EUROPEAN COMMISSION
DG Competition
Case M.8124 –
Microsoft / LinkedIn
Only the English text is available and authentic.
REGULATION (EC) No 139/2004
MERGER PROCEDURE
Article 6(1)(b) in conjunction with 6(2)
Date: 6.12.2016
Commission européenne, DG COMP MERGER REGISTRY, 1049 Bruxelles, BELGIQUE
Europese Commissie, DG COMP MERGER REGISTRY, 1049 Brussel, BELGIË
Tel: +32 229-91111. Fax: +32 229-64301. E-mail: COMP-MERGER-REGI[email protected]pa.eu.
EUROPEAN COMMISSION
Brussels, 6.12.2016
C(2016) 8404 final
To the Notifying party:
Dear Sir/Madam,
Subject: Case M.8124 – Microsoft / LinkedIn
Commission decision pursuant to Article 6(1)(b) in conjunction with
Article 6(2) of Council Regulation No 139/2004
1
and Article 57 of the
Agreement on the European Economic Area
2
(1) On 14 October 2016, the Commission received notification of a proposed
concentration pursuant to Article 4 of Council Regulation (EC) No 139/2004
(the "Merger Regulation") by which the undertaking Microsoft Corporation
(“Microsoft” or the "Notifying Party", USA) acquires within the meaning of
Article 3(1)(b) of the Merger Regulation control of the whole of the undertaking
LinkedIn Corporation ("LinkedIn", USA) by way of purchase of shares.
3
Microsoft and LinkedIn together are designated hereinafter as the "Parties".
1 THE PARTIES AND THE CONCENTRATION
(2) Microsoft is a global technology company, whose product offering includes
operating systems ("OSs") for personal computers ("PCs"), servers and mobile
devices, related services, cross-device productivity applications and other
1
OJ L 24, 29.1.2004, p. 1. With effect from 1 December 2009, the Treaty on the Functioning of the
European Union ('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.
2
OJ L 1, 3.1.1994, p. 3 (the 'EEA Agreement').
3
Publication in the Official Journal of the European Union No C 388, 21.10.2016, p. 4.
In the published version of this decision, some
information has been omitted pursuant to Article
17(2) of Council Regulation (EC) No 139/2004
concerning non-disclosure of business secrets and
other confidential information. The omissions are
shown thus […]. Where possible the information
omitted has been replaced by ranges of figures or a
general description.
PUBLIC VERSION
2
software solutions, hardware devices, cloud-based solutions, online advertising
(primarily with its web search engine, Bing).
(3) LinkedIn operates a professional social network ("PSN") and generates
revenues through the following product lines: (i) "Talent Solutions" (63% of its
revenue), which include recruiting tools and online education courses; (ii)
"Marketing Solutions" (19% of its revenue), which allow individuals and
enterprises to advertise to LinkedIn's PSN members; and (iii) "Premium
Subscriptions" for both consumer and businesses (18% of its revenue).
(4) On 11 June 2016, Microsoft and LinkedIn signed an Agreement Plan of Merger,
by which Microsoft will acquire all the shares of LinkedIn (the "Transaction").
As a result of the Transaction, Microsoft will acquire sole control over
LinkedIn.
(5) The Transaction therefore constitutes a concentration within the meaning of
Article 3(1)(b) of the Merger Regulation.
2 EU DIMENSION
(6) The undertakings concerned have a combined aggregate world-wide turnover of
more than EUR 5 000 million (Microsoft: EUR 78 223 million; LinkedIn: EUR
2 697 million).
4
Each of them has an EU-wide turnover in excess of EUR 250
million (Microsoft: EUR […] million; LinkedIn: EUR […] million), but each
does not achieve more than two-thirds of its aggregate EU-wide turnover within
one and the same Member State.
(7) The Transaction therefore has an EU dimension within the meaning of Article
1(2) of the Merger Regulation.
3 RELEVANT MARKETS
3.1 PC OSs
(8) OSs are system software products that control the basic functions of computing
devices
5
such as servers, PCs, tablets and mobile devices and enable the user to
use the device and run application software on it.
(9) Microsoft develops and offers OSs for different devices, including PCs, tablets,
and mobile devices, under the “Windows” name.
6
Microsoft’s latest version of
its PC OS, Windows 10, was launched in July 2015. Microsoft’s latest version
of its mobile OS, Windows 10 Mobile, was released in November 2015.
7
(10) LinkedIn does not offer any OS.
4
Turnover calculated in accordance with Article 5(1) of the Merger Regulation and the Commission
Consolidated Jurisdictional Notice, OJ C95, 16.4.2008, p. 1.
5
Case C-3/37.792 – Microsoft, Commission decision of 24 March 2004, recital 37.
6
In addition, Microsoft also develops and offers: an OS that runs its line of gaming consoles (Xbox); an
OS called Windows IoT for use in embedded systems (for instance, in automotive applications, cash
registers, ATMs or self-checkout machines); and an OS for servers called Windows Server. These
OSs, which are not for PCs, are not further discussed in this Decision.
7
Windows 10 Mobile shares user interface elements and apps with Windows 10 for PCs, but retains
distinct OS release.
3
3.1.1 Product market definition
3.1.1.1 Notifying Party's view
(11) The Notifying Party states that the market definition of OSs can be left open.
However, in relation to a possible distinction between mobile OSs and PC OSs,
the Notifying Party stresses that OS providers for PCs are constrained by OS
providers for mobile devices. This is a consequence of the diminishing
popularity of PCs in comparison to mobile devices and smartphones. Moreover,
this effect is strengthened by the fact that Google offers the mobile OS Android
free of charge.
3.1.1.2 Commission's assessment
(12) In Microsoft, the Commission found that OSs for client PCs and OSs for other
client appliances (such as PDAs and smart mobile devices) were not part of the
same product market, given the lack of demand-side substitutability.
8
(13) In Google/Motorola Mobility, based on the responses of the market
investigation, the Commission took the view that OSs for PCs and OSs for smart
mobile devices belong to separate product markets, given that both used
different hardware and had different performance capacities. However, the
Commission ultimately left the exact scope of the product market definition
open in that case.
9
(14) The Commission adopted a similar approach in Microsoft/Nokia, where it found
no indication that the relevant product market for OSs for mobile devices should
be broadened to include PC OSs.
10
(15) For the purpose of this Decision, the Commission retains its previous product
market definition and will carry out its competitive assessment on a conservative
basis on the product market of OS for PCs.
3.1.2 Geographic market definition
3.1.2.1 Notifying Party's view
(16) On the basis of previous Commission decisions, the Notifying Party states that
the scope of the market for OSs is at least EEA-wide, but could also be world-
wide.
3.1.2.2 Commission's assessment
(17) In Microsoft, the Commission found that the relevant geographic market for
client PC OSs was worldwide in scope.
11
In Google/Motorola Mobility and
Microsoft/Nokia, which concerned mobile OS, the Commission found that the
relevant geographic market was at least EEA-wide, or even worldwide, in scope,
but ultimately left the question open.
12
(18) For the purpose of this Decision, the Commission considers that the relevant
geographic market for OSs for PCs is EEA-wide, in line with a more
8
Case C-3/37.792 – Microsoft, Commission decision of 24 March 2004, recital 324-330.
9
Case M.6381 – Google/Motorola Mobility, Commission decision of 13 February 2012, paragraphs 26-
30.
10
Case M.7047 – Microsoft/Nokia, Commission decision of 4 December 2013, paragraph 27.
11
Case C-3/37.792 – Microsoft, Commission decision of 24 March 2004, recital 427.
12
Case M.6381 – Google/Motorola Mobility, Commission decision of 13 February 2012, paragraphs 33-
35; Case M.7047 – Microsoft/Nokia, Commission decision of 4 December 2013, paragraphs 74-77.
4
conservative approach, and will carry out its competitive assessment on that
basis.
3.2 Productivity software
(19) Productivity software consists of applications that enable users to create
documents, databases, graphs, worksheets and presentations or other data
structures used to exchange information. Productivity software includes
applications such as word processing and spreadsheet applications. Productivity
software can be used either on-premises on the user’s own hardware or in the
cloud, hosted by a third-party vendor.
(20) Microsoft is active in the provision of productivity software for PCs and mobile
devices, including by offering the productivity and collaboration suite Office
(through perpetual licenses for on-premises use) and its cloud-based version
Office 365 (through subscriptions). The Office and Office 365 suites include a
number of productivity software solutions, the core of which are: (i) Word
(word processing); (ii) Excel (spreadsheets); (iii) PowerPoint (presentations);
and (iv) Outlook (emails, tasks, contacts and calendar).
13
In the remainder of this
Decision, unless indicated otherwise, the term "Office" will be used to refer to
the Microsoft productivity suites consisting of these four core products
(including both on-premises and cloud versions).
14
(21) LinkedIn does not offer productivity software.
3.2.1 Product market definition
3.2.1.1 Notifying Party's view
(22) The Notifying Party submits that the exact product market definition with
regards to productivity software should be left open. The Notifying Party recalls
that in previous decisions the Commission identified a market for productivity
software, of which personal productivity applications were a segment.
Moreover, the Notifying Party states that even though mobile productivity apps
have previously been recognized as a separate market compared to productivity
applications for desktops and laptops, these markets are converging because of
the shift in usage from mobile devices to PC. The Notifying Party argues that in
any event the precise product market definition can be left open.
13
The data provided by the Notifying Party indicate that Word, Excel, PowerPoint and Outlook are
present in all or most versions of the Microsoft Office and Office 365 suites and, at the same time, they
are the most widely used products by the customers of such suites (Notifying Party's submission of 2
December 2016 at 11:58). In addition, other Office-branded productivity software solutions include
Access (database management), OneNote (multi-user collaboration), Publisher (publishing), Project
(projects, portfolio and resource management), Visio (diagramming) and Sway (interactive content /
storytelling). However, the data provided by the Notifying Party indicate that these other Office-
branded productivity software products are either not included in all or most Office suites (for
example, Access is only included in some Office versions targeted at businesses, and Project is not
part of the Office suites but purchased on a standalone basis), or they account for a much more limited
usage of the Office suite compared to Word, Excel, PowerPoint and Outlook (such as OneNote and
Access).
14
Microsoft also offers Outlook.com (previously called Hotmail), which is a web-based suite of email,
contacts, tasks, and calendaring services. Outlook.com is available free of charge on the web.
Outlook.com is not part of the Microsoft productivity software suite and is not further discussed in this
Decision.
5
3.2.1.2 Commission's assessment
(23) In Oracle/Peoplesoft, the Commission found that business application software
comprises software programs that address some aspect of planning, execution or
collaboration in a business, government or other organisation. The Commission
noted that this software was distinguished from consumer software.
15
The
Commission found that, within business application software, a distinction could
be drawn between personal productivity applications (word processing,
spreadsheets and client-side collaborative applications), as opposed to enterprise
application software (“EAS”). However, the Commission did not further discuss
personal productivity applications in that decision.
16
(24) More recently, in Microsoft/Nokia, the Commission considered that mobile
productivity applications constituted a separate market from other types of
applications for smartphone devices and from productivity applications for
desktops and laptops, i.e. PCs.
17
In Microsoft/Nokia, the Commission also
considered that mobile productivity applications for corporate users may
constitute a separate product market, but ultimately left the question open. The
Commission also left open the question whether mobile productivity
applications could be segmented by OS or functionality.
18
(25) For the purpose of this Decision, the Commission retains its previous product
market definition of productivity software for PCs, without it being necessary to
consider whether the market could be further segmented depending on the type
of productivity software, given that the remedies address the competition
concerns identified by the Commission even on some of these narrower markets.
3.2.2 Geographic market definition
3.2.2.1 Notifying Party's view
(26) The Notifying Party does not take a view on the geographic market definition
for productivity software.
3.2.2.2 Commission's assessment
(27) In previous decisions, the Commission found that the market for mobile
productivity applications was at least EEA-wide, if not worldwide, in scope, but
ultimately left the precise geographic market definition open.
19
(28) For the purpose of this Decision, the Commission considers the geographic
market for productivity software for PCs as EEA-wide, in line with a more
conservative approach, and will carry out its competitive assessment on that
basis.
3.3 Customer relationship management ("CRM") software solutions
(29) 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, in particular the
15
Case M.3216 – Oracle/Peoplesoft, Commission decision of 26 October 2004, recital 15.
16
Case M.3216 – Oracle/Peoplesoft, Commission decision of 26 October 2004, recital 15.
17
Case M.7047 – Microsoft/Nokia, Commission decision of 4 December 2013, paragraphs 46-55.
18
Case M.7047 – Microsoft/Nokia, Commission decision of 4 December 2013, paragraph-56.
19
Case M.7047 – Microsoft/Nokia, Commission decision of 4 December 2013, paragraphs 80-81.
6
sales department, to improve customer relationships, to better manage accounts,
to enhance sales effectiveness, to optimise data quality, and to mitigate
regulatory compliance risks.
(30) CRM software solutions can offer different functionalities, such as (i) Customer
Sales and Support ("CSS"); (ii) Sales (including sales force automation ("SFA")
and sales force enablement ("SFE")), (iii) Marketing; and (iv) Digital Commerce
("DC").
20
Most CRM software solutions offer several functionalities and can be
on-premises or cloud-based.
(31) Microsoft is active in the provision of CRM software solutions, where it offers
its CRM software solution “Microsoft Dynamics”. LinkedIn is not active in this
market.
3.3.1 Product market definition
3.3.1.1 Notifying Party's view
(32) On the basis of previous Commission decisions, the Notifying Party argues that
CRM software solutions form a distinct product market. Moreover, the
Notifying Party states that this market should not be further subdivided on the
basis of the functionality of the software, the industry sector or the mode of
deployment.
(33) As regards a possible segmentation of CRM software solutions on the basis of
functionality, in categories such as customer sales and support, sales, marketing
and digital commerce, the Notifying Party states that CRM customers generally
view different CRM software solutions as valid substitutes for one another.
While industry reports segment CRM software solutions on the basis of
functionality, most CRM software solutions have capabilities in all of these
categories and therefore it is not appropriate to distinguish further segments on
the basis of those functionalities.
(34) Furthermore, the Notifying Party submits that no distinction on the basis of
industry sector should be made. Whereas certain CRM providers only serve one
particular industry, customers of those industries are still able to choose general
purpose CRM software solutions. Due to this demand-side substitutability, the
Notifying Party claims that the market should not be subdivided any further.
(35) Lastly, the Notifying Party states that it is not difficult for a provider of on-
premises software to adapt its product to cloud-based software, which indicates
20
These functionalities are the most relevant ones identified under the industry classification of the
analyst Gartner.
In the market investigation, Salesforce also explained that another segmentation may be relevant from
the functionality view-point, that is to say between (i) Business-to-Business ("B2B") and (ii) Business-
to-Consumer ("B2C"). According to Salesforce, the functionality of the processes and the software
between these two segments are different and provides the following examples "small contact
database for B2B vs large contact database for B2C; longer sales time for B2B vs shorter sales time
for B2C; fewer leads for B2B vs more leads for B2C; higher level of data detail required in B2B vs
lower level of data detail required in B2C; etc." (Salesforce's response to questionnaire to CRM
Competitors Q5 of 14 October 2016, question 3). Indications of the relevance of this segmentation for
the competitive assessment of the Transaction have been provided also in the responses of other CRM
competitors, in particular with respect to the input needed for certain advanced functionalities, such as
those based on machine learning (in this regard see footnote 230), see Zoho's and E-Deal's responses
to questionnaire to CRM Competitors Q5 of 14 October 2016. Therefore, while not assessing further
whether a market segmentation between B2B and B2C is warranted on the basis of demand and supply
side considerations, in its competitive assessment of the Transaction the Commission will take into
consideration also this segmentation.
7
a high degree of supply substitutability. As a result, the Notifying Party argues
that no distinction between the modes of deployment of CRM software solutions
should be made.
3.3.1.2 Commission's assessment
(36) The Commission previously considered the market for the provision 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”).
21
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 sales force automation or marketing
automation; (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
("SaaS"); or (iv) customisation, i.e. custom-built solutions as opposed to
standardised software.
22
(37) The Commission considered CRM software solutions as a relevant product
market, without further segmentations, also in more recent cases.
23
(38) 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 relevant industry 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.
24
(39) With respect to a possible distinction of CRM software solutions on the basis of
functionality, most CRM providers responding to the market investigation
indicated that they provide most or all of the relevant functionalities identified
by the Gartner reports.
25
Those respondents also indicated that they offer those
functionalities to customers either separately, via different licences or
subscriptions, or bundled together under the same licence or subscription.
26
Certain CRM customers indicated that they use more than one CRM
functionality.
27
The majority of those CRM customers indicated that they
purchase these functionalities bundled together under a same licence or
subscription. However, some CRM customers indicated that they purchase the
CRM functionalities based on separate licences.
28
The market investigation
yielded mixed results as to whether CRM customers purchase different CRM
functionalities from different CRM providers: some CRM customers indicated
21
Case M.3978 – Oracle/Siebel, Commission decision of 22 December 2005, paragraphs 11-16.
22
Case M.3978 – Oracle/Siebel, Commission decision of 22 December 2005, paragraph 16.
23
Case M.4944 – SAP/Business Objects, Commission decision of 27 November 2007, paragraph 7; Case
M.4987 – IBM/Cognos, Commission decision of 4 January 2008, paragraphs 7-9; Case M.5904 –
SAP/Sybase, Commission decision of 20 July 2010, paragraph 21; Case M.7334 – Oracle/Micros,
Commission decision of 29 August 2014, paragraph 9.
