointment of
such technical expert and/or advisors instead, after having heard Meta. Only the
Monitoring Trustee shall be entitled to issue instructions to the technical expert and/or
advisors. Paragraph 16 of these Commitments shall apply mutatis mutandis.
18. Meta agrees that the Commission may share Confidential Information proprietary to Meta
with the Monitoring Trustee. The Monitoring Trustee shall not disclose Confidential
Information received from the Commission, Meta, or Kustomer to any third party other
than the Commission and a technical expert appointed pursuant to paragraph 17. The
principles contained in Article 17(1) and (2) of the Merger Regulation apply mutatis
mutandis.
19. Meta agrees that the contact details of the Monitoring Trustee shall be published on the
website of the Commission's Directorate-General for Competition and they shall inform
interested third parties of the identity and the tasks of the Monitoring Trustee.
B. 4 Replacement, discharge and reappointment of the Monitoring Trustee
20. If the Monitoring Trustee ceases to perform its functions under the Commitments or for
any other good cause, including the exposure of the Monitoring Trustee to a Conflict of
Interest:
a. The Commission may, after hearing the Monitoring Trustee and Meta, require Meta
to replace the Monitoring Trustee; or
b. Meta may, with the prior approval of the Commission, replace the Monitoring
Trustee.
21. If the Monitoring Trustee is removed according to paragraph 20 of these Commitments,
the Monitoring Trustee may be required to continue its function until a new Monitoring
Trustee is in place to whom the Monitoring Trustee has affected a full hand over of all the
relevant information. The new Monitoring Trustee shall be appointed in accordance with
the procedure referred to in paragraphs 7 to 12 of these Commitments.
22. Unless removed according to paragraph 20 of these Commitments, the Monitoring Trustee
shall cease to act as Monitoring Trustee only after the Commission has discharged it from
its duties after all the Commitments with which the Monitoring Trustee has been entrusted
have been implemented.
Section C. Dispute Resolution
23. In the event that a Third Party CS CRM Provider, showing a sufficient legitimate interest,
claims that Meta is failing to comply with its obligations arising from the Commitments, the
Fast-Track Dispute Resolution Procedure described in Annex 1 shall apply.
Section D. General Provisions
24. Notwithstanding any other provision of these Commitments, Meta shall be permitted under
these Commitments to take steps (including, for the avoidance of doubt, by amending the
Terms of Service and any relevant Meta or Kustomer policies) to: (i) comply with any
applicable law, regulation, legal process, or governmental request; (ii) detect, prevent,
promote, or otherwise address fraud, security, integrity, or technical issues; (iii) improve
user or platform privacy or security; (iv) enforce the relevant Terms of Service or policies;
or (v) protect against harm to the rights, property, or safety of Meta, Meta users, or the
public as required or permitted by law.
25. Subject to paragraph 24, Meta shall not make changes to the Terms of Service in a
manner that would be equivalent to a breach of paragraph 1 or paragraph 2 of these
Commitments.
26. The Commitments shall take effect upon the Effective Date.
27. The Commitments shall remain in effect for ten years from the Closing Date.
Section E. Review
28. The Commission may in response to a reasoned request from Meta showing good cause,
waive, modify, or substitute, in exceptional circumstances, one or more of the
undertakings in these Commitments. This request shall be accompanied by a report from
the Monitoring Trustee, who shall at the same time send a non-confidential copy of the
report to Meta. The request shall not have the effect of suspending the application of the
undertaking and, in particular, of suspending the expiry of any time period in which the
undertaking has to be complied with.
Section F. Definitions
29. For the purpose of the Commitments, the following terms shall have the following
meaning:
Affiliated Undertakings: undertakings controlled by the Parties and/or by the ultimate
parents of the Parties, whereby the notion of control shall be interpreted pursuant to
Article 3 of the Merger Regulation and in light of the Commission Consolidated
Jurisdictional Notice under Council Regulation (EC) No 139/2004 on the control of
concentrations between undertakings.
API Development and Testing: the process by which Meta develops and tests new B2C
Messaging Channel APIs (or new functionalities or improvements), subject in each case
to the procedure in Annex 4. For the avoidance of doubt, API Development and Testing
may include Meta’s use of the Kustomer Service or certain third party developers and/or
products to develop and test new B2C Messaging Channel APIs (or new functionalities or
improvements).