24
Case M.7337 – IMS Health/Cegedim Business, Commission decision of 19 December 2014,
paragraphs 91-92.
25
Responses to questionnaire to CRM Competitors Q5 of 14 October 2016, question 3.
26
Responses to questionnaire to CRM Competitors Q5 of 14 October 2016, question 4.
27
Responses to questionnaire to CRM Customers Q7 of 14 October 2016, question 3.
28
Responses to questionnaire to CRM Customers Q7 of 14 October 2016, question 4.
8
that they source CRM functionalities from the same provider, whereas others
source different functionalities from separate providers.
29
(40) From a demand-side perspective, all CRM providers were of the opinion that
customers view the different functionalities of CRM software solutions as
complements to each other, rather than substitutes.
30
One CRM provider
explained that “[a] CRM product supporting specific CRM functionalities by
definition cannot be substituted with another product that does not support these
functionalities. Thus, a customer wishing fully to automate the CRM function
would need to acquire complementary functionality, if that functionality is not
included within the scope of an existing implemented solution.” Most CRM
customers confirmed that they view the different functionalities of a CRM
software solution as complements for their different needs, rather than as
substitutes to each other.
31
(41) From a supply-side perspective, when asked to comment on whether they would
be able to start providing a different CRM functionality in the short term and
without incurring significant investments, CRM providers’ responses were
mixed.
32
One respondent, Salesforce, commented that given the developments
of the market and the requirements of customers, CRM providers are not able to
start providing other functionalities in the short term and without incurring
significant investments. This respondent explained that, since the core
technology of a CRM software solution is well-established, introducing a
minimum viable product or minimum saleable product is not an easy task and
requires significant material investment. The same respondent added that
usually a new functionality is developed by one CRM provider (often through
the acquisition of a start-up), which then enjoys a first mover advantage.
(42) As regards a possible distinction depending on the type of service, namely cloud
or on-premises, the market investigation indicated that most CRM providers
offer either on-premises or cloud-based solutions, but not both.
33
Most CRM
providers also responded that they provide these solutions by means of different
licences and subscriptions, rather than under the same arrangement.
34
CRM
customers responding to the market investigation indicated that they use both
types of services.
35
These services are usually purchased separately through
different licences.
36
CRM customers also indicated that, when they use both an
on-premises and a cloud-based CRM, they usually purchase them from different
providers.
37
(43) From a demand-side perspective, most CRM providers responding to the market
investigation indicated that in their view CRM customers view on-premise and
cloud-based CRM software solutions as substitutes rather than complements.
One CRM provider indicated that for the same CRM functionality customers
typically choose either a cloud-based solution or an on-premises solution, but
29
Responses to questionnaire to CRM Customers Q7 of 14 October 2016, question 5.
30
Responses to questionnaire to CRM Competitors Q5 of 14 October 2016, question 6.
31
Responses to questionnaire to CRM Customers Q7 of 14 October 2016, question 6.
32
Responses to questionnaire to CRM Competitors Q5 of 14 October 2016, question 5.
33
Responses to questionnaire to CRM Competitors Q5 of 14 October 2016, question 7.
34
Responses to questionnaire to CRM Competitors Q5 of 14 October 2016, question 8.
35
Responses to questionnaire to CRM Customers Q7 of 14 October 2016, question 7.
36
Responses to questionnaire to CRM Customers Q7 of 14 October 2016, question 8.
37
Responses to questionnaire to CRM Customers Q7 of 14 October 2016, question 9.
9
not both. Another respondent indicated that the market for CRM software
solutions is shifting towards cloud-based solutions, whereas on-premises
services are usually still provided to legacy customers or customers that need an
on-premises solution for regulatory purposes.
38
CRM customers responding to
the market investigation mostly confirmed that they consider cloud and on-
premises CRM software solutions as substitutes to each other rather than
complements.
39
(44) From a supply-side perspective, most CRM providers indicated that they would
not be able to start supplying a different type of CRM software solution shortly
and without significant costs.
40
One respondent explained that it would take a
significant development and sales and marketing effort to build a new product
on a different platform.
(45) When considering a possible segmentation based on customer size, from a
demand-side perspective, the views of CRM providers were mixed. Some
commented that the requirements and sourcing patterns of large enterprises
differ from those of small and medium enterprises, whereas others did not find
this distinction relevant. For instance, one respondent explained that “[t]he
requirements and suppliers of SMEs and large enterprises by and large tend to
be the same with regard to CRM”, whereas another explained that “[l]arge
enterprises usually have complex workflows, integrations with their internal
databases or other 3rd party software, requirement for a sandbox environment,
need for more customization/fields and API [application program interface]
calls”. Moreover, another competitor explained that “[v]ery large customers are
looking for both a solution and a vision…On the contrary, very small and small
business customers are generally more interested in ease of use and
maintainability. They will focus on the solution with usually no deep integration
between processes. They are less likely to be looking for a vision”.
41
(46) From a supply-side perspective, most CRM providers indicated that they are
equally capable of providing CRM software solutions to all CRM customers,
irrespective of their size.
42
This is also indirectly confirmed by the fact that
CRM providers actually do supply both larger and medium-small customers,
and have revenues from both types of customers.
43
(47) Finally, with respect to whether it would be appropriate to distinguish CRM
software solutions based on the type of industry of the customers, respondents
among CRM providers had mixed views. From a demand-side perspective,
those respondents commented that customers may have different needs and
requirements depending on their industry sector, whereas others found that,
while these differences may exist, the requirements remain broadly similar for
all customers. For instance, one respondent was of the view that “[e]ach
industry has unique requirements that necessitate customization of the CRM
solution specifically for the industry”, whereas another explained that
“[d]epending on the industry and the particular business model of the particular
customer, some requirements, or the emphasis on some requirements, may be
38
Responses to questionnaire to CRM Competitors Q5 of 14 October 2016, question 10.
39
Responses to questionnaire to CRM Customers Q7 of 14 October 2016, question 10.
40
Responses to questionnaire to CRM Competitors Q5 of 14 October 2016, question 9.
41
Responses to questionnaire to CRM Competitors Q5 of 14 October 2016, question 11.
42
Responses to questionnaire to CRM Competitors Q5 of 14 October 2016, question 12.
43
Responses to questionnaire to CRM Competitors Q5 of 14 October 2016, question 13.
10
different […] However, differences and [sic] business model and associated
requirements also exist among businesses active in the same sector and core
needs are similar across different organisations. By and large, sourcing of CRM
solutions will be similar across different types of businesses in different
sectors”.
44
(48) In any event, from a supply-side perspective all CRM providers indicated that
they are equally able to provide CRM software solutions to customers active in
different business sectors. For instance, one respondent commented that “[m]ost
CRM [software] solutions share the same core characteristics for all types of
industries and solutions sold in one specific industry can therefore very easily
be adapted to be sold in other industries. SAP also offers industry-specific
functionality for certain industries.”
45
In particular, the Commission's notes that
the Notifying Party's and its main competitors' CRM software solution offerings
do not have a specific industry focus.
(49) Based on the results of the market investigation, the Commission therefore
considers that CRM software solutions may likely be further distinguished on
the basis of functionality, as customers view the different functionalities as
complementary, rather than substitutable, and there appears to be also limited
scope for supply-side substitution by CRM providers. Conversely, it appears
that CRM customers tend to purchase various functionalities, and most CRM
providers offer them. As regards the distinction between cloud-based and on-
premises solutions, the market investigation indicates that there is demand-side
substitutability between these services, as CRM customers usually see these
options as alternatives. However, from a supply-side perspective, CRM
providers appear to not be able to easily develop and supply another type of
CRM software solution. Finally, based on the responses to the market
investigation, it appears that a segmentation of CRM software solutions on the
basis of the customer’s size or industry sector is not relevant, given that, while
there may be some differences in terms of demand, most customers have the
same requirements and needs (irrespective of their size and industry), and all
CRM providers appear capable to offer CRM software solutions to all types of
customers.
(50) In any event, for the purpose of this Decision, the exact product market
definition for CRM software solutions can be left open, as the Transaction does
not raise serious doubts as to its compatibility with the internal market with
respect to CRM software solutions under any alternative product market
definition.
3.3.2 Geographic market definition
3.3.2.1 Notifying Party's view
(51) The Notifying Party submits that the geographic market for CRM software
solutions is worldwide in scope for four reasons. Firstly, a CRM software
solution does not vary across different regions. Secondly, there are no
regulatory/technical differences or local customisation when a CRM software
solution is provided among different regions. Thirdly, there are no transport
costs, particularly for cloud-based software. Lastly, the main providers of CRM
software conclude their contracts on a global or multi-country basis. However,
44
Responses to questionnaire to CRM Competitors Q5 of 14 October 2016, question 14.
45
Responses to questionnaire to CRM Competitors Q5 of 14 October 2016, question 15.
11
the Notifying Party reiterates that the Commission has left open the precise
scope in previous decisions, and this can be done in the present case as well.
3.3.2.2 Commission's assessment
(52) 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.”
46
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.
47
The same conclusion was reached in
IBM/Cognos.
48
(53) 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 provision of CRM software among Member States. The
Commission ultimately left the exact geographic market definition open, as the
transaction did not raise concerns irrespective of the precise geographic scope of
the market.
49
(54) In the present case, respondents to the market investigation among CRM
providers unanimously agreed that the geographic market for CRM software
solutions is worldwide in scope, as there are no differences in sourcing patterns
or requirements.
50
One respondent explained that “[c]ustomer-demand does not
vary across different geographic regions except for certain language or other
local requirements which can easily be customized in the software applications.
Almost all customers have the ability to purchase their software applications
without being confined to the suppliers that have a physical presence in their
territory. Most software vendors are global players and compete with each
other across the world. Given that the software applications can easily be
adapted to local requirements (primarily language requirements), software
vendors have the ability to expand the scope of their activities to territories in
which they are not currently present”. Furthermore, all CRM providers
indicated that they are equally able to offer their products in the EEA and in the
rest of the world.
51
Indeed, CRM providers’ EEA revenues appear to be a
fraction of worldwide revenues.
52
(55) Most CRM customers also indicated that they equally consider possible
suppliers of CRM software solutions within the EEA and worldwide.
53
(56) Based on the results of the market investigation, the Commission concludes that
the market for CRM software solutions is EEA-wide, if not worldwide, in scope.
46
Case M.3978 – Oracle/Siebel, Commission decision of 22 December 2005, paragraph 18.
47
Case M.3978 – Oracle/Siebel, Commission decision of 22 December 2005, paragraph 19.
48
Case M.4987 – IBM/Cognos, Commission decision of 4 January 2008, paragraph 16.
49
Case M.7337 – IMS Health/Cegedim Business, Commission decision of 19 December 2014, paragraph
96.
50
Responses to questionnaire to CRM Competitors Q5 of 14 October 2016, question 16.
51
Responses to questionnaire to CRM Competitors Q5 of 14 October 2016, question 17.
52
Responses to questionnaire to CRM Competitors Q5 of 14 October 2016, question 18.
53
Responses to questionnaire to CRM Customers Q7 of 14 October 2016, question 11.
12
The precise scope of the geographic market can be left open, as the Transaction
does not raise serious doubts as to its compatibility with the internal market with
respect to CRM software solutions under any alternative geographic market
definition.
3.4 Sales intelligence solutions
(57) Sales intelligence solutions provide sales professionals with background and
contact information about individuals (such as name, address, phone number,
place of employment, title and position, etc.) or companies (such as financial
information and metrics, organisational hierarchy and leadership structure,
company’s products and services, industry background, etc.). Sales
professionals use this information to identify new leads or to update information
about existing contacts, and to more effectively identify, reach out to, and win
business from potential customers and relevant decision makers.
(58) Microsoft is not active in the provision of sales intelligence solutions. Among its
"Premium Subscriptions", LinkedIn offers a sales intelligence solution branded
“Sales Navigator”. Sales Navigator is a subscription-based solution that draws
from LinkedIn’s database of user data (“LinkedIn full data”
54
) and displays a
subset of this database to its users, mainly sales professionals, to allow for the
quick identification and creation of new customer leads and sales opportunities.
Sales Navigator is available both in a desktop and mobile version.
3.4.1 Product market definition
3.4.1.1 Notifying Party's view
(59) The Notifying Party argues that a separate product market for sales intelligence
solutions could exist. However, sales intelligence solutions could also be part of
a larger market covering sales acceleration technology.
(60) The Notifying Party refers to a report by InsideSales, “Sales Acceleration
Technology Market Size Study”, where it is explained that “[t]he sales
acceleration technology industry is the business space between CRM and
marketing automation which facilitates, and thereby accelerates, all processes
pertinent to the sales pipeline”.
55
The Notifying Party explains that sales
personnel use both CRM software solutions and marketing automation
technologies to interact with potential clients. While a CRM software solution
organizes the interactions between a company and its customers, marketing
automation software aids in scheduling, tracking, and task automation. Sales
intelligence solutions create a connection between these two areas with the goal
of making sales processes faster. As such, in the Notifying Party’s view, sales
intelligence solutions could amount to a type of sales acceleration technology
and thus should be included in a broader market covering sales acceleration
technology.
(61) The Notifying Party submits that in any event, the exact product market
definition for sales intelligence solutions can be left open.
54
LinkedIn full data refers to all the data that LinkedIn collects, or could collect, and store about its users
and their activity, such as professional details, connections, interests, posts, endorsements.
55
Form CO, Annex 8 – Document 2.
13
3.4.1.2 Commission's assessment
(62) The Commission has not considered the market for the provision of sales
intelligence solutions, or the possibly broader market for sales acceleration
technology, in previous decisions.
(63) Most CRM customers responding to the market investigation indicated that they
use sales acceleration technologies, including specifically sales intelligence
solutions.
56
As regards sales intelligence solutions, CRM customers indicated
that they rely on several third-party sources, as well as internal and self-
developed inputs.
57
(64) From a demand-side perspective, when asked how they view the various types
of sales acceleration technology, including sales intelligence solutions, most
CRM customers replied that these different products are complements, rather
than substitutes, as they are used for different purposes. One respondent
explained that the various technologies are used for different purposes and so
are complements to each other. Another commented that “[e]ach of the
individual products that we use are either separate or add additional
functionality or insight to one another and so are not substitutes”.
58
(65) Most CRM providers replying to the market investigation also indicated that, in
their opinion, CRM customers view the different types of sales acceleration
technology as complement rather than substitutes. One respondent explained
that each type of sales acceleration technology “respond[s] to a specific need
from a customer standpoint. By way of example, chats do not respond to the
same needs than Sales Intelligence Solutions. These are two different product
categories which are complimentary and can form part of the same CRM
[software] solution”.
59
(66) From a supply-side perspective, most CRM providers commented that it is not
easy for a supplier of one type of sales acceleration technology to start providing
a different type of sales acceleration technology.
60
One respondent stated that
“[i]t would take considerable time and financial investment to development and
successfully offer products in other categories in the short term.” Another CRM
provider commented that “[a]ny significant software development project
requires significant investments. The time and cost required depends on the
particular product to be developed, and the relative burden involved will be
greater for small recent entrants than for established market participants. It is
likely that a recent entrant focusing on one category would not have the
resources required additionally to focus on a different product category”.
Finally, Salesforce explained that “[t]hese are very different technologies. The
basic principles, core functionalities and the overall structure of these
technologies differ significantly. By way of example, analytics technologies and
SMS/Communications technologies do not rely on the same concepts, the same
know-how, the same principles, etc. Therefore, a supplier of sales acceleration
technology will not be able to start providing products of other categories in the
56
Responses to questionnaire to CRM Customers Q7 of 14 October 2016, question 12.
57
Responses to questionnaire to CRM Customers Q7 of 14 October 2016, question 12.2.
58
Responses to questionnaire to CRM Customers Q7 of 14 October 2016, question 13.1.
59
Responses to questionnaire to CRM Competitors Q5 of 14 October 2016, question 19.
60
Responses to questionnaire to CRM Competitors Q5 of 14 October 2016, question 20.
14
short term and without incurring significant investments”. Responses from
providers of sales intelligence solutions were not conclusive.
61
(67) With respect to a possible further segmentation of sales intelligence solutions by
industry type, CRM providers gave mixed views. Some respondents indicated
that the basic needs are similar across business sectors, but also highlighted that
there may be certain industry-specific requirements.
62
(68) Based on the results of the market investigation, the Commission considers that
sales intelligence solutions may constitute a separate product market within
sales acceleration technology, given that there appears to be limited demand-
side and supply-side substitutability with respect to the other types of sales
acceleration technology. With respect to the segmentation by industry type, the
Commission considers that for the purposes of this Decision such segmentation
is not relevant as only LinkedIn is active in the provision of sales intelligence
solutions and the market investigation did not provide indications that LinkedIn
is particularly strong in one or more specific sectors.
(69) For the purpose of this Decision, the Commission therefore considers the
relevant product market to be that for the provision of sales intelligence
solutions, without further segmentations.
3.4.2 Geographic market definition
3.4.2.1 Notifying Party's view
(70) The Notifying Party submits that the market for sales intelligence solutions is
likely to be worldwide in scope. However, the exact scope of the market can be
left open.
3.4.2.2 Commission's assessment
(71) All CRM customers replying to the market investigation indicated that, for the
purpose of acquiring sales intelligence solutions, they consider both EEA and
worldwide vendors.
63
(72) Most CRM providers were of the opinion that generally there is no difference in
the provision of sales intelligence solutions at the EEA and worldwide level,
with respect to sourcing process. However, some respondents highlighted that
there may be differences in terms of regulatory and compliance requirements in
certain areas.
64
(73) Based on the above, the Commission considers that the relevant geographic
scope of the market for sales intelligence solutions is at least EEA-wide, if not
worldwide in scope. In any event, the precise geographic market definition can
be left open, as the Transaction does not raise serious doubts as to its
compatibility with the internal market in relation to sales intelligence solutions.
3.5 Online communications services
(74) Online communications services are multimedia communications solutions that
allow people to communicate by means of an application or software in real
time. They can be distinguished between consumer and enterprise
61
Responses to questionnaire to Sales Intelligence Solutions Competitors Q6 of 14 October 2016,
question 5.
62
Responses to questionnaire to CRM Competitors Q5 of 14 October 2016, question 21.
63
Responses to questionnaire to CRM Customers Q7 of 14 October 2016, question 14.
64
Responses to questionnaire to CRM Competitors Q5 of 14 October 2016, question 22.
15
communications services. Consumer communications services enable people to
reach out to their friends, family members and other contacts in real time.
Enterprise communications services are used by companies for business and
professional purposes.
(75) Microsoft offers online communications services, notably “Skype” and Skype
for Business (“SfB”). LinkedIn does not offer online communications services.
3.5.1 Product market definition
3.5.1.1 Notifying Party's view
(76) The Notifying Party submits that the definition of the product market for online
communications services can be left open.
3.5.1.2 Commission's assessment
(77) In previous decisions, the Commission distinguished between consumer
communications services and enterprise communications services (also referred
to as “unified communications” or "UC"), viewing the two as belonging to
distinct product markets.
(78) In Microsoft/Skype, the Commission noted that enterprise communications
services are more sophisticated and reliable than consumer communications
services. Enterprises have different and higher service requirements, such as
redundancy and robustness, security, reliability, ancillary functionality,
management and support requirements. The Commission further noted that
enterprise communications services offer additional features in terms of
collaborating tools, such as the possibility to share and edit a document in real-
time from different places. As regards voice calls, enterprise-grade
communications services require features and functionality which are not
available in consumer-grade communications services, such as park/hold, mute,
simultaneous ringing or integrated voicemail. The Commission ultimately
concluded that consumer communications services and enterprise
communications services form two distinct product markets.
65
(79) The Commission maintained the same distinction in Microsoft/Nokia and
Facebook/WhatsApp, where it analysed consumer communications services as a
separate product market from enterprise communications services.
66
(80) With respect to possible further segmentations of consumer communications
services, in Microsoft/Skype the Commission left open the question whether the
market for consumer communications services needed to be further segmented
by platform or by OS.
67
In Microsoft/Nokia, the Commission noted that there
were indications that there may be a separate market for consumer
communications apps, which could be segmented by platform. With respect to
segmentation by OS, the Commission noted that several communications
applications are available on several OSs. The Commission ultimately left the
product market definition open in that case.
68
In Facebook/WhatsApp, the
Commission considered that a segmentation based on platforms was
appropriate, and therefore carried out its assessment with respect to consumer
65
Case M.6281 – Microsoft/Skype, Commission decision of 7 October 2011, paragraphs 14 and 17.
66
Case M.7047 – Microsoft/Nokia, Commission decision of 4 December 2013, paragraphs 43-45; Case
M.7217 – Facebook/WhatsApp, Commission decision of 3 October 2014, paragraphs 20-34.
67
Case M.6281 – Microsoft/Skype, Commission decision of 7 October 2011, paragraphs 10-43.
68
Case M.7047 – Microsoft/Nokia, Commission decision of 4 December 2013, paragraphs 41-45.
16
communications applications for smartphones. However, based on the results of
the market investigation, it found that a distinction of consumer communications
services on the basis of the OS or the functionality would not be appropriate.
69
(81) With respect to enterprise communications services, in previous decisions the
Commission defined enterprise communications services as “products and
services […] used by business customers of all sizes to improve workgroup and
collaborative communications, and […] designed to provide a simple and
consistent user experience across all types of communications (telephone, fax,
email, voicemail, voice and videoconference, instant messaging, etc.)”.
70
(82) In Microsoft/Skype, the Commission considered, but ultimately left open,
whether it would be appropriate to further segment enterprise communications
services on the basis of, for example, functionality, OS and platform.
71
With
respect to functionality, the Commission noted among the possible
functionalities of enterprise communications: advanced telephony, unified
messaging – email, fax, voice messaging combined, web, voice and
videoconferencing, IM/presence, collaborating tools.
72
(83) For the purpose of this Decision, the Commission retains the product market
definition adopted in previous cases, and will carry out its assessment with
respect to the separate product market for enterprise communications services.
The question whether enterprise communications services should be further
segmented depending on platform, OS or functionality can be left open, as the
Transaction does not raise concerns with respect to enterprise communications
services irrespective of the exact product market definition.
3.5.2 Geographic market definition
3.5.2.1 Notifying Party's view
(84) The Notifying Party argues that the definition of the geographic market for
online communications services should be left open.
3.5.2.2 Commission's assessment
(85) In previous decisions, the Commission found that the market for enterprise
communications services was EEA-wide, if not worldwide, in scope, but
ultimately left the precise geographic market definition open.
73
With respect to
consumer communications services, the Commission also found that the market
was EEA-wide, if not worldwide in scope, but ultimately left the geographic
market definition open.
74
In Facebook/WhatsApp, the Commission noted that
the market for consumer communications applications was likely worldwide in
scope, but defined it as EEA-wide, in line with a more conservative approach.
75
(86) For the purpose of this Decision, the Commission considers the relevant
geographic market for consumer communications services and enterprise
69
Case M.7217 – Facebook/WhatsApp, Commission decision of 3 October 2014, paragraphs 21-22.
70
Case M.5669 – Cisco/Tandberg, Commission decision of 29 March 2010, paragraph 9.
71
Case M.6281 – Microsoft/Skype, Commission decision of 7 October 2011, paragraphs 44-63.
72
Case M.6281 – Microsoft/Skype, Commission decision of 7 October 2011, paragraph 51.
73
Case M.6281 – Microsoft/Skype, Commission decision of 7 October 2011, paragraphs 67-68; Case
M.5669 – Cisco/Tandberg, Commission decision of 29 March 2010, paragraph 33.
74
Case M.6281 – Microsoft/Skype, Commission decision of 7 October 2011, paragraphs 66 and 68.
75
Case M.7217 – Facebook/WhatsApp, Commission decision of 3 October 2014, paragraphs 36-44.
17
communications services is EEA-wide, in line with a more conservative
approach.
3.6 PSN services
(87) Social networking ("SN") services can be generally described as multi-sided
platforms
76
that enable users to connect, share, discover and communicate with
each other across multiple devices (mobile and desktop) and means (e.g., via
chats, posts, videos, recommendations). SN services are used to build social
relations among people who share similar personal and career interests,
activities, backgrounds or real-life connections. A sub-set of SN services are
focused on connecting with professional contacts and are therefore typically
referred to as PSN services. The vast majority of SN services are provided free
of monetary charges. They can however be monetised through other means,
such as advertising or charges for premium services.
(88) LinkedIn offers PSN services. By contrast, Microsoft does not offer PSN
services, but it operates an enterprise social network, “Yammer”.
77
3.6.1 Product market definition
3.6.1.1 Notifying Party's view
(89) The Notifying Party submits that social networking services should not be
segmented according to intended use (e.g., professional versus private), or on
any other basis. The Notifying Party explains that whereas some SN services
cover a broad range of use cases, others, such as LinkedIn, simply focus on a
subset of use cases. The Notifying Party argues that there is an ever increasing
overlap in the use cases offered by competing SN services: services initially
targeting non-professional use have expanded their appeal such that they also
target professional development and career opportunities. The Notifying Party
argues that the multi-sided nature of SN services, their variety of monetization
opportunities and their user reach support this argument. As a result, no
distinction between professional or private use of SN services should be made.
(90) Additionally, the Notifying Party considers that so called “vertical” social
networks should also be considered within the overall market for SN services.
The Notifying Party explains that vertical SN services are specialized services
that focus on connecting the members of a particular profession: as such, these
social networks are limited to professionals active within the same field and so
members are trusted peers (for instance, Academia for academics, ResearchGate
for scientists and researchers, Github for programmers, Doximity for doctors).
Nevertheless, the Notifying Party submits that these vertical SN services offer
recruiting services, job listings, and search and discovery services like those
offered by LinkedIn.
(91) The Notifying Party further argues that SN services should be distinguished as a
separate product market from enterprise social networks, because the latter lack
the external dimension of the former.
(92) Enterprise social networks are social software platforms that facilitate
communication and collaboration amongst employees within the same
76
Typically, social networks provide the services described in this paragraph to consumers (B2C side)
while they provide other services to businesses (B2B side) such as online advertising and recruitment
services (in the case of LinkedIn).
77
On the meaning of "enterprise social network", see paragraph (92) below.
18
organization. They may include a variety of features and functionalities (such as
newsfeeds, groups, search, file sharing and content management, messaging and
notifications).
(93) The Notifying Party argues that, while some providers offer both social
networks and enterprise social networks (for instance, Facebook also offers
“Facebook Workplace”) and there is some commonality in terms of
functionality, enterprise social networks, in their current form, and social
networks do not exhibit the necessary degree of substitutability to form part of
the same market.
(94) In that respect, the Notifying Party explains that enterprise social networks lack
the external dimension that social networks have. In particular, they do not
feature some of the key social functions of a social network, such as the
possibility for employees to create a public or semi-public profile, to compile
lists of external connections and to engage with them. In addition, enterprise
social networks are also unsuited for most use cases of social networking, such
as building an external professional identity, showcasing expertise and external
recruitment, and thus lack the requisite degree of demand-side substitutability.
Rather, enterprise social networks are used for internal communications and
relations within the same company.
3.6.1.2 Commission's assessment
(95) In Facebook/WhatsApp, the Commission found that SN services should be
distinguished from consumer communications apps. While these two services
may offer a certain number of common functionalities, and the line between the
two services is increasingly blurred, the Commission also noted that SN services
tend to offer a richer social experience compared to consumer communications
apps. Furthermore, while both social networks and consumer communications
apps enable communication between users, the communications functionalities
and their usage differ. Finally, the Commission noted that social networks tend
to enable communication and information sharing with a wider audience than
consumer communications apps, which are more personal and targeted.
78
(96) In Facebook/WhatsApp the Commission also considered whether social
networking services should be further segmented according to the platform (that
is, PC, smartphone and tablet) or an OS (such as Windows, MacOS, Android or
iOS). Based on the results of the market investigation, the Commission found
that social networking services should not be further segmented according to a
platform or an operating system.
79
(97) Finally, in Facebook/WhatsApp, the Commission also considered, but ultimately
left open, the question whether SN services could be segmented according to
their intended use, namely SN services promoting interpersonal contact and
services used for professional purposes. The Commission noted that the market
investigation suggested that such a distinction might be drawn, but also that
there were overlaps between the purposes of intended use.
80
(98) Respondents to the market investigation in the present case identified several
functionalities as essential in order for a platform to be considered as a social
78
Case M.7217 – Facebook/WhatsApp, Commission decision of 3 October 2014, paragraphs 52-56.
79
Case M.7217 – Facebook/WhatsApp, Commission decision of 3 October 2014, paragraphs 57-59.
80
Case M.7217 – Facebook/WhatsApp, Commission decision of 3 October 2014, paragraphs 59-60.
19
network by an end-user. In particular, the most relevant ones were considered to
be the creation of a user profile and the possibility to send/receive messages,
closely followed by several others (search for other people in the network,
send/accept invitations to connect with new contacts, post/share content, post
comments on items posted by others, interact with other users through private or
public groups and have a newsfeed displaying news from the user's
connections).
81
Most respondents indicated that they offer most, if not all, of
these functionalities.
82
(99) In the present case, most respondents to the market investigation indicated that
they offer their SN services to a target user group consisting of both private and
business users, whereas some respondents indicated that they focused on
professionals only, or on a specific sub-set of professionals. For instance, one
respondent explained that it targets the “users that are (visually) creative, such
as architects, photographers, interior designers. This user group is more narrow
or specific than the user group LinkedIn is targeting”. Another respondent
indicated that it targets IT professionals.
83
When asked to identify themselves,
certain respondents identified themselves as “professional” social networks,
whereas others did not take a view.
84
One respondent explained this self-
qualification by referring to the fact that it requires its users to create public
profiles with work experience, language, education and other professional skills,
and that its users create a network inviting colleagues and business partners, and
post professional content.
(100) When asked to identify what factors would distinguish a PSN, respondents
mentioned several elements, such as: the fact that a user can apply directly to job
offers through the profile; the ability to indicate professional experience,
education, and skill; the different type of connections on the network
(colleagues and business partners rather than family and friends), as well as the
creation of business-oriented relationships; the different type of content shared
(professional rather than connected to private life); the use and maintenance of a
business identity.
85
(101) Additionally, respondents also identified the functionalities that would be
essential for a PSN. In this respect, most respondents pointed to the functionality
of creating and updating a detailed resume or CV. Other prominent essential
functionalities were the search for jobs, asking to be introduced to new contacts
through a common connection, recommending contacts and receiving
recommendations, and a search for jobs functionality.
86
(102) Those respondents that identified themselves as PSNs
87
indicated that they offer
the aforementioned “essential” functionalities,
88
including the creation and
update of a CV, searching for jobs, receiving alerts and ads about jobs, and
asking to be introduced to new contacts through a common connection.
89
81
Responses to questionnaire to Social Networks Q3 of 17 October 2016, question 5.
82
Responses to questionnaire to Social Networks Q3 of 17 October 2016, question 4.
83
Responses to questionnaire to Social Networks Q3 of 17 October 2016, question 6.
84
Responses to questionnaire to Social Networks Q3 of 17 October 2016, question11.
85
Responses to questionnaire to Social Networks Q3 of 17 October 2016, question 7.
86
Responses to questionnaire to Social Networks Q3 of 17 October 2016, question 10.
87
Responses to questionnaire to Social Networks Q3 of 17 October 2016, question11.
88
Responses to questionnaire to Social Networks Q3 of 17 October 2016, question 10.
89
Responses to questionnaire to Social Networks Q3 of 17 October 2016, question 8.
20
Conversely, those respondents that did not qualify themselves as “professional”
social networks indicated that they do not offer these essential functionalities.
90
Respondents however also noted that they would be able to offer additional
functionalities in the short term and without significant investment.
91
(103) Importantly, in addition to indicating various differentiating and essential
functionalities that qualify and distinguish a social network as “professional”,
respondents significantly emphasised that PSNs are different from other social
networks because of the way they are used.
(104) For instance, one respondent explained that a “personal” social network is used
to “spend time”, whereas a PSN is used to “invest time”. That respondent further
commented that “the user profile of a PSN focusses on highlighting the
professional attributes of a user (e.g. stating the company, industry, function, or
employment status). PSNs even make the filling in of some of those fields
mandatory, setting the tone of the entire product experience accordingly.
Personal Networks instead shift the emphasis to more private matters (e.g. links
to private photo albums)”.
92
In another reply, that same respondent also noted
that “[t]wo functionalities of PSNs are worth stressing explicitly: The user
profile of a PSN focusses on representation of the user as a professional.
Features of the PSN are geared towards highlighting the professional attributes
of a user (e.g. stating the company, industry, function, or employment status)
rather than more private matters (e.g. private photo album). The professional
social network allows the user to reach out to and connect with the broader
member base, irrespective of affiliations to companies/institutions/etc.”.
93
Another respondent also explained that users usually provide more accurate and
complete information and professional and educational background for a PSN,
which is not always the case for other social platforms.
(105) Other respondents’ replies also highlighted this difference in usage. For
instance, one respondent explained that users insert on a PSN a public profile
with work experience, invite colleagues and share professional content, as
opposed to private content and contacts. Another respondent also commented
that it is not usually used to “post private posts and pictures (e.g., vacation
pictures) but rather to post creative content to share with the creative Behance
community.”
94
Finally, one respondent commented that a PSN’s focus is on
professional networking and career/work areas as opposed to the broader social
context. A further indication of the different, more professional-centric, focus of
PSN services as opposed to “personal” SN services can be inferred from the
responses of recruiters to the market investigation, mentioned in paragraphs
(144) and (145) below. Those respondents expressed the view that, both from
the perspective of a jobseeker and of a recruiter, PSN websites are more
substitutable to online recruitment services than other SN services.
95
(106) The Commission further notes that LinkedIn also distinguishes PSNs from
personal social networks on the basis of the different use purposes. For instance,
in a presentation entitled “the Mindset Divide”, LinkedIn notes that the reasons
90
Responses to questionnaire to Social Networks Q3 of 17 October 2016, questions 8, 10 and 11.
91
Responses to questionnaire to Social Networks Q3 of 17 October 2016, question 9.
92
Responses to questionnaire to Social Networks Q3 of 17 October 2016, question 7.
93
Responses to questionnaire to Social Networks Q3 of 17 October 2016, question 10.1.
94
Response to questionnaire to Social Networks Q3 of 17 October 2016, question 7.