B2C Messaging: the over the top (OTT) messaging services:
(a) offered through WhatsApp Business Platform, Messenger Platform, and Instagram
Messaging that provide businesses the means to receive and send private
messages from and to B2C Messaging Channel Users at scale; and/or
(b) that provide businesses the means to read and respond via private message to
public comments from B2C Messaging Channel Users on Facebook and
Instagram posts at scale.
B2C Messaging Channel API(s): APIs (including any future APIs) that Meta offers which
are necessary to implement B2C Messaging (other than in the context of API Development
and Testing), access to which is subject to the applicable eligibility criteria and Terms of
Service and the Third Party and App Assessment Process.
B2C Messaging Channel Users: individual (non-business) users of the consumer
versions of WhatsApp, Messenger, Facebook and Instagram.
Closing Date: closing date of the Concentration.
Confidential Information: any business secrets, know-how, commercial information, or
any other information of a proprietary nature that is not in the public domain.
Conflict of Interest: any conflict of interest that impairs the Monitoring Trustee's
objectivity and independence in discharging its duties under the Commitments.
Core B2C Messaging Channel API Functionalities: (i) the functionalities of the B2C
Messaging Channel APIs listed at Annex 3; (ii) functionalities of B2C Messaging Channel
APIs accessed by the Kustomer Service after the Effective Date as set out in Annex 4;
and (iii) in each case, any improvements of those functionalities accessed by the Kustomer
Service as set out in Annex 4. Core B2C Messaging Channel API Functionalities shall not
include any functionalities which are primarily associated with other Meta Services (e.g.
where the B2C Messaging Channel Users view or access another Meta Service via a
distinct interface such as an app switch, modal, popup, plugin or webview).
CS CRM: a software application used by businesses for the purpose of collecting,
storing, and organising customer data for use by customer service agents or functions
to manage and respond to customer service communications.
EEA: the 27 Member States of the European Union (Austria, Belgium, Bulgaria, Croatia,
Cyprus, Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hungary,
Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta, Netherlands, Poland, Portugal,
Romania, Slovakia, Slovenia, Spain and Sweden), Iceland, Liechtenstein and Norway.
Effective Date: the date of adoption of the Decision.
Instagram Messaging: Instagram messaging on Messenger Platform, currently
described at https://developers.Meta.com/docs/messenger-platform/instagram
.
Kustomer: Kustomer, Inc. or any Affiliated Undertakings of Kustomer, Inc, or any
successor entities or businesses (irrespective of branding).
Kustomer Service: any CS CRM product or service provided by Kustomer at any time
during the term of the Commitments.
Kustomer Business User: Any business customer of the Kustomer Service that
subscribes to the Kustomer Terms of Service.
Kustomer Terms of Service: the terms of service applicable to the use of the Kustomer
Service.
Monitoring Trustee: one or more natural or legal persons who are approved by the
Commission and appointed by Meta, and who have the duty to carry out the functions
specified in the Commitments for a Monitoring Trustee.
Messenger Platform: Messenger Platform, currently described at
https://developers.facebook.com/products/messenger/
.
Meta: Meta Platforms, Inc. or any Affiliated Undertakings of Meta Platforms, Inc.
Meta’s Developer Website: The version of https://developers.facebook.com/
available
without logging in, and any successor site.
Meta Services: any product or service (including algorithms) operated by Meta at any
time during the term of the Commitments. For the avoidance of doubt, Meta Services
include any product or service Meta may develop in the future, including via acquisitions,
regardless of its branding.
Parties: Meta and Kustomer.
Publicly Available B2C Messaging Channel API(s): B2C Messaging Channel APIs that
are publicly available on Meta’s Developer Website as may be updated from time to time,
excluding any APIs made available for API Development and Testing.
Terms of Service: the terms of service applicable to the use of the B2C Messaging
Channel APIs as may be updated from time to time, for instance to comply with legal
requirements or policy objectives as set out in paragraph 24 of these Commitments.
Third Party and App Assessment Process: the review process by which Meta verifies
(including on an ongoing basis) that a third party’s use of its B2C Messaging channels
complies with Meta’s applicable policies.
Third Party CS CRM Provider: Provider, other than Kustomer (and for the avoidance of
doubt other than Meta), of a CS CRM which has sales in the EEA, or which actively targets
customers in the EEA (e.g. through entering into a contract to provide CS CRM services
in the EEA or establishing a local sales team in the EEA) and that meets Meta’s Third
Party and App Assessment Process and agrees to the applicable Terms of Service.