95
Responses to questionnaire to Recruiters Q4 of 17 October 2016, questions 12 and 13.
21
for using a PSN are different from those for using a personal social network.
While the former is used to “1. Maintain professional identity, 2. Make useful
contacts, 3. Search for opportunities, 4. Stay in touch”, the latter is used to “1.
Socialize, 2. Stay in touch, 3. Be entertained, 4. Kill time”. The same document
also emphasises differences in the type of content people expect on a PSN (“1.
Career info, 2. Updates on brands, 3. Current affairs”) as opposed to a personal
network (“1. Info on friends and family, 2. Info on personal interests, 3.
Entertainment updates”).
96
(107) A subsequent presentation by LinkedIn, entitled “Spotlight on Content”, also
highlights that PSNs are also different in terms of content. For instance,
LinkedIn notes that “professionals’ response to content varies hugely depending
on which social media platform they are using […] professional content on
Facebook typically invites only superficial engagement, with users more likely
to ‘like’ an item of content than they to actually read it. On LinkedIn and
Twitter, users consume first and share later yet a difference in emphasis
remains. Twitter users prioritise speed of action, retweeting content before
commenting on it in depth; LinkedIn users are far more likely to post their own
comments, adding greater social validation as the content travels across their
networks”.
97
(108) Additionally, the Commission notes that, when asked to identify what social
networks were substitutable to those of LinkedIn, respondents mostly indicated
other providers that present themselves as PSNs, such as XING, Viadeo and
GoldenLine.
98
Most respondents indicated that social networks such as
Facebook and Twitter were not substitutes to LinkedIn. Also in this context,
respondents commented that the latters’ focus was on personal and private use,
rather than professional. For instance, one respondent explained that Facebook
is “a social network where people share content connected with their personal
life, hobbies, family, friends”. Another also reiterated that Facebook is for
personal lives.
(109) Respondents also emphasised that vertical social networks, while also being
geared for professional usage, do not appear to have the same scope of a general
PSN, such as LinkedIn. Rather, they appear to target and serve specific
professions and niches.
99
For instance, respondents commented that a social
network such as Academia is geared for academics and serves the specific use
case of sharing publications. One respondent commented that Academia is a
“[s]pecialist vertical network, which focusses on a specific target group
(academics) via a specialized primary use case (sharing publications); a
complement, not a substitute to LinkedIn’s horizontal PSN”. Thus, an academic
96
LinkedIn document, “the Mindset Divide”, 20 September 2012, slides 10-11, available at
https://business.linkedin.com/content/dam/business/marketing-
solutions/global/en US/site/pdf/wp/linkedin-marketing-solutions-mindset-divide.pdf. The
presentation’s objective was to “deeply understand the motivating factors for how and why people use
personal and professional networks”, as explained at https://business.linkedin.com/marketing-
solutions/blog/t/todays-mindset-divide.
97
LinkedIn document, “the Mindset Divide – Spotlight on content”, 28 March 2014, slide 9, available at
http://www.slideshare.net/linkedineurope/mindset-divide-
whitepaperlinkedin?ref=https://business.linkedin.com/en-uk/marketing-
solutions/blog/posts/2014/How-social-media-puts-professional-content-centre-
stage&from embed lead cta=true&tracking id=97eb1eb1a58743f0.
98
Responses to questionnaire to Social Networks Q3 of 17 October 2016, question 12.
99
Responses to questionnaire to Social Networks Q3 of 17 October 2016, question 12.
22
would not rely on LinkedIn for professional goals. Similarly, respondents
explained that Behance is a platform for the sharing of artistic work, which is
not substitutable to LinkedIn. Finally, one respondent explained that a platform
such as Doximity is a “vertical” social network serving healthcare professionals
specifically. Therefore, it appears that “vertical” social networks, while also
having a “professional” focus, have a narrower scope and target more specific
users and audience than a PSN service provider, such as LinkedIn, Viadeo or
XING.
(110) Finally, most respondents indicated that, for a social network currently not
substitutable with LinkedIn, it would not be possible to develop and become
substitutable to LinkedIn in the short term and without significant
investments.
100
Respondents indicated that the most significant challenge would
be to attract a sufficiently broad user base, which could not be done without
investments. One respondent explained that it may be difficult for a social
network to reconcile a “personal” and a “professional” intended use. Therefore,
while the introduction of additional functionalities may be simpler, as mentioned
in paragraph (102), it appears that the process of transformation of a social
network from “personal” into “professional” social network is more complex, as
it requires a change of approach, usage and general functionalities.
(111) With regard to a possible distinction of SN services on the basis of the device or
OS, most respondents indicated that social networks and PSNs are available
both on mobile and desktop devices, and on the main OSs.
101
(112) The Commission also investigated whether enterprise social networks, as
defined in paragraph (92) above, should be included in the same product market
as PSNs. Respondents to the market investigation all expressed the view that,
for a user, “public” social networks are not substitutable with a “closed” social
network such as an enterprise social network.
102
One respondent explained that
“a ‘closed’ professional social networks is only used for corporate purposes
(such as internal communication or organisation of working methods) and only
by employees”. Another respondent, by referring to “Facebook workplace”,
highlighted the main features of an enterprise social network as: “workplace
account is tied to a user’s company e-mail address and is completely
independent of a Facebook account; access to a company’s workplace limited to
a company’s employees, the company can control who has access to the
workplace and who has not; any activity on Workspace is confined to stay
within the walled-garden of the company”.
(113) The same respondent, in a separate submission, further illustrated the key
differences between a PSN and an enterprise social network. That respondent
explained that an enterprise social network has different functionalities, and is
based on a different business model (closed services offered on a licence basis to
enterprises). An overview of these differences is reproduced in Table 1 below.
100
Responses to questionnaire to Social Networks Q3 of 17 October 2016, question 13.
101
Responses to questionnaire to Social Networks Q3 of 17 October 2016, questions 16 and 17.
102
Responses to questionnaire to Social Networks Q3 of 17 October 2016, question 14.
24
more narrow focus on certain categories of professionals, which distinguish
them from general PSNs. Therefore, the extent to which a vertical social
network could be a substitute to a PSN is unclear. Furthermore, it appears that
the conversion of a “personal” social network into a professional one is not as
immediate and straightforward. There are, however, some indications that
existing social network may find it easier than other possible entrants to develop
into professional social networks.
(116) The Commission finds that it is not appropriate to further segment PSNs
depending on the device or OS on which they are available, as these services are
available on both mobile and desktop devices, and on most OSs. Finally, the
Commission considers that enterprise social networks do not form part of the
same relevant product market as social networks and PSNs.
(117) Therefore, in this Decision the Commission will conduct its assessment on the
basis of the narrowest possible product market, that is to say the market for PSN
services, excluding those services that target specific professions and niches.
3.6.2 Geographic market definition
3.6.2.1 Notifying Party's view
(118) The Notifying Party submits that the market for SN services is worldwide in
scope. The Notifying Party explains that SN services are similar among different
regions and countries, with respect to (i) price, (ii) offered functionalities, (iii)
the platforms and (iv) the OSs operated on. Moreover, all services are Internet-
based, and are available across all platforms, devices and OSs, with no
distinction on the basis of location. Lastly, the Notifying Party states that local
language considerations, user preferences, and local regulatory requirements do
not create any significant obstacles for the cross-border provision of these
services. While certain providers may have a more regional focus, the dimension
remains global for all providers. The Notifying Party submits that, in any event,
the exact geographic market definition can be left open.
3.6.2.2 Commission's assessment
(119) In Facebook/WhatsApp, the Commission noted that SN services are commonly
available worldwide, given the global scope of the Internet. Furthermore, there
were generally no differences in SN services offered in different geographic
regions or countries, although limited adjustments, such as language and minor
functionalities, were present. Other possible identified differences included
marketing costs, legal/regulatory requirements and customers' preferences.
(120) The Commission ultimately concluded that, while there were indications that the
geographic scope of the market could be global, the relevant geographic market
for the assessment of the case was EEA-wide, in line with a more conservative
approach.
104
(121) In the present case, most respondents to the market investigation indicated that
customers do not require significant differences as regards SN services in the
EEA compared to the rest of the world. Similarly, most respondents identified
no significant differences comparing one EEA country to another in terms of
customer requirements.
105
However, respondents also highlighted several
possible differences between EEA countries. Respondents mentioned regulatory
104
Case M.7217 – Facebook/WhatsApp, Commission decision of 3 October 2014, paragraphs 64-68.
105
Responses to questionnaire to Social Networks Q3 of 17 October 2016, question 18.
25
and privacy requirements, as well as cultural and linguistic differences. One
respondent commented that, unlike general or personal social networks, PSN
have a much more localized footprint, which is in inherent to fact that they are
work-oriented. In that respondent’s view, as work is a local matter, PSNs also
have a local dimension, in terms of contacts, job opportunities, language and
professional and educational background. That respondent also added that
privacy considerations may also play an important role as a local requirement.
(122) In addition to the above, the Commission notes the market investigation showed
that several SN services providers appear to specifically target customers from
one or more particular countries in the EEA.
106
Respondents identified several
examples of such social networks, such as XING in Germany/Austria, Viadeo in
France and GoldenLine in Poland. These identified social networks are all social
networks that qualified themselves as “professional”. The Commission therefore
notes that most PSNs active in the EEA appear to focus on a subset of EEA
countries, without having an EEA-wide presence (LinkedIn being the
exception).
(123) Furthermore, when asked whether a company already offering a PSN in one
country could successfully start this in another country without significant
investments, most respondents answered negatively.
107
One respondent
explained that when a PSN is limited to one country, expanding to another
country requires time and resources. Another respondent mentioned the
examples of XING and Viadeo, which are not successful outside of Germany
and France respectively, given their national focus. These responses indicate
that certain barriers between different EEA countries exist with respect to PSNs.
(124) Based on the results of the market investigation with regard to PSNs, the
Commission notes that the differences in terms of language, functionalities,
legal/regulatory requirements and customers' preferences among EEA countries
appear to be relevant for PSNs and influence the geographic scope of their
activities. This may be linked, among other things, to the fact that professional
relations and employment tend to have a more local dimension and use.
(125) Based on the above, the Commission considers that the relevant geographic
market for PSN services for the assessment of this case is national, in line with a
more conservative approach.
3.7 Online recruitment services
(126) Online recruitment services enable individuals, who are searching for
employment or who could be induced to consider an opportunity, to connect
with recruiters who seek to fill vacancies. A job seeker typically creates a profile
and resume on the online recruitment services and may then sign up for updates
in relation to relevant job opportunities.
(127) Providers of online recruitment solutions maintain databases of profiles and
resumes for use by both job seekers and recruiters and create a system in which
job seekers and recruiters can be matched according to mutually defined criteria.
(128) LinkedIn offers online recruitment services. Microsoft is not active in this
sector.
106
Responses to questionnaire to Social Networks Q3 of 17 October 2016, question 19.
107
Responses to questionnaire to Social Networks Q3 of 17 October 2016, question 20.
26
3.7.1 Product market definition
3.7.1.1 Notifying Party's view
(129) The Notifying Party submits that there is an overall market for recruitment
solutions, which should not be further divided between offline and online
recruitment solutions. The Notifying Party argues that a jobseeker will make use
of both offline and online recruitment solutions, and that these two solutions are
not materially different in terms of process. The Notifying Party further submits
that companies that offer jobs or recruiters who offer jobs on the companies’
behalf will also use both types of recruitment solutions in their recruitment
process. In the Notifying Party’s view, while both may have a preference for
offline or online recruitment, it is unlikely that they would not generally
consider all available options to create matches between candidates looking for
employment and recruiters or companies offering jobs. In a forward-looking
perspective, the Notifying Part submits that any differences between online and
offline recruitment services will tend to disappear, as recruitment companies
pursue models based on both channels.
(130) Moreover, the Notifying Party submits that, if there were to be a separate market
for online recruitment services, this market should not be further segmented by
business model or industry.
(131) With respect to a possible segmentation by business model, the Notifying Party
explains that this would not be appropriate. While online recruitment services
are offered based on many different business models that often integrate other
services (such as career advice, interview preparation, resume building, job
alerts, professional networking opportunities and online advertising
opportunities), all providers offer the same basic functionalities for both sides of
the online recruitment market, irrespective of the business model. Accordingly,
in the Notifying Party’s view, providers of online recruitment services include
companies and recruitment agencies advertising job opportunities on their own
site, online versions of newspaper classifieds, generalist jobsites operating
across the full spectrum of industry categories and qualification levels, sector-
focused job sites, job aggregators, and SN services such as Facebook and
LinkedIn.
(132) As regards a possible segmentation of online recruitment solutions on the basis
of the industry sector, the Notifying Party submits that this would not be
appropriate, neither from the point of view of the company (or recruiter) seeking
to fill a vacancy or that of the job seeker. The Notifying Party explains that a
company seeking to hire employees would nevertheless generally recruit across
multiple sectors. For instance, a company seeking to fill vacancies (directly or
via a recruiter) would not rely on LinkedIn only with regard to candidates from
one particular sector. Moreover, it would not be viable for LinkedIn’s Talent
Solutions business to differentiate between jobs in one sector and jobs in another
sector and, based on this differentiation, charge different prices for its
recruitment tools. In the Notifying Party’s view, there is no need to differentiate
according to sectors from a job seeker’s perspective either. While certain
candidates will seek jobs according to their qualification, and eventually rely on
specific recruitment tools (for instance, an engineer may take particular interest
in a niche online recruitment solution specialised in jobs for engineers), this
does not mean that those candidates would not consider other providers of
recruitment solutions.
27
(133) The Notifying Party concludes that in any event the precise scope of the product
market can be left open.
3.7.1.2 Commission's assessment
(134) In Randstad/VNU/JV, the Commission considered that online recruitment
services could be distinguished from offline recruitment services, but ultimately
left the market definition open.
108
In the same decision, the Commission also
considered the market for online job advertisement as a separate market from
general advertising, but also ultimately left the definition open.
109
(135) In Randstad Holding / Monster Worldwide, the Commission analysed the
market for online job board services, which employers use to find job seekers by
offering vacancies on the online job board or searching through candidates’ CVs
based on a number of search criteria, but ultimately left the market definition
open.
110
(136) Respondents to the market investigation in the present case identified several
functionalities as being essential for recruitment services, including: access to
databases of profiles and resumes, the possibility to provide online job
advertisements, the creation of a profile and resume, the possibility to sign up
for job alerts, and CV-search services. Other features, such as career advice,
interview preparation and face-to-face interviews, were considered less
important.
111
(137) When asked to comment on the degree of substitutability between online and
offline recruitment solutions, the view of the majority of respondents was that
the two channels were only substitutable to a certain extent, and mostly
complementary, with online becoming increasingly important.
(138) From the perspective of the jobseeker, most respondents were of the opinion that
offline recruitment tools cannot substitute online recruitment tools, but rather
complement them. Some respondents highlighted the different features of the
two channels. For instance, one respondent explained that online recruitment
services are more noticeable and allow faster and more comprehensive access to
information, while offline services are used purely for gathering information and
can be replaced by online services. Another respondent commented that online
services are an impersonal "behind a computer" only experience, whereas
offline services consist of a personal contact and relationship. Another explained
that the medium and mode of interaction are different. Other respondents
explained that job seekers would use the two channels in a complementary
manner, as each channel is a means to connect a suitable applicant with a
suitable role. Finally, certain respondents highlighted the easier and broader
accessibility of online recruitment sources, the higher frequency of updates, the
broader scope in terms of search and access, as well as other distinctive
functions.
112
(139) Most respondents expressed the view that online recruitment solutions are
increasing in importance and use, while offline job ads are decreasing, as
108
Case M.2057 – Randstad/VNU/JV, Commission decision of 30 August 2000, paragraphs 11-13.
109
Case M.2057 – Randstad/VNU/JV, Commission decision of 30 August 2000, paragraphs 14-15.
110
Case M.8201 – Randstad Holding/Monster Worldwide, Commission decision of 26 October 2016,
paragraphs 18-20.
111
Responses to questionnaire to Recruiters Q4 of 17 October 2016, question 8.
112
Responses to questionnaire to Recruiters Q4 of 17 October 2016, question 10.1.
28
jobseekers tend more and more to rely on online channels rather than offline
channels, given the superiority of the former in terms of efficiency, scope, speed
and accessibility. One respondent explained that while offline recruiting services
can still somewhat substitute online services, the latter are much more powerful
and are rapidly replacing almost all offline services. Another respondent stated
that online services cannot be replaced by offline services anymore, except for
lower level and very local jobs.
113
(140) Respondents expressed the same view when considering the perspective of the
recruiter. Also in this context, responses to the market investigation emphasised
the different features and the complementary use of the online and offline
channels, with online gaining an increasing importance in usage and scope.
(141) For instance, one respondent explained that online and offline recruitment
services complement each other and can be used together or separately
depending on the type of role. Another commented that online recruitment
services give the flexibility to announce open positions to a wider audience,
while offline recruitment services focus more on the specific range of people
who are suitable to a role, and include face to face interviews, personal
interaction and more in-deep analysis of a candidate’s qualification. Another
respondent also stated that “[f]rom an employer’s perspective online and offline
recruitment services can only be complementary. Thereby employers prefer
using online recruitment services. To reach as many applicants as possible job
offers are mostly published on online platforms and online job exchanges.
Online Networks are another possibility of recruiting. Using online channels
allows to reach a much wider group of applicants. […] Offline services are
supplementally used by the employers, whereby offline services such as phone
contacts and job interviews, but also actions like speed dating, job fairs etc. are
explicitly asked for and used, as the recruiting process continues”.
114
(142) Most respondents highlighted the higher speed and broader scope of online
recruiting, and expressed the view that offline recruiting is becoming more
marginal, while online recruiting channels become increasingly important.
115
(143) Therefore, based on the results of the market investigation, the Commission
considers that online recruitment services should be distinguished from offline
recruitment services, in light of the different use, features and functions, which
make the two services complementary rather than substitutable.
(144) Respondents to the market investigation expressed mixed views with regards to
a possible segmentation of online recruitment services based on the business
model, including the business models mentioned in paragraph (131) above.
From the jobseeker’s perspective, respondents to the market investigation had
divided opinions as to whether business models, such as sector-focused job sites,
online versions of newspaper classifieds or companies and recruitment agencies
advertising job opportunities on their own site could actually be substitutable to
online recruitment services such as those offered by LinkedIn. Respondents
more clearly indicated that in their view non-professional SN websites are not
perceived by a jobseeker as an online recruitment tool. Respondents also were of
the opinion that PSN websites offered services that were substitutable to those
113
Responses to questionnaire to Recruiters Q4 of 17 October 2016, question 10.1.
114
Responses to questionnaire to Recruiters Q4 of 17 October 2016, question 10.2.
115
Responses to questionnaire to Recruiters Q4 of 17 October 2016, question 10.2.
29
of an online recruitment tool. One respondent explained that PSN websites
“offer job seekers the ability to view job openings and apply for jobs. They also
enable recruiters to directly contact potential candidates to gauge their interest
in a particular opening.”
116
(145) Similarly, from the recruitment perspective, respondents also gave mixed views.
On the question whether different business models such as online versions of
newspaper classifieds, generalist jobsites, sector-focused jobsites, or companies
and recruitment agencies advertising job opportunities on their own site would
be substitutable with online recruitment services such as those of LinkedIn,
respondents of the market investigation were divided. Similarly to the
jobseekers’ perspective, the respondents indicated that non-professional SN
websites are not substitutable with online recruitment services, whereas PSN
websites were found to be more substitutable. One of the respondents explained
this with the fact that most other PSN websites offer similar solutions. One of
the respondents explained this with the different focus of such websites.
117
(146) With respect to a possible segmentation of online recruitment services on the
basis of the industry sector, most respondents to the market investigation
indicated that recruitment services differ depending on the relevant industry
sector.
118
Respondents indicated that the differentiating factors may include
cannels of recruiting and methods of selection, customer expectations, education
and experience level, and the level of technicality of the industry. One
respondent commented that the nature of the industry and the types of job may
even require a different route for recruiting, for instance lower skilled positions
can be filled by means of job advertising, whereas more specialist and skilled
roles may require a more proactive approach and reliance on professional
recruitment. However, most respondents also indicated that they charge the
same price for their recruitment services to customers, irrespective of the
relevant industry.
119
(147) For the purpose of this Decision, the Commission therefore considers that the
relevant product market is that for online recruitment services, which should be
distinguished from offline recruitment services. The question whether online
recruitment services should be further segmented depending on the business
model or the relevant industry sector can be left open, as the Transaction does
not raise serious doubts as to its compatibility with the internal market with
respect to online recruitment services irrespective of the exact product market
definition.
3.7.2 Geographic market definition
3.7.2.1 Notifying Party's view
(148) The Notifying Party submits that the market for recruitment solutions should be
either regional or national in scope. The Notifying Party explains that it is
unlikely that jobseekers would be willing to move across countries, due to
language barriers, education, qualifications and administrative hurdles. This
results in the fact that recruiters will typically attempt to recruit employees
116
Responses to questionnaire to Recruiters Q4 of 17 October 2016, question 12.
117
Responses to questionnaire to Recruiters Q4 of 17 October 2016, question 13.
118
Responses to questionnaire to Recruiters Q4 of 17 October 2016, question 14.
119
Responses to questionnaire to Recruiters Q4 of 17 October 2016, question 15.
30
within a single country. Finally, the Notifying Party ultimately concludes that
the precise definition for the market can be left open.
3.7.2.2 Commission's assessment
(149) In previous cases, the Commission left open the exact geographic scope for the
provision of online recruitment services.
120
(150) In the present case, the responses to the market investigation were mixed. There
was no clear indication on whether the provision of online recruitment services
in the EEA differs from the rest of the world. Similarly, respondents did not take
a clear view on whether online recruitment services differ between EEA
countries.
121
Some respondents indicated that relevant differentiating factors
include regulatory requirements, language, and local knowledge.
(151) For the purpose of this Decision, the precise scope of the geographic market for
the provision of online recruitment services can be left open, as the Transaction
does not raise serious doubts as to its compatibility with the internal market
under any geographic market definition.
3.8 Online advertising services
(152) The Parties both offer online advertising services. Microsoft provides both non-
search and search advertising services, LinkedIn only offers online non-search
advertising services.
3.8.1 Product market definition
3.8.1.1 Notifying Party's view
(153) The Notifying Party submits that online non-search advertising should not be
further segmented according to the type of service and device platform, the
audience type or by the fact that the advertising is targeted.
(154) Firstly, the Notifying Party states that the market for online non-search
advertising should not be subdivided by website type. In particular, online
advertising on SN websites should not be identified as a separate product
market. The Notifying Party argues that, from an advertiser’s perspective, online
advertising on SN websites does not differ from online advertising on other
websites. Moreover, the ad formats and the products which are offered on both
types of websites are largely similar. This results in a high degree of demand
substitution for online advertising on different types of websites. The Notifying
Party submits that this substitutability has been further increased by the
introduction of ad intermediation tools, which target ads to specific users
wherever they are on the Internet. In case that SN websites increase their prices
for advertising space, these tools allow advertisers to switch to other websites
and continue targeting the same customers.
(155) Secondly, the Notifying Party submits that the market should not be further
subdivided according to device platforms, such as mobile or PC. The reason for
this is the growing convergence between these platforms in recent years. In the
Notifying party’s view, this is shown by the fact that intermediation providers
120
Case M.2057 – Randstad/VNU/JV, Commission decision of 30 August 2000, paragraph 10; Case
M.8201 – Randstad Holding/Monster Worldwide, Commission decision of 26 October 2016,
paragraphs 28-29.
121
Responses to questionnaire to Recruiters Q4 of 17 October 2016, question 18.
31
have adjusted their ad intermediation tools to enable advertisers to manage
campaigns across different platforms, such as desktop and mobile.
(156) Thirdly, the Notifying Party submits that no distinction within the non-search
online advertising market should be made based on audience types, such as
businesses or customers. That is because web publishers selling online
advertising will sell ads independently of the audience type. The Notifying Party
submits that, although certain web publishers might have a more specific
audience, and thus be more attractive for certain types of advertising, advertisers
target their audience through numerous and various channels. For instance,
LinkedIn may be more attractive for B2B advertising, but advertisers can reach
the same audience also through different, non-professional, websites.
Furthermore, the Notifying Party argues that it would be inappropriate to
attempt to differentiate websites depending on their audience type, as this
segmentation would result in an infinite amount of submarkets, with an unclear
distinction of the precise target audience. Finally, the Notifying Party adds that
the increase of ad-targeting tools enhances the interchangeability of websites for
advertising purposes. This means that advertisers could easily switch from one
website to others, in case of an increase of advertising price, and keep targeting
the same audience.
(157) Fourthly, the Notifying Party states that no distinction should be made on the
basis of targeting. The Notifying Party submits that the large majority of
respondents to the market investigation in Telefónica UK/Vodafone
UK/Everything Everywhere/JV in 2012
122
considered that targeted marketing
messaging constituted a separate market.
123
However, the Notifying Party argues
that there have been rapid technological developments since then and that
targeting is a prerequisite to sell online advertising today. Therefore, all main
providers of display advertising solely offer targeted display advertising, based
on different criteria. Therefore, no meaningful distinction can be drawn between
targeted and non-target online non-search advertising anymore.
(158) The Notifying Party submits that in any event the exact scope of the product
market for online non-search advertising can be left open.
3.8.1.2 Commission's assessment
(159) In previous decisions, the Commission distinguished between the provision of
online and offline advertising space.
124
The Commission further considered
whether the market for online advertising could be sub-segmented into search
and non-search advertising, but ultimately left this question open.
125
(160) In Facebook/WhatsApp, the Commission noted, on the basis of the market
investigation, that a further sub-segmentation of online advertising between
search and non-search advertising could be drawn, as respondents to a large
extent considered that search and non-search ads are not substitutable since they
122
Case M.6314 – Telefonica UK/Vodafone UK/Everything Everywhere/JV, Commission decision of 4
September 2002.
123
Form CO, Section 6, paragraph 357.
124
Case M.5727 – Microsoft/Yahoo! Search Business, Commission decision of 18 February 2010,
paragraph 61; Case M. 4731 – Google/DoubleClick, Commission decision of 11 March 2008, recitals
45-46; 56.
125
Case M.5727 – Microsoft/Yahoo! Search Business, Commission decision of 18 February 2010,
paragraphs 71-75; Case M. 4731 – Google / DoubleClick, Commission decision of 11 March 2008,
recitals 49-56.
32
serve different purposes and, as a result, most advertisers would not be likely to
switch from one type to another.
126
The Commission also examined whether a
separate product market should be defined for the provision of online non-search
advertising services on SN websites, and whether it would be appropriate to
distinguish online advertising on the basis of the platform (PC versus mobile),
but did not reach a conclusion on these two possible distinctions.
127
The
Commission ultimately left open the questions as to whether the market for
online advertising should further be segmented on the basis of search and non-
search, whether a separate segment for advertising on SN websites could be
identified, and whether a distinction should be drawn on the basis of the
platform.
128
(161) For the purpose of this Decision, the question whether the market can be further
segmented between search and non-search advertising, and whether there is a
specific segment for advertising on SN websites and by device platform can be
left open, as the Transaction does not raise serious doubts as to its compatibility
with the internal market, irrespective of the product market definition.
3.8.2 Geographic market definition
3.8.2.1 Notifying Party's view
(162) The Notifying Party submits that the exact scope of the geographic market can
be left open.
3.8.2.2 Commission's assessment
(163) In previous decisions, the Commission concluded that the market for online
advertising is to be divided alongside national or linguistic borders within the
EEA, although it ultimately left the geographic market definition open in one
case.
129
Factors pointing to a national or linguistic geographic market definition
included customers' purchasing preferences and languages, and the presence of
support and sales networks located at national level
(164) For the purpose of this Decision, the Commission concludes that the online
advertising market and its possible sub-segments should be defined as national
in scope or alongside linguistic borders within the EEA.
4 COMPETITIVE ASSESSMENT
(165) According to the Notifying Party, by acquiring LinkedIn, Microsoft aims to
expand the standalone business of LinkedIn and integrate LinkedIn’s services
and products within its own offering, enriching both Microsoft and LinkedIn
user experience.
(166) The Transaction will therefore mainly combine Microsoft's and LinkedIn’s
complementary offerings. While both Microsoft and LinkedIn provide some
online advertising services, this is not their core business. Hence the Transaction
126
Case M.7217 – Facebook/WhatsApp, Commission decision of 3 October 2014, paragraph 76.
127
Case M.7217 – Facebook/WhatsApp, Commission decision of 3 October 2014, paragraphs 77-78.
128
Case M.7217 – Facebook/WhatsApp, Commission decision of 3 October 2014, paragraph 79.
129
Case M.6314 – Telefónica UK/Vodafone UK/Everything Everywhere/JV, Commission decision of 4
September 2012, recitals 226-229; Case M.5727 – Microsoft/Yahoo! Search Business, Commission
decision of 18 February 2010, paragraphs 91-93; Case M.4731 – Google/DoubleClick, Commission
decision of 11 March 2008, recitals 83-84; Case M.7217 – Facebook/WhatsApp, Commission decision
of 3 October 2014, paragraph 83.
33
only gives rise to limited horizontal overlaps while, at the same time, giving rise
to a number of non-horizontal relationships.
4.1 Assessment of horizontal effects
(167) Microsoft and LinkedIn are both active in the provision of online advertising
services.
130
(168) However, as mentioned in paragraph (152) above, while Microsoft provides
both non-search and search advertising services, LinkedIn is active only in non-
search advertising services. Therefore, the Parties' activities overlap only with
respect to non-search advertising services.
(169) In the overall market for online advertising services, the Parties’ combined
market share post-Transaction would be low. Based on the Parties’ information,
in 2015 worldwide Microsoft’s share of digital ads revenues amounted to [0-
5]% and LinkedIn’s to [0-5]%.
131
(170) With respect to online search advertising services specifically, where LinkedIn
is not active, Microsoft, with a market share no greater than [5-10]% in any
EEA Member State, is a small player, well behind Google.
(171) With respect to online non-search advertising, both Microsoft and LinkedIn are
active. However, the Parties are small players, with a combined share by
revenue no greater than [5-10]% in any EEA Member State (with the exception
of France, where the combined share is [5-10]%).
132
The online non-search
advertising sector is fragmented and led by Facebook and Google, each with
revenues several times greater the Parties’ combined revenues.
(172) Moreover, Microsoft’s presence in online non-search advertising is limited to
advertising space offered on its own properties (websites), since it has
outsourced other non-search advertising activities to Verizon's AOL until 2025,
thus withdrawing from selling online display ads.
(173) Finally, the Commission notes that LinkedIn’s online non-search advertising
business specifically relates to non-search advertising services on (professional)
social networks. Based on the Parties’ estimates,
133
in a possible market segment
including all social networks, LinkedIn would be a much smaller player in terms
of advertising revenues than Facebook or Twitter. LinkedIn’s share in this space
would be below 5% in each key EEA country, in most cases being limited to [0-
5]%, well behind the market leader Facebook, whose shares are above [60-70]%
in most countries.
(174) LinkedIn would arguably have a strong position in a hypothetical market limited
to online non-search advertising on PSNs. However, the Transaction would not
have any impact on LinkedIn’s position, given that Microsoft is not active in
online non-search advertising on PSNs.
130
The Transaction does not give rise to other horizontal overlaps. In addition to online advertising
services, as mentioned in paragraph (3) above, LinkedIn offers online recruitment solutions, PSN
services, which Microsoft does not offer.
131
Form CO, Section 6, paragraph 583.
132
Form CO, Section 6, Table 26.
133
Form CO, Section 6, paragraph 588.
34
(175) In addition to the low market shares, the Commission further notes that
respondents to the market investigation confirmed that the Transaction did not
raise concerns with respect to online advertising.
134
(176) Moreover, the Transaction does not raise competition concerns resulting from
the possible post-merger combination of the "data" (essentially consisting of
personal information, such as information about an individual's job, career
history and professional connections, and/or her or his email or other contacts,
search behaviour etc. about the users of their services) held by each of the
Parties in relation to online advertising.
(177) As a preliminary remark, it should be noted that any such data combination
could only be implemented by the merged entity to the extent it is allowed by
applicable data protection rules. In this respect, the Commission notes that,
today, Microsoft and LinkedIn are subject to relevant national data protection
rules with respect to the collection, processing, storage and usage of personal
data, which, subject to certain exceptions, limit their ability to process the
dataset they maintain.
135
Currently, the data protection rules of the EU Member
State(s) where Microsoft and LinkedIn have their registered seat and/or where
they have subsidiaries processing data apply. Since LinkedIn with regard to the
EU is not processing personal data of its customers outside of Ireland where it
has its registered seat, it is currently only subject to Irish data protection rules.
Likewise, the Notifying Party submits that Microsoft is currently subject to Irish
data protection rules.
(178) Moreover, the Commission notes that the newly adopted General Data
Protection Regulation ("GDPR"),
136
which will establish one single set of rules
for companies processing personal data in the EU and entered into force on 24
May 2016, will apply from 25 May 2018. The GDPR provides for a harmonised
and high level of protection of personal data and fully regulates the processing
of personal data in the EU, including inter alia the collection, use of, access to
and portability of personal data as well as the possibilities to transmit or to
transfer personal data. This may further limit Microsoft's ability to have access
and to process its users' personal data in the future since the new rules will
strengthen the existing rights and empowering individuals with more control
over their personal data (i.e. easier access to personal data; right to data
portability; etc.).
(179) Assuming such data combination is allowed under the applicable data protection
legislation, there are two main ways in which a merger may raise horizontal
issues as a result of the combination under the ownership of the merged entity of
two datasets previously held by two independent firms. First, the combination of
two datasets post-merger may increase the merged entity's market power in a
hypothetical market for the supply of this data or increase barriers to
134
Responses to questionnaire to advertisers Q8 of 14 October 2016, questions 2 and 3.
135
Directive 95/46/EC of the European Parliament and the Council of 24 October 1995 on the protection
of individuals with regard to the processing of personal data and the free movement of such data
("Data Protection Directive"), OJ L 281, 23.11.1995, pp.31-39. Article 7(b)-(f) of the Data Protection
Directive lays out the situations where the personal data of a data subject may be processed without the
consent of the data subject.
136
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" or "GDPR"),
repealing the Data Protection Directive, OJ L 119, 4.5.2016, pp. 1-88.