References in these Commitments to Third Party CS CRM Providers shall (i) apply only
to their CS CRM activities; (ii) apply to both direct and indirect access by Third Party CS
CRM Providers to B2C Messaging Channel APIs; and (iii) include the arrangements by
which business users of that Third Party CS CRM Provider access B2C Messaging via
that Third Party CS CRM Provider.
WhatsApp Business Platform: The WhatsApp Business Platform, currently described
at https://www.whatsapp.com/business/api
.
duly authorised for and on behalf of Meta Platforms, Inc.
Signature Page for Case No. M.10262 - Meta / Kustomer Commitments to the European
Commission
[signed]
Annex 1 - Fast Track Dispute Resolution Procedure
1. A Third Party CS CRM Provider showing a sufficient legitimate interest that wishes to avail
itself of the fast track dispute resolution procedure (the "Requesting Party") shall inform
Meta and the Monitoring Trustee in writing, setting out in detail the reasons why the
Requesting Party believes that Meta is failing to comply with the Commitments. The
Requesting Party and Meta will use commercially reasonable efforts to settle all disputes
that may arise through cooperation and consultation within a reasonable period of time
not exceeding fifteen working days (such period being extendable by mutual consent of
Meta and the Requesting Party) (the “Consultation Period”) after receipt of the request.
2. The Monitoring Trustee shall present its own proposal (the “Trustee Proposal”) for
resolving the dispute within ten working days of the request to Meta, the Requesting Party
and the Commission, specifying in writing the rationale for the Trustee Proposal, and the
action, if any, to be taken by Meta in order to ensure compliance with Commitments vis-
à-vis the Requesting Party.
3. Should the Requesting Party and Meta (together, the “Parties to the Arbitration”) fail to
resolve their dispute in the Consultation Period, then the Requesting Party may, within
20 working days after the Consultation Period has expired, serve a notice (the “Notice”),
in the sense of a request for arbitration, to the International Chamber of Commerce (the
“Arbitral Institution”), with a copy of such Notice and request for arbitration to Meta.
4. The Notice shall set out in detail the dispute, difference or claim (the “Dispute”) and shall
contain, inter alia, all issues of both fact and law, including any suggestions as to the
procedure, and all documents relied upon shall be attached, e.g., documents, agreements,
expert reports, and witness statements. The Notice shall also contain a detailed
description of theTrustee Proposal, including a comment as to its appropriateness.
5. Meta shall, within ten working days from receipt of the Notice, submit its answer (the
“Answer”), which shall provide detailed reasons for its conduct and set out, inter alia, all
issues of both fact and law, including any suggestions as to the procedure, and all
documents relied upon, e.g., documents, agreements, expert reports, and witness
statements. The Answer shall, if appropriate, contain a detailed description of the action
which Meta proposes to undertake vis-à-vis the Requesting Party and a detailed
description of the Trustee Proposal (if not already submitted), including a comment as to
its appropriateness.
Appointment of the Arbitrators
6. The Arbitral Tribunal shall consist of three persons. The Requesting Party shall nominate
its arbitrator in the Notice; Meta shall nominate its arbitrator in the Answer. The arbitrator
nominated by the Requesting Party and by Meta shall, within five working days of the
nomination of the latter, nominate the chair, making such nomination known to the parties
and the Arbitral Institution which shall forthwith confirm the appointment of all three
arbitrators.
7. Should the Requesting Party wish to have the Dispute decided by a sole arbitrator it shall
indicate this in the Notice. In this case, the Requesting Party and Meta shall agree on the
nomination of a sole arbitrator within five working days from the communication of the
Answer, communicating this to the Arbitral Institution.
8. Should Meta fail to nominate an arbitrator, or if the two arbitrators fail to agree on the chair,
or should the Parties to the Arbitration fail to agree on a sole arbitrator, the default
appointment(s) shall be made by the Arbitral Institution.
9. The three-person arbitral tribunal or, as the case may be, the sole arbitrator, are herein
referred to as the “Arbitral Tribunal”.
Arbitration Procedure
10. The Dispute shall be finally resolved by arbitration under the International Chamber of
Commerce Rules of Arbitration, with such modifications or adaptations as foreseen herein
or necessary under the circumstances (the “Rules”). The arbitration shall be conducted
in New York, New York, United States of America (or, at the option of the Requesting
Party, Dublin, Ireland), in the English language. For good cause, any Party may apply to
the Arbitral Institution (or Arbitral Tribunal as may be appropriate) for an extension of the
timelines provided in this Annex.