35
entry/expansion in the market for actual or potential competitors, which may
need this data to operate on this market. Competitors may indeed be required to
collect a larger dataset in order to compete effectively with the merged entity
than absent the merger. Second, even if there is no intention or technical
possibility to combine the two datasets, it may be that pre-merger the two
companies were competing with each other on the basis of the data they
controlled and this competition would be eliminated by the merger.
(180) In the case at hand, however, the Transaction does not give rise to this type of
concerns in relation to online advertising. First, Microsoft and LinkedIn do not
make available their data to third parties for advertising purposes, with very
limited exceptions.
137
Second, the combination of their respective datasets does
not appear to result in raising the barriers to entry/expansion for other players in
this space, as there will continue to be a large amount of internet user data that
are valuable for advertising purposes and that are not within Microsoft's
exclusive control. Third, the Parties are small market players and compete with
each other only to a very limited extent in online advertising and its possible
segments.
(181) Therefore, the Transaction does not raise serious doubts as regards its
compatibility with the internal market with respect to online advertising.
4.2 Assessment of non-horizontal effects
4.2.1 Legal framework
4.2.1.1 Vertical non-coordinated effects
(182) According to the Non-Horizontal Guidelines,
138
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.
139
Such foreclosure may discourage entry or expansion of competitors
or encourage their exit.
140
(183) 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.
141
(184) 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
137
Under partnership agreement whereby Microsoft outsourced its online non-search advertising services
to AOL, for AOL to provide the outsourced services, […]. See Form CO, Annex 7.
138
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, p. 6-
25.
139
See Non-Horizontal Guidelines, paragraph 18.
140
See Non-Horizontal Guidelines, paragraph 29.
141
See Non-Horizontal Guidelines, paragraph 30.
36
foreclosure strategy would have a significant detrimental effect on
competition.
142
4.2.1.2 Conglomerate non-coordinated effects
(185) According to the Non-Horizontal Guidelines, in the majority of circumstances,
conglomerate mergers will not lead to any competition problems.
143
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 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.
144
(186) In assessing the likelihood of anticompetitive foreclosure effects, the
Commission examines, first, whether the merged firm would have the ability to
foreclose its actual or potential competitors, second, whether it would have the
economic incentive to do so and, third, whether a foreclosure strategy would
have a significant detrimental effect on competition, thus causing harm to
consumers.
145
4.2.2 Foreclosure of competing providers of CRM software solutions
4.2.2.1 Industry overview and trends
(a) CRM software solutions
(187) The worldwide market for CRM software solutions has being growing in the last
years and it is set to grow further in the near future. Figure 1 below shows the
evolution of the aggregated revenues of providers of CRM software solutions in
the period 2007 to 2015 as reported by the third party analyst Gartner.
146
142
See Non-Horizontal Guidelines, paragraph 32.
143
See Non-Horizontal Guidelines, paragraph 92.
144
See Non-Horizontal Guidelines, paragraphs 91 and 93.
145
See Non-Horizontal Guidelines, paragraph 94.
146
Form CO, Annex 10, Figure 1; Notifying Party's response to RFI n. 24, question 5; Schaeffer, C.,
"CRM Software Market Share Report 2016", CRMSearch, available at
http://www.crmsearch.com/crm-software-market-share.php, viewed on 10 November 2016.
37
Figure 1: Evolution of the worldwide market for CRM software solutions 2007-2020
(revenues, billion USD)
Source: Gartner, Notifying Party's estimates.
(188) Figure 1 shows the steep growth of revenues for CRM software solutions at
worldwide level, set at 16% in 2015 over 2014 and 12.3% in 2016 over 2015.
147
The growth of the market for CRM software solutions was even greater in the
EEA, where it was above 24% in 2015 over 2014.
148
(189) On the basis of the data collected by Gartner, the significant growth of the
market for CRM software solutions is driven by the growth of the revenues for
cloud services, which grew by 27% in 2015 over 2014, while revenues for on-
premises new licenses declined by 1% over the same period,
149
as shown in
Figure 2 below.
147
Compared to 2015, in 2016 the revenues for CRM software solutions grew more than any other
segment within EAS, which overall grew only 3.4%. Form CO, Annex 10, Table 1.
148
Form CO, Annex 10, Table 3. Gartner does not report figures for the EEA, so this is a proxy based on
Western and Eastern Europe data.
149
Form CO, Annex 10, page 12 and Figure 1.
0
10
20
30
40
50
60
2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020
40
Figure 5: CRM competitive landscape
Source: Form CO, Section 6, paragraph 492.
(194) In this fragmented competitive landscape, the major players are Salesforce,
SAP, Oracle, Microsoft, and Adobe.
(195) Salesforce entered the CRM software solutions market in 1999 and grew
significantly ever since. Currently Salesforce is the largest vendor overall in the
market for CRM software solutions worldwide (the second in the EEA), as well
as in the CSS and Sales segments (both worldwide and in the EEA). It is also
growing in the marketing segment. Salesforce pioneered the use of cloud
computing and all of its CRM software solutions run entirely in the cloud. As a
result, today Salesforce is the leader by revenues in the provision of CRM cloud
services, as illustrated by Figure 6 below.
Figure 6: Revenues of the main providers of CRM software solutions - cloud segment
(worldwide, billion USD)
Source: Form CO, Annex 10.
(196) In 2010 Salesforce acquired Jigsaw, rebranded Data.com, an online business
directory of companies and business professionals. In September 2016,
41
Salesforce introduced Einstein, a functionality powered by machine learning
("ML") and predictive analytics built into the core of Salesforce's platform,
which "leverages all the data in Salesforce—customer data; activity data from
Chatter [Salesforce's enterprise social network], email, calendar and
ecommerce; social data streams such as tweets and images; and even [Internet
of things] signals."
154
It has a premium price positioning. While over 70% of its
revenues come from North America, Salesforce is expanding its footprint within
Europe with additional data centers, new offices and sales staff.
155
Salesforce is
considered to have the highest priced services in the market.
156
CRM
competitors and customers responding to the market investigation have rated
Salesforce as the most competitive player with respect to all parameters of
competition but price.
157
(197) Headquartered in Germany, SAP is the leading provider of EAS and in the
overall market for CRM software solutions in the EEA. It is the second largest
player in the market for CRM software solutions worldwide. It offers a variety
of CRM software solutions (on premises, cloud and hybrid) in the following
segments: (i) Sales; (ii) Marketing; (iii) DC; and (iv) CSS. SAP incorporated
ML capabilities in their products.
158
In particular its application "Automated
SalesForecast" offers ML based on data in CRM cloud software solution and
unstructured text from emails and the Web.
159
CRM competitors and customers
responding to the market investigation have rated SAP third in terms of
competitiveness of its offering with respect to the functionalities offered.
160
(198) Oracle is the third player in the market for CRM software solutions, both
worldwide and in the EEA. Oracle offers a complete and integrated CRM
solution, including Sales, Marketing, DC and CSS as well as social and pricing
information in relation to an organisation’s customers.
161
In 2016, to improve its
offering in EAS, including CRM software solutions, Oracle announced
"Adaptive Intelligent Applications", providing ML powered by digital consumer
and business data from Oracle’s Data Cloud.
162
Oracle’s Data Cloud provides
access to over 5 billion global consumer profiles and 400 million B2B
163
profiles from more than 1 500 data providers.
164
Based on the Gartner 2015
report, Oracle is offering aggressive pricing, with most clients being able to get
154
http://www.salesforce.com/company/news-press/press-releases/2016/09/160919.jsp
155
Gartner Report: Market Share Analysis -- Customer Relationship Management Software, Worldwide
2015. Form CO, Annex 10, pages 17-18; https://www.salesforce.com/
156
Gartner Report: Market Share Analysis -- Customer Relationship Management Software, Worldwide
2015. Form CO, Annex 10, page 19.
157
Responses to questionnaire to CRM Competitors Q5 of 14 October 2016, question 26, and
questionnaire to CRM Customers Q7 of 14 October 2016, question 26.
158
Agreed non-confidential minutes of the conference call with SAP held on 3 November 2016, point 2.
159
SAP "Machine learning goes mainstream" available at: http://news.sap.com/machine-learning-goes-
mainstream/
160
Responses to questionnaire to CRM Competitors Q5 of 14 October 2016, question 26, and
questionnaire to CRM Customers Q7 of 14 October 2016, question 26.
161
Oracle CRM sales prospector, http://www.oracle.com/us/products/applications/social-
crm/054233 html
162
Oracle "Adaptive Intelligent Apps" available at https://cloud.oracle.com/en US/adaptive-intelligent-
apps. See also: "Oracle Unveils Its Next Generation Cloud Strategy: Intelligent Applications", at
https://www.oracle.com/corporate/pressrelease/oracle-unveils-intelligent-applications-091916.html
163
See https://www.oracle.com/corporate/pressrelease/oracle-launches-b2b-data-marketplace-
081216.html
164
See https://www.oracle.com/corporate/pressrelease/oracle-revolutionizes-data-analytics-091916.html
42
attractive discounts and leverage promotional programs.
165
CRM competitors
and customers responding to the market investigation have rated Oracle the
second most competitive player in terms of price (on par with Adobe) and third
in terms of competitiveness of its offering with respect to the functionalities
offered and the ease of use.
166
(199) Microsoft is the fourth operator in the market with revenues of approximately
USD 1 billion with its Dynamics 365 solution available both on premises and as
cloud solution. While Microsoft's CRM software solutions cover all CRM
segments, the majority of its revenues are generated in the Sales segment. In
spring 2016, Microsoft launched several ML services for CRM users based on
customers' organization data and accessible third-party data. CRM competitors
and customers responding to the market investigation have rated Microsoft as
the most competitive player with respect to price and ease of use (the latter on
par with Salesforce) and second to Salesforce in terms of functionality and
support.
167
(200) Adobe is a U.S. software company that offers creative, marketing and document
management solutions. In the CRM software solution market, Adobe's cloud
offering is limited to the Marketing segment.
168
Through Adobe Analytics,
Adobe offers solutions for applying real-time analytics and detailed
segmentation across all marketing channels.
169
CRM competitors and customers
responding to the market investigation have rated Adobe the second most
competitive player in terms of price (on par with Oracle) and reputation and
third with respect to ease of use.
170
(201) While the market for CRM software solutions appears to be competitive when
customers look for a CRM software solution for the first time, the ability of
customers to switch providers of CRM software solutions once the choice has
been made appears to be limited.
171
Indeed, all CRM customers responding to
the market investigation have indicated that switching is not easy. While several
examples of customers switching have been brought to the Commission's
attention,
172
it appears that switching implies for customers certain non-trivial
costs, such as the cost of training users on a new system, the cost of integrating
the new CRM software solution with the existing systems, etc. Interoperability
issues have been raised with respect to the transfer of the customer's
organization data from one CRM software solution to another.
173
In this context,
165
Gartner Report: Market Share Analysis -- Customer Relationship Management Software, Worldwide
2015. Form CO, Annex 10, page 21.
166
Responses to questionnaire to CRM Competitors Q5 of 14 October 2016, question 26, and
questionnaire to CRM Customers Q7 of 14 October 2016, question 26.
167
Responses to questionnaire to CRM Competitors Q5 of 14 October 2016, question 26, and
questionnaire to CRM Customers Q7 of 14 October 2016, question 26.
168
Adobe Marketing Cloud, available at: http://www.adobe.com/uk/marketing-cloud.html
169
Adobe Analytics, available at: http://www.adobe.com/marketing-cloud/web-analytics html
170
Responses to questionnaire to CRM Competitors Q5 of 14 October 2016, question 26, and
questionnaire to CRM Customers Q7 of 14 October 2016, question 26.
171
One possible explanation for this is that future profits to be realised on customers not switching
suppliers are competed away ex ante when CRM software solutions providers compete for new clients.
172
See, in particular, responses to questionnaire to CRM Competitors Q5 of 14 October 2016, question
27.
173
Responses to questionnaire to CRM Customers Q7 of 14 October 2016, question 18, questionnaire to
additional CRM Customers Q18 of 28 October 2016, question 3, and questionnaire to CRM
Competitors Q5 of 14 October 2016, question 27.
43
IBM explained that "[o]ur sales processes and tooling are now very intertwined.
It would be a multi-year effort to move our sales force from one platform to
another"
174
, while Mapei stated that "it is difficult to replace only single
functionality [sic] inside the CRM platform."
175
A CRM supplier stated that
"[f]or customers of CRM software solutions, switching provider is not easy and
requires significant cost and time."
176
(b) Sales intelligence solutions
(202) The sales intelligence solutions market is highly fragmented and features many
providers, such as Dun & Bradstreet, Data.com, Nimble, Avention,
DiscoverOrg, ZoomInfo, InsideSales, InsideView, Madison Logic, and
LinkedIn's Sales Navigator, as shown in Figure 7 below.
Figure 7: Sales Intelligence solutions competitive landscape
Source: G2 Crowd Report for Sales Intelligence, Form CO, Annex 8.
(203) Each provider of sales intelligence solutions has a specific focus and a
differentiated offering: some focus on discrete industries and provide in-depth
information about selected or major industry participants; others provide general
market intelligence and basic contact information about potential points of
contact for sales generation. For example, Dun & Bradstreet offers data on more
than 85 million companies, 100 million people across 900 industries.
177
Zoominfo has data on almost 200 million business people across 100 countries
with 50 thousand contacts added daily.
178
InsideSales maintains a sales database
with over 100 billion interactions.
179
Avention provides data on over 53 million
companies, sourced from over 100 data partners, from searching over 1 billion
web pages and from 20,000 real-time news sources, including social media,
180
174
IBM's response to questionnaire to CRM Customers Q7 of 14 October 2016, question 18.
175
Mapei's response to questionnaire to CRM Customers Q7 of 14 October 2016, question 18.
176
Agreed non-confidential minutes of the conference call held on 28 October 2016.
177
http://www.dnb.co.in/IndiaSite/sales marketing solutions/sms-hoovers.aspx
178
http://www.zoominfo.com/business/why-zoominfo#coverage
179
https://uk.insidesales.com/products/platform/
180
https://www.avention.com/onesource-advantage
44
as well as data on over 40 million executives in the worlds.
181
InsideView
provides contacts on almost 5 million companies and 8 million contacts in
Europe and it has been built specifically for sales and marketing
professionals.
182
LinkedIn's Sales Navigator draws from LinkedIn’s database of
430 million profiles (of which 105 million are actively maintained).
(204) The market investigation showed that, in this highly fragmented and
differentiated market, customers tend to use different sales intelligence solutions
depending on the specific needs of the sales team (e.g. financial companies
sourcing sales intelligence solutions with access to financial datasets) and to
multisource.
183
(205) In the market investigation, all responding providers of sales intelligence
solutions have indicated that entry in the market is easy.
184
In this respect, Dun
& Bradstreet stated that "[t]he proliferation of entrants into the sales
intelligence solutions market demonstrates the ease with which a new player can
create and distribute a new and competitive solution. Due to the availability of
SaaS hosted solutions, as well as marketplaces such as Salesforce’s
appexchange or Microsoft’s AppSource, a new sales intelligence solutions
provider can launch with little to no up-front investment."
185
In the same vein,
Madison Logic explained that "[m]any data providers and platforms have
provided APIs to connect into their data set. This makes creating a sales
intelligence platform relatively straight forward to bring to market."
186
(206) Sales intelligence solutions can be used as a complement to CRM software
solutions, and in particular CRM Sales solutions, as they provide useful insights
which can increase the productivity and effectiveness of sale forces. In this
respect an industry report, based on the results of a customer survey, reported
that "[u]sers appreciated sales intelligence products that connected directly to
their CRM programs, allowing them to easily build lists and manage contacts.
Users noted that this added integration can offer increased automation and do
more to organize leads."
187
Such complementarity is also shown by the
existence of partnerships between sales intelligence solutions providers and
providers of CRM software solutions, such as those between LinkedIn, on the
one hand, and Microsoft and Salesforce, on the other hand. Notwithstanding
this, sales intelligence solutions have only played a limited role historically, in
particular in the EEA compared to North America.
188
181
https://www.quora.com/What-is-onesource-database-most-effectively-used-for
182
https://www.insideview.com/relevance/
183
Responses to questionnaire to CRM Customers Q7 of 14 October 2016, questions 12 and 25.
184
Responses to questionnaire to Sales Intelligence Solutions Competitors Q6 of 14 October 2016,
question 15.
185
Dun & Bradstreet's response to questionnaire to Sales Intelligence Solutions Competitors Q6 of 14
October 2016, question 15.
186
Madison Logic's response to questionnaire to Sales Intelligence Solutions Competitors Q6 of 14
October 2016, question 15.
187
Form CO, Annex 8,G2 Crowd Report for Sales Intelligence.
188
For example, see Salesforce's response to questionnaire to CRM Competitors Q5 of 14 October 2016,
question 30.
With specific regard with Sales Navigator, this is confirmed by the fact that today more than […]% of
Sales Navigator's sales are to customers in the United States, while customers in the EMEA regions,
including Europe, Middle-East and Africa, account for approximately […]% of the total. Based on
LinkedIn's estimate, these figures would not change in the next years with most of Sales Navigator's
46
Microsoft
[0-5]%
[0-5]%
[10-
20]%
[10-
20]%
[0-5]%
[0-5]%
[0-5]%
[0-5]%
[0-5]%
[0-5]%
Adobe
[0-5]%
[0-5]%
[0-5]%
[0-5]%
[10-
20]%
[10-
20]%
[0-5]%
[0-5]%
[0-5]%
[0-5]%
IBM
[0-5]%
[0-5]%
[0-5]%
[0-5]%
[10-
20]%
[10-
20]%
[0-5]%
[0-5]%
[0-5]%
[0-5]%
Others
[50-
60]%
50-
60]%
[20-
30]%
[20-
30]%
[40-
50]%
[40-
50]%
[60-
70]%
[60-
70]%
[70-
80]%
[60-
70]%
Source: Form, CO, Annex 9, documents 4, 5 and Annex 10; Salesforce's response to RFI n. 1.