11. The procedure shall be a fast-track procedure. For this purpose, the Arbitral Tribunal shall
shorten all applicable procedural time-limits under the Rules as far as admissible and
appropriate in the circumstances. The Parties to the Arbitration shall consent to the use
of email for the exchange of documents.
12. The Arbitral Tribunal shall, as soon as practical after the confirmation of the Arbitral
Tribunal, hold an organizational conference to discuss any procedural issues with the
Parties to the Arbitration. Terms of Reference shall be drawn up and signed by the Parties
to the Arbitration and the Arbitral Tribunal at the organizational meeting or thereafter and
a procedural time-table shall be established by the Arbitral Tribunal. An oral hearing shall,
as a rule, be established within two months of the confirmation of the Arbitral Tribunal.
13. In order to enable the Arbitral Tribunal to reach a decision, it shall be entitled to request
any relevant information from the Parties to the Arbitration, to appoint experts and to
examine them at the bearing, and to establish the facts by all appropriate means. The
Arbitral Tribunal is also entitled to ask for assistance by the Monitoring Trustee in all stages
of the procedure if the Parties to the Arbitration agree.
14. The Arbitral Tribunal shall not disclose confidential information and apply the standards
attributable to confidential information under the Merger Regulation. The Arbitral Tribunal
may take the measures necessary for protecting confidential information in particular by
restricting access to confidential information to the Arbitral Tribunal, the Monitoring
Trustee, and outside counsel and experts of the opposing party.
15. The burden of proof in any dispute under these Rules shall be borne as follows: (i) the
Requesting Party must produce evidence of a prima facie case; and (ii) if the Requesting
Party produces evidence of a prima facie case, the Arbitral Tribunal must find in favor of
the Requesting Party unless Meta can produce evidence to the contrary.
Involvement of the Commission
16. The Commission shall be allowed and enabled to participate in all stages of the procedure
by
a. receiving all written submissions (including documents and reports, etc.) made by
the Parties to the Arbitration;
b. receiving all orders, interim and final awards and other documents exchanged by
the Arbitral Tribunal with the Parties to the Arbitration (including Terms of
Reference and procedural time-table);
c. filing amicus curiae briefs; and
d. being present at the hearing(s) and with the permission of the Arbitral Tribunal, it
may also make oral observations.
17. The Arbitral Tribunal shall forward, or shall order the Parties to the Arbitration to forward,
the documents mentioned to the Commission without delay.
18. In the event of disagreement between the Parties to the Arbitration regarding the
interpretation of the Commitments, the Arbitral Tribunal may seek the Commission's
interpretation of the Commitments before finding in favour of any Party to the Arbitration
and shall be bound by the interpretation.
Decisions of the Arbitral Tribunal
19. The Arbitral Tribunal shall decide the dispute on the basis of the Commitments and the
Decision. Issues not covered by the Commitments and the Decision shall be decided (in
the order as stated) by reference to the Merger Regulation, EU law and general principles
of law common to the legal orders of the Member States without a requirement to apply a
particular national system. The Arbitral Tribunal shall take all decisions by majority vote.
20. Upon request of the Requesting Party, the Arbitral Tribunal may make a preliminary ruling
on the Dispute. The preliminary ruling shall be rendered within one month after the
confirmation of the Arbitral Tribunal, shall be applicable immediately and, as a rule, remain
in force until a final decision is rendered.
21. The Arbitral Tribunal shall, in the preliminary ruling as well as in the final award, specify
the action, if any, to be taken by Meta in order to comply with the Commitments vis-à-vis
the Requesting Party (e.g., specify that the Requesting Party gain access to the relevant
Publicly Available B2C Messaging Channel API or Core B2C Messaging Channel API
Functionality). The final award shall be final and binding on the Parties to the Arbitration
and shall resolve the Dispute and determine any and all claims, motions or requests
submitted to the Arbitral Tribunal. The arbitral award shall also determine the
reimbursement of the costs of the successful party and the allocation of the arbitration
costs. In case of granting a preliminary ruling or if otherwise appropriate, the Arbitral
Tribunal shall specify that terms and conditions determined in the final award apply
retroactively.