(211) Microsoft is the fourth operator in the market, with worldwide and EEA market
shares by revenue both at [0-5]%. Microsoft has a larger segment share in the
CRM Sales segment where it is the second operator both worldwide at EEA
level after Salesforce, the leader.
(212) On the basis of the data presented in the above Table 2, the concentration level
in the market for CRM software solutions and in all segments thereof is below
2000, except in the CRM Sales segment worldwide where it is above 2000 due
to Salesforce's high market share.
(213) On the basis of the estimates provided by the Notifying Party, Microsoft's
position would not be different even if looking at market segmentations by type
of service, customer size, customer industry or between B2B and B2C
functionalities.
194
Indeed, the Notifying Party estimates that its segment share in
cloud-based and on premise CRM solutions in the EEA was respectively [5-
10]% and [5-10]% in 2015.
195
At worldwide level in 2015, in relation to cloud-
based CRM software solutions, Salesforce is the leader with a [30-40]%
segment share by revenue, followed at distance by Adobe, Oracle and then
Microsoft, which was the fourth operator with a segment share of approximately
[0-5]%.
(214) Moreover, the Notifying Party estimates that its market share does not exceed
[10-20]% in CRM software solutions purchased by small and medium business
nor in CRM software solutions purchased by large enterprises worldwide or in
the EEA.
196
(215) In addition, with respect to a potential segmentation by customer industry, the
Notifying Party submits that it would not be active in the distinct market
segment for CRM focusing on the healthcare sector,
197
which is the only
segment identified in the Commission's past decisions.
(216) Finally, on the basis of the estimates provided by the Notifying Party,
Microsoft's segment shares in the overall B2B and B2C CRM segments appear
to be in line with Microsoft's market shares in the market for CRM software
194
Concentrations levels could not be calculated in any of these possible segments as the Notifying Party
only provided data on Microsoft's share and the total market size.
195
Form CO, Section 6, paragraph 488.
196
Form CO, Section 6, paragraph 489 and Table 19.
197
Form CO, Section 6, paragraph 494. The Notifying Party was not aware and not able to estimate the
market share figures for other possible industry segments (such as healthcare or financial). The
Notifying Party submitted that, while there are certain customers which purchase Microsoft's CRM
software solutions as base platform to build their own industry specific solutions, such sales represent
[10-20]% or less of Microsoft's worldwide and EEA-wide CRM revenues.
47
solutions. Likewise, Microsoft's shares in the B2B and B2C sub-segments
within Sales, CSS, Marketing and DC appear to be in line with Microsoft's
segment shares in, respectively, the Sales, CSS, Marketing and DC segments.
(b) Sales intelligence solutions
(217) Based on the Notifying Party’s estimates, LinkedIn’s market share in the market
for sales intelligence solutions would be well below [20-30]% at worldwide
level and in the EEA, and most likely below [10-20]%,
198
with the rest of the
market being highly fragmented.
4.2.2.3 Conglomerate non-coordinated effects
(218) As mentioned in paragraph (206), sales intelligence solutions, such as Sales
Navigator offered by LinkedIn, can be used as a complement to CRM software
solutions, and in particular CRM Sales solutions, such as those offered by
Microsoft. Therefore, sales intelligence solutions and CRM software solutions
can be considered complementary or at least closely related products within the
meaning of paragraph 91 of the Non-Horizontal Guidelines.
(219) During the market investigation a concern has been raised that, post-
Transaction, Microsoft could bundle or tie Sales Navigator with its CRM
software solutions, or undertake other exclusionary practices, so that customers
of a competing CRM software solution would not have access to Sales
Navigator. As a result of this strategy, customers of competing CRM software
solutions would shift to Microsoft and other providers of CRM software
solutions would be foreclosed.
199
(220) Accordingly, the Commission analysed whether the Transaction could confer on
the merged entity the ability and the incentive to leverage its market position
from sales intelligence solutions to the CRM software solution market by means
of bundling, tying or other exclusionary practices, as well as whether such
potential conduct would have an effect on competition.
(a) Notifying Party's view
(221) The Notifying Party submits that Microsoft would not have the ability and the
incentive to foreclose access to Sales Navigator to its competitors in the CRM
software solutions market. First, neither Microsoft nor LinkedIn have significant
market power in respectively the CRM software solutions market and sales
intelligence solutions market. Second, LinkedIn's Sales Navigator is not an
important asset in the CRM software solutions market. Third, several alternative
providers would remain active in the sales intelligence solutions market.
Moreover, Microsoft would not have any incentive to foreclose since most of
Sales Navigator's revenues come from customers of competing CRM software
198
The Parties were not aware and not able to estimate the total figure for revenues from sales intelligence
solutions in the EEA and provided the above estimates using two methodologies: (i) as a proxy based
on LinkedIn' market position in North America, as they do not have any reasons to believe that it is
likely that LinkedIn has a higher share in the EEA than it does in North America (North America
market shares as based on the third party report "InsideSales Sales Acceleration Technology Market
Size Study", Form CO, Annex 8), and (ii) estimating a total market size by summing up the turnovers
of LinkedIn's competitors active in the market (Form CO, Annex 29).
Concentrations levels could not be calculated as the Notifying Party only provided data on Microsoft's
share and the total market size.
199
No concerns have been raised with respect to the reverse conduct, that is to say a possible foreclosure
of competing providers of sales intelligence solution as a result of bundling or tying of Microsoft's
CRM software solutions with Sales Navigator or other exclusionary practices.
48
solutions and any foreclosure strategy would negatively impact Microsoft's
overall revenue.
(b) Commission's assessment
(i) Ability to foreclose
(222) The Commission considers that post-Transaction the merged entity would not
have the ability to foreclose CRM software solutions competitors for the
following reasons.
(223) First, taking into consideration paragraph 25 of the Non-Horizontal Guidelines,
the Commission considers that LinkedIn does not have a strong market position
within the meaning of paragraph 93 of the Non-Horizontal Guidelines in the
sales intelligence solution market, which it could leverage to foreclose
competing providers of CRM software solutions. Indeed, as illustrated in
Section 4.2.2.2(b), on the basis of the Parties' best estimates, LinkedIn’s current
market share in the sales intelligence solutions market would be well below [20-
30]% at worldwide level and in the EEA, and most likely below [10-20]%.
(224) Moreover, as described in Section 4.2.2.1(b), the market for sales intelligence
solutions is a highly fragmented and differentiated market where LinkedIn’s
Sales Navigator appears to be only one of the many alternative solutions
available.
200
In this respect the Commission notes that none of the CRM
customers responding to the market investigation has indicated that it currently
purchases Sales Navigator, and they have instead stated to be using alternative
products such as Avention, Data.com, Dun & Bradstreet, InsideView, as well as
self-developed solutions.
(225) Likewise, while respondents have acknowledged the specific qualities of
LinkedIn’s database (in particular its accuracy and updated nature), when asked
about alternatives to Sales Navigator in a post-Transaction scenario, the majority
of the respondents to the market investigation noted that there will remain many
alternative solutions in the market including Avention, Data.com, Dun &
Bradstreet, InsideView, Twitter and others.
201
As illustrated in paragraph (203),
those alternative providers of sales intelligence solutions have also access to
significant databases, which are comparable to LinkedIn.
(226) Second, even if, notwithstanding the above, Sales Navigator were to be
considered a particularly important product for CRM customers, in particular for
customers of CRM Sales software solutions, the Commission considers that the
pool of customers of Sales Navigator that purchase also CRM software solutions
would not be large enough for the analysed risk of foreclosure to arise.
(227) Indeed, as stated in paragraph (206), generally sales intelligence solutions have
only played a limited role historically, in particular in the EEA. With specific
regard to Sales Navigator, at the end of 2015 on a worldwide basis, only
200
These alternative solutions are viewed as substitutes even if they are differentiated as they can cater for
the different CRM-needs of individual companies. As a result, the market for sales intelligence
solutions is a competitive one with differentiated products where the supplier of each individual
product does not enjoy "local" market power as customers view the alternative products as substitutes.
201
Responses to questionnaire to CRM Customers Q7 of 14 October 2016, question 25, questionnaire to
CRM Competitors Q5 of 14 October 2016, question 36, questionnaire to additional CRM Customers
Q18 of 28 October 2016, question 9, and questionnaire to Sales Intelligence Solutions Competitors Q6
of 14 October 2016, question 21.
49
[…]%
202
and […]%
203
of respectively Microsoft's and Salesforce’s CRM
customers (accounting together for around [50-60]% of the CRM Sales segment
both worldwide and in the EEA) were also LinkedIn's Sales Navigator
customers.
(228) While the importance of sales intelligence solutions for CRM customers is
expected to grow over the next two to three years,
204
the evidence in the
Commission’s file suggests that such increase would not significantly affect the
number of CRM software solutions customers which are also using Sales
Navigator. Notably, while Microsoft’s internal documents show an increase by a
factor of […] in the penetration of Sales Navigator on its CRM online customer
base, thus excluding Microsoft's on-premises customers, in 7 years post-
Transaction, such increase is from […]% in 2015 to only […]% in 2021 for the
CRM software solutions which should benefit from the analysed foreclosure
strategy, as shown in Figure 8 below.
Figure 8: […]
[…]
Source: Form CO, Annex 5, Document 8.
(229) Third, the Commission considers that competing CRM software solution
providers may have the ability to undertake effective and timely counter
strategies which could defeat the merged entity’s foreclosure effort. Indeed, the
Commission notes that, in a scenario where the importance of sales intelligence
solutions increases, given the differentiated nature of these products and the
tendency of customers to multisource,
205
also other solutions than the one
offered by LinkedIn may become more important than they are today. In this
context, providers of CRM software solutions could start partnering with sales
intelligence solutions providers other than LinkedIn. In this respect the
Commission notes that already today Salesforce has partnerships not only with
LinkedIn but also with Dun & Bradstreet.
206
(ii) Incentive to foreclose
(230) As regards Microsoft's incentive to foreclose competing providers of CRM
software solutions through bundling or tying the Parties' offerings, or through
other exclusionary practices, post-Transaction, the Commission notes the
following.
(231) The results of the market investigation were not conclusive regarding the
merged entity's incentive to foreclose. While half of the responding CRM
competitors expressed the view that Microsoft would have such incentive, the
202
Form CO, Annex 5, Document 46, spreadsheet "Synergies", Sales Navigator % penetration in
Dynamics Online CRM accounts.
203
The number of Sales Navigator's seats with Salesforce's customers is based on Figure 9 and data
contained in Form CO, Annex 5, Document 46, spreadsheet "Synergies". The total number of
Salesforce's CRM seats has been calculated dividing Salesforce's revenue ([…]) by the value of
Salesforce's most expensive CRM product ([…]).
204
See paragraph (207) above.
205
See paragraph (204) above.
206
Dun & Bradstreet provides information on several million companies to be included in Data.com,
Salesforce's sales intelligence solution. See
http://investor.dnb.com/releasedetail.cfm?releaseid=602322
50
other half did not express a view in this regard.
207
Likewise, the overwhelming
majority of customers were not able to provide a view as to whether Microsoft
will continue offering Sales Navigator on a standalone basis post-Transaction.
208
Competing providers of sales intelligence solutions, on the contrary,
unanimously considered that Microsoft would not have any reasons to stop
offering Sales Navigator to customers of competing CRM software solutions.
209
In this respect, MadisonLogic stated that "Microsoft CRM solution today is not
the market leader. Salesforce.com dwarfes Microsoft's market share [sic]. If
Microsoft were to cut off access to Sales Navigator from any other CRM
solution they would alienate themselves from where their customers are and
need that data".
210
Similarly, Dun & Bradstreet noted that "limiting access to
Sales Navigator would reduce [Microsoft’s] ability to drive revenues from
LinkedIn’s properties since Dynamics CRM has a relatively limited reach
according to most analyst reports."
211
(232) The Commission also notes that it is not clear to what extent such foreclosing
strategy would be profitable for Microsoft.
(233) Indeed, on the basis of the estimates carried out in a pre-Transaction standalone
scenario, Sales Navigator's revenues, accounting for almost […]% of LinkedIn's
total revenue, are expected to significantly increase to approximately […] by
2019.
212
These revenues would be generated mostly (over […]) from customers
of other providers of CRM software solutions, as it is the case today where only
[…] of Sales Navigator's customer base uses Microsoft’s CRM software (Figure
9).
Figure 9: […]
[…]
Source: […], Annex to RFI n. 11.
(234) The fact that access to the customer base of competing CRM software solutions
providers is an important distribution channel for Sales Navigator is confirmed
by the partnership agreements into which LinkedIn has entered. Indeed, pre-
Transaction LinkedIn had entered into partnerships with both Microsoft and
Salesforce and the partnership with latter has been renewed […] in July 2016
213
after the announcement of the Transaction. […].
214
Moreover, in September
2016, LinkedIn announced new integration partnerships also with other
providers of CRM software solutions including SAP, Oracle, NetSuite, Hubspot,
207
Responses to questionnaire to CRM Competitors Q5 of 14 October 2016, questions 32, 33 and 34.
208
Responses to questionnaire to CRM Customers Q7 of 14 October 2016, question 23, and questionnaire
to additional CRM Customers Q18 of 28 October 2016, question 7.
209
Responses to questionnaire to Sales Intelligence Solutions Competitors Q6 of 14 October 2016,
questions 16.
210
Madison Logic's response to questionnaire to Sales Intelligence Solutions Competitors Q6 of 14
October 2016, question 16.
211
Dun & Bradstreet's response to questionnaire to Sales Intelligence Solutions Competitors Q6 of 14
October 2016, question 16.
212
Form CO, Annex 5, Document 46.
213
On 19 July 2016, LinkedIn and Salesforce entered a new partnership agreement, for […].
214
See Salesforce and LinkedIn partnership agreement "Copy of ISVForce Reseller Agreement dated 15
January 2013; Amendment to ISVForce Reseller Agreement dated 28 January 2016", Form CO,
Annex 16. Response to RFI 15, question 2.
51
and SugarCRM.
215
Finally, an internal document of LinkedIn shows that
integrating Sales Navigator with CRM software solutions delivers significant
value by improving product engagement, reducing the number of customers
leaving the service and preventing customers from working with other sales
intelligence solutions that scrape LinkedIn's data from the web.
216
(235) Moreover, the Commission notes that LinkedIn's Sales Navigator has higher
yearly average revenue per user (“ARPU”) and a higher profit margin compared
to Microsoft's CRM software solutions.
217
Further, Microsoft's current revenues
from the licensing of cloud-based CRM software solutions are expected to grow
up to USD […] by 2019. By that time, Sales Navigator's standalone revenue will
grow to approximately […]. Therefore, revenues generated from Sales
Navigator are estimated to account for […]% of Microsoft's cloud-based CRM
revenues by 2019.
218
(236) In this context, considering the tendency of CRM customers not to change
providers of CRM software solutions
219
and the limited forecast in the expansion
of the number of customers CRM software solutions that also purchase Sales
Navigator, any strategy reducing or denying sales of Sales Navigator to
customers of competing providers of CRM software solutions risks translating
in significant losses which may not be compensated by the gains from
expanding market shares in the CRM software solutions market.
(237) In line with this, the Commission notes that internal documents of Microsoft
evidence its incentive to continue LinkedIn's collaboration with other providers
of CRM software solutions. For example, the minutes of an executive meeting
of June 2016, after the Transaction was agreed, report a recommendation to
maintain "partnerships with key CRM providers".
220
Likewise, in an internal
email of May 2016 it is highlighted that "Salesforce sees value in LinkedIn and
we [Microsoft] want that relationship to continue".
221
In the same vein, in its
synergy documents, Microsoft estimates that, in the next years […].
222
Therefore, Microsoft, similar to LinkedIn absent the Transaction, is likely to
have the incentive to further improve Sales Navigator and grow its customer
base, irrespective of a customer's provider of CRM software solutions, in order
to meet its revenue targets.
(238) In light of the above, the Commission considers that, post-Transaction,
Microsoft is unlikely to have the incentive to foreclose competing providers of
CRM software solutions by engaging in bundling, tying or other exclusionary
practices in relation to Sales Navigator and its CRM software solution.
(iii) Overall likely impact on effective competition
(239) As regards the impact of a foreclosure strategy by Microsoft on the market for
CRM software solutions, in the market investigation the overwhelming majority
215
See https://business.linkedin.com/sales-solutions/blog/linkedin-sales-navigator/linkedin-launches-crm-
partner-program.
216
LSS CRM Partner Program Deck Final, Annex to RFI 11.
217
[…]. See Form CO, Annex 5, Document 46, spreadsheet "Synergies".
218
Form CO, Annex 5, Document 46, spreadsheet "Synergies".
219
See paragraph (201) above.
220
Form CO, Annex 5 – Document 38.
221
Form CO, Annex 17.
222
Form CO, Annex 5, Document 46, spreadsheet "Synergies".
52
of CRM customers and all providers of sales intelligence solutions considered
that the Transaction would not have any impact on their company or on the
market for CRM software solutions.
223
Likewise, CRM competitors, i.e. those
companies which would be allegedly foreclosed, did not raise concerns as
regards possible conglomerate foreclosure effects stemming from the
Transaction, and initial concerns raised by a CRM competitor were eventually
superseded by later submissions, where that competitor minimized its prognosis
of the anticompetitive effects of any bundling, tying and other exclusionary
practices undertaken by Microsoft post-Transaction.
224
(240) The Commission considers that the Transaction is unlikely to have a negative
impact on effective competition, in particular with regard to prices and choice in
the market for the provision of CRM software solutions, as the effect of
Microsoft's potential exclusionary practices is unlikely to be enough to reduce
the ability and incentives to compete of other providers of CRM software
solutions.
(241) In this respect the Commission notes, first, that is unlikely that if Sales
Navigator were to be available only with Microsoft's CRM software solution,
this would prompt CRM customers to switch to Microsoft. No CRM customer
responding to the market investigation stated that it would stop considering
acquiring CRM software solutions from other providers in case Microsoft were
to start offering Sales Navigator only with its CRM software solution.
225
In this
regard, a customer explained that it would not switch because "the ecosystem for
other providers is more robust and provides much better opportunity to leverage
best in class partner solutions (e.g., Sugar, [Salesforce], Oracle Sales CRM). In
addition the flexibility to customize with other competitors is greater
([Salesforce]and Sugar)".
226
227
Likewise IBM stated that it would not switch as
LinkedIn is not the only provider of business data and other suppliers are likely
to offer competing solutions.
228
(242) Such statements made by customers in the market investigation are in line with
the Commission's findings as regards the unwillingness of CRM customers to
switch providers of CRM software solutions and the existence of several
alternatives to Sales Navigator.
(243) Second, the Commission notes that, even if all customers of Sales Navigator,
current and estimated in the near future, were to switch to Microsoft's CRM
software solution, quod non, the increase in Microsoft’s market share in the
CRM software solutions market and Sales segment thereof would be limited.
223
Responses to questionnaire to CRM Customers Q7 of 14 October 2016, questions 30 and
31,questionnaire to additional CRM Customers Q18 of 28 October 2016, questions 14 and
15,questionnaire to Sales intelligence solutions Competitors Q6 of 14 October 2016, question 24, 25
and 26.
224
See responses to questionnaire to CRM Competitors Q5 of 14 October 2016, question 47.2.
225
Responses to questionnaire to CRM Customers Q7 of 14 October 2016, question 24, and responses to
questionnaire to additional CRM Customers Q18 of 28 October 2016, question 8.
226
Automatic Data Processing's response to questionnaire to additional CRM Customers Q18 of 28
October 2016, question 8.
227
More generally, diversion from Salesforce to Microsoft may stay low as CRM customers buy a bundle
of CRM software solutions (see responses to questionnaire to CRM Customers Q7 of 14 October
2016, question 4) and are unlikely to switch to Microsoft because of just one particular CRM software
solutions, especially if switching is costly as described in Section 4.2.2.1 (a).
228
Responses to questionnaire to CRM Customers Q7 of 14 October 2016, question 24.
53
Indeed, today the total number of Sales Navigator's seats is approximately[…],
while Microsoft has approximately […] with its CRM software solution:
therefore, even assuming that no customer of Sales Navigator was a customer of
Microsoft's CRM software solution, a switch of all Sales Navigator's customers
to Microsoft's CRM software solution would increase Microsoft's customer base
only by […] and have a negligible impact in Microsoft's market shares in the
CRM software solutions market and Sales segment. On the basis of the Parties'
estimates, by 2021 the total number of Sales Navigator's seats would amount to
[…], while Microsoft would have approximately […] seats with its CRM
software solution: therefore, even assuming that all customers of Sales
Navigator would constitute additional customers for Microsoft CRM's software
solution, a switch of all Sales Navigator's customers to Microsoft's CRM
software solution would increase Microsoft's customer base by only […] and
have a negligible impact in Microsoft's market shares in the CRM software
solutions market and Sales segment.
(244) Finally, as described in Section 4.2.2.1(a), the Commission notes that the CRM
software solutions market is competitive and is characterised by the presence of
strong competitors, such as Salesforce, Oracle and SAP, with market shares
much larger than Microsoft.
(iv) Conclusion
(245) In light of the above, the Commission considers that the Transaction does not
give rise to serious doubts with regard to its compatibility with the internal
market as a result of conglomerate foreclosure effects to the detriment of
competing providers of CRM software solutions.
229
4.2.2.4 Vertical non-coordinated effects (input foreclosure)
(246) During the market investigation, competing CRM software solutions providers
claimed that LinkedIn full data, including but not limited to those displayed via
Sales Navigator, would constitute in the near future an important input within
the meaning of paragraph 31 and 34 of the Non-Horizontal Guidelines for the
provision of advanced functionalities in CRM software solutions through so
called ML.
230
On this basis a concern has been raised mainly by one third party
that post-Transaction Microsoft could restrict access to LinkedIn full data for the
purposes of ML in competing CRM software solutions, thereby making it harder
for other providers of CRM software solutions to compete and to bring
innovation in the market.
(247) As a preliminary remark, the Commission notes that, currently, LinkedIn does
not make available LinkedIn full data, or a subset thereof, to third parties for
ML. Sales Navigator is the only offering through which LinkedIn currently
229
For the sake of clarity, antitrust rules, in particular article 102 TFEU will continue to apply to the
merged entity after the closing of the Transaction, regardless of the outcome of the present assessment
under the Merger Regulation.
230
ML is based on algorithms that can learn from, process and rank data to make useful predictions to its
users. Applied to CRM software solutions, ML can sort, integrate and understand data of CRM
customers' organizations and other data sets available in the market to then suggest the best next action
for the CRM product user. The market investigation has confirmed the general market expectation that
ML will increase its importance for CRM software solutions in the next two to three years; see
Responses to questionnaire to CRM Customers Q7 of 14 October 2016, question 28, questionnaire to
additional CRM Customers Q18 of 28 October 2016, question 12, and questionnaire to CRM
Competitors Q5 of 14 October 2016, question 39.
54
makes available a subset of LinkedIn full data to third parties for CRM
purposes. In this context, the above-mentioned concern expressed by one third
party is predicated on the assumption that, absent the Transaction, LinkedIn
would have started monetising LinkedIn full data. This monetization strategy
would have concerned in particular the CRM industry, with, on the one side, an
inclusion of more LinkedIn's data in Sales Navigator, to make it the primary
sales intelligence solution and, on the other, an increase in the integration with
CRM software solutions.
231
(248) In this respect, the Commission notes that, first, it is unclear at this stage
whether LinkedIn would indeed have started licensing LinkedIn full data to third
parties absent the Transaction. Indeed as mentioned in paragraph (254),
LinkedIn’s pre-Transaction internal documents contain no reference to the
possible licensing of LinkedIn full data, or a subset thereof, to any third party,
including CRM software solution providers. This is in particular because,
LinkedIn's business model, and the value of the data it collects, are based on its
users' willingness to update their profiles and engage with LinkedIn services,
which may be undermined should users become aware that LinkedIn grants third
parties access to their data, albeit in a way that is consistent with applicable data
protection laws.
(249) Second, if LinkedIn did not have the incentive to start monetising LinkedIn full
data on a stand-alone basis, and if Microsoft post-Transaction had access to
these data and started using them to improve its own CRM software solutions,
the Transaction may even have pro-competitive effects, as it would allow for the
possibility of new products, or improvements to existing products in the market,
to the benefit of consumers, based on a dataset to which otherwise no one would
have had access.
(250) Therefore, the Commission considers that there is uncertainty as to whether in
the near future LinkedIn full data would effectively become an important input
within the meaning of paragraph 31 and 34 of the Non-Horizontal Guidelines.
(251) Nonetheless, the Commission has analysed whether the Transaction could
confer on the merged entity the ability and the incentive to reduce competition
by restricting access to an important input to its downstream competitors, in the
event that, absent the Transaction, LinkedIn would have started monetising its
data by licensing the LinkedIn full data to CRM customers and competing
providers of CRM software solutions
232
and/or, post-Transaction, Microsoft
would start using these data for its own CRM software solution.
(a) Notifying Party's view
(252) The Notifying Party submits that Microsoft would not have the ability and
incentive to foreclose its competitors from the provision of ML in the CRM
software solutions market. First, LinkedIn full data, or a subset thereof, are not
an important input to develop ML for CRM software solutions. Second, there
are alternative data available in the market from other vendors. Third,
231
Salesforce's response to questionnaire to CRM competitors Q5 of 14 October 2016, question 42.
232
The Commission understands that for LinkedIn full data to be used for ML on a given CRM software
solution a licence may be needed not only for the vendor of such CRM software solution but also for
the customer using that CRM software solution.
55
Microsoft's CRM competitors already provide ML for their CRM software
solutions.
233
(b) Commission's assessment
(i) Ability to foreclose
(253) The Commission considers that post-Transaction the merged entity would not
have the ability to foreclose competing providers of CRM software solutions as,
in any event, by reducing access to LinkedIn full data, it is unlikely to
negatively affect the overall availability of data for ML in CRM software
solutions.
(254) First, taking into consideration paragraph 25 of the Non-Horizontal Guidelines,
the Commission considers that LinkedIn does not appear to have a significant
degree of market power within the meaning of paragraph 35 of the Non-
Horizontal Guidelines in any potential relevant upstream market, which in this
case would be an hypothetical market or segment for the provision of data for
the purposes of ML in CRM software solutions.
234
Indeed, as stated in paragraph
(203), LinkedIn does not currently license any data to any third party and the
only data valuable for CRM purposes that it makes available to third parties are
those displayed to users of Sales Navigator, irrespective of whether or not Sales
Navigator is integrated with CRM software solutions (none of which are,
however, currently used for ML). Moreover, LinkedIn’s internal documents
show that, absent the Transaction, while LinkedIn was planning to enhance the
sales of Sales Navigator's subscriptions, […],
235
no reference was made to the
possible licensing of LinkedIn full data, or a subset thereof, to any third party,
including for ML purposes.
236
(255) In this context, the Commission notes that, as mentioned in paragraph (177),
Microsoft is subject to European data protection laws which limit its ability to
undertake any treatment of LinkedIn full data. While, today's LinkedIn's privacy
policy allows to share the personal data it collects, processes, stores and uses
with its controlling companies, this is only for the purposes described in the
privacy policy itself.
237
Moreover, the Commission notes that the newly adopted
General Data Protection Regulation, which will apply from 25 May 2018, may
further limit Microsoft's ability to undertake any treatment of LinkedIn full data
by strengthening the existing rights and empowering individuals with more
control over their personal data (i.e. easier access to personal data; right to data
portability; etc.).
(256) Second, the Commission considers that LinkedIn full data, or a subset thereof,
cannot be qualified as, and is not likely to become in the next two to three years,
233
Notifying Party's response to RFI n.21 and Annex Q.1 of 16 November 2016.
234
If LinkedIn was active in the provision of data to third parties for ML, the hypothetical market or
segment for the provision of data for the purposes of ML in CRM software solutions would be the
narrowest segment where LinkedIn would be active and which would be relevant for the assessment in
this Section.
235
See paragraph (234) above.
236
See LinkedIn's response to RFI n.11 and Annexes.
237
According to its current privacy policy LinkedIn will provide personal data to third parties (i) with the
user's consent, (ii) where it is necessary to carry out the user's instructions, (iii) in order to provide
LinkedIn's current features and functionality to the user, (iv) when LinkedIn believes it is required by
law, or (v) as necessary to enforce LinkedIn's user agreement or protect the rights, property, or safety
of LinkedIn, its members and visitors and the public.
56
an important input within the meaning of paragraph 34 of the Non-Horizontal
Guidelines with respect to the provision of ML in CRM software solutions.
(257) Indeed, the Commission notes that, although in the market investigation all
CRM competitors and half of the customers considered that LinkedIn full data
may be, or will be, important for ML in CRM software solutions,
238
all major
CRM vendors have already started offering advanced functionalities to their
CRM customers based on ML, or plan to do so in the next two to three years,
and none of these offerings has been developed or requires for its use access to
LinkedIn full data.
239
(258) Furthermore, Microsoft’s internal documents concerning the Transaction,
including the synergies plans, do not mention the use of LinkedIn full data, or a
subset thereof, for ML (be it for CRM software solutions or other purposes).
240
Likewise, [CONFIDENTIAL].
241
(259) In addition, the Commission understands that, even if LinkedIn full data, or a
subset thereof, were to be used in the near future for ML in CRM software
solutions, it would constitute only one of the many types of data which are
needed for this purpose. Indeed, the data that are needed for ML in CRM
software solutions come from essentially two data sources: in-house customer
data uploaded in the CRM software and complementary third party data.
242
(260) In-house customer data uploaded in the CRM software relates to accounts,
service tickets, interactions, leads, etc. These data are by definition available to
each relevant provider of CRM software solutions and availability of such data
will not be affected by the Transaction.
(261) Third party data relevant for ML can be different depending to the use case and
the relevant industry. The data collected by LinkedIn are one source of the third
party data which could be used for ML and may be relevant for certain use cases
in certain industry sectors,
243
but not for others. Considering the quality of the
data collected by LinkedIn
244
and the submissions of the Notifying Parties
245
and
third parties in the context of the market investigation,
246
the Commission
understands that LinkedIn full data may be relevant for the CRM B2B Sales and
B2B Marketing sub-segments, but not for others. In this regard, SAP stated that
"LinkedIn is only one data source. Depending on the use case, other types of
238
Responses to questionnaire to CRM Competitors Q5 of 14 October 2016, question 40, questionnaire to
CRM Customers Q7 of 14 October 2016, question 29, and questionnaire to additional CRM Customers
Q18 of 28 October 2016, question 13.
239
See Section 4.2.2.1(a) above, as well as, Agreed non-confidential minutes of the conference call with
SAP held on 3 November 2016 and responses to questionnaire to CRM Competitors Q5 of 14 October
2016, question 37.
240
Form CO, Annex 5.
241
Response to RFI n.10 and Annexes.
242
In this regards see Agreed non-confidential minutes of the conference call with SAP held on 3
November 2011, point 4.
243
For example, a customer active in the financial sectors, Provident Financial Management Services
Ltd., explained that "We currently do not use social media as a way to communicate with our
customers and do not recruit new customers by using Linkedin," response to questionnaire to Sales
Intelligence Solutions Competitors Q6 of 14 October 2016, question 22.
244
See footnote 54.
245
Notifying Party's response to RFI n. 15.
246
For example, Response to RFI n. 16; Agreed non-confidential minutes of the conference call with SAP
held on 3 November 2016.
57
data might be more relevant than LinkedIn. It is difficult to predict how this will
evolve in the future."
247
In the same vein Oracle explained that "there is not one
dataset with the highest value [as input for ML], but that it is about having
numerous types of data. Therefore, not only the quality, but also the quantity
and the variety are important."
248
(262) Finally, the Commission notes that there are many other possible sources of data
which are already available for ML. In this respect all providers of sales
intelligence solutions replying to the market investigation stated that the data
included in their sales intelligence solutions could be used by providers of CRM
software solutions to provide ML.
249
Dun & Bradstreet, whose data are currently
available for ML, explained that ML "allows the onboarding [sic] and analysis
of data from any source. To that end, data acquired from any sales intelligence
provider could be used in that fashion if licensed for such a purpose either
through a partnership or by the end customer. Avention Data.com Dun &
Bradstreet InsideView Twitter And others." Likewise, Madison Logic stated that
"[i]nsights from sales intelligence would be very valuable for usage inside of
CRM with a machine learning approach."
(263) Moreover, CRM customers, even those considering that LinkedIn full data may
be, or will be, important for ML in CRM software solutions, explained that there
are alternatives.
250
For example, IBM explains that "LinkedIn data is very useful
but is not the only source of data. There are many sources of unstructured
information about commercial markets and cognitive solutions can interrogate
and make sense of those." Automatic Data Processing, Inc. stated that "[t]he
data needed to leverage predictive analytics to target companies using
qualitative needs-based solutions is already readily available. This data
includes company firmographics [sic], news, regulatory change, information
about companies similar to the prospect, competitor announcements, internal
intelligence on win loss reasons, publicly available leadership change
information, pending legal issues and financial earnings announcements."
(264) Among the CRM competitors, while there is a general acknowledgment of the
specific qualities of LinkedIn full data (in particular its accuracy and updated
nature), when asked about the existence of alternatives the views were mixed.
SAP, Oracle and two other providers of CRM software solutions stated that
there are alternatives,
251
while Salesforce, Zoho and E-Deal expressed the
opposite view.
252
In this respect, the Commission notes that Salesforce, Zoho
and E-Deal already offer, or plan to do so in the next two to three years, ML
without access to LinkedIn's full data.
253
247
Agreed non-confidential minutes of the conference call with SAP held on 3 November 2011, point 5.
248
Agreed non-confidential minutes of the conference call with Oracle held on 7 November 2011, point 8.
249
Responses to questionnaire to Sales Intelligence Solutions Competitors Q6 of 14 October 2016,
question 22.
250
Responses to questionnaire to CRM Customers Q7 of 14 October 2016, question 29, and questionnaire
to additional CRM