22. The final award shall, as a rule, be rendered within six months after the confirmation of
the Arbitral Tribunal. The time-frame shall, in any case, be extended by the time the
Commission takes to submit an interpretation of the Commitments if asked by the Arbitral
Tribunal.
23. The Parties to the Arbitration shall prepare a non-confidential version of the final award,
without business secrets. The Commission may publish the non-confidential version of
the award. The Parties to the Arbitration, the Arbitral Tribunal, all other persons
participating in the proceedings and all further persons involved, i.e. in the administration
of the arbitral proceedings, shall maintain confidentiality towards all persons regarding the
conduct of arbitral proceedings. All proceedings will be held in private and remain
confidential.
24. Nothing in the arbitration procedure shall affect the power to the Commission to take
decisions in relation to the Commitments in accordance with its powers under the Merger
Regulation.
Annex 2 - Information to be provided by Meta to the Monitoring Trustee
1. Meta shall notify the Monitoring Trustee on a quarterly basis of:
a. any non-trivial changes to the terms of access to Publicly Available B2C
Messaging Channel API(s);
b. the nature, extent, cause and duration of any non-trivial degradation or non-
availability of the Publicly Available B2C Messaging Channel APIs within the
meaning of paragraph 5 of these Commitments;
c. any B2C Messaging Channel API functionalities or improvements of such
functionalities made available to the Kustomer Service (other than those made
available to the Kustomer Service only through Publicly Available B2C Messaging
Channel APIs or API Development and Testing), including the percentage and
number of Kustomer Business Users using such functionality or improvement of
such functionalities on a weekly basis in the previous quarter; and
d. any API Development and Testing, including the percentage and number of
Kustomer Business Users using such functionality. If any such API Development
and Testing has exceeded six (6) months in duration, Meta shall also provide the
Monitoring Trustee an explanation of why the testing period has exceeded such
duration.
Annex 4 - B2C Messaging Channel API functionalities and improvements accessed by
the Kustomer Service after the Effective Date
1. B2C Messaging Channel API functionalities or improvements not included within the
scope of Core B2C Messaging Channel API Functionalities as at the Effective Date shall
become Core B2C Messaging Channel API Functionalities pursuant to the following
procedure:
a. Subject to paragraph 1c. of this Annex 4, B2C Messaging Channel API
functionalities or improvements accessed by the Kustomer Service through any
means other than a Publicly Available B2C Messaging Channel API (including for
API Development and Testing) shall constitute Core B2C Messaging Channel API
Functionalities and shall be made available to Third Party CS CRM Providers as
soon as reasonably practicable and not later than twelve (12) months of meeting
the Usage Threshold.
b. On a quarterly basis for the twelve (12) month period following the date of the
Usage Threshold being met, Meta shall provide the Monitoring Trustee with a
progress update in relation to the status of the roll-out to Third Party CS CRM
Providers of access to such functionality.
c. Meta may notify the Monitoring Trustee in writing of any B2C Messaging Channel
API functionality accessed by the Kustomer Service after the Effective Date that
meets the criteria of paragraph 1a. of this Annex 4, but which Meta does not
consider to be essential for Third Party CS CRM Providers to implement B2C
Messaging and should not constitute a Core B2C Messaging Channel API
Functionality, providing detailed reasons. Such notification shall be made within
one month following the date of such functionality meeting the Usage Threshold.
Following consultation with Meta, the Monitoring Trustee in agreement with the
Commission shall confirm within one month of receipt of such notice whether the
relevant B2C Messaging Channel API functionality is essential for Third Party CS
CRM Providers to implement B2C Messaging (in which case it shall constitute a
Core B2C Messaging Channel API Functionality pursuant to paragraph 1a. of this
Annex 4 and shall be made available to Third Party CS CRM Providers as soon as
reasonably practicable and not later than twelve (12) months of the date of the
Monitoring Trustee’s confirmation).
2. For the avoidance of doubt, Core B2C Messaging Channel API Functionalities shall
remain Core B2C Messaging Channel API Functionalities unless excepted pursuant to
paragraph 1c. of this Annex 4 (or deprecated in accordance with paragraph 2 of these
Commitments), irrespective of whether their usage falls beneath the Usage Threshold.
3. For the purposes of this Annex 4, “Usage Threshold” shall mean used by more than 25%
of individual Kustomer Business Users via the Kustomer Service on a weekly basis during
the previous quarter.
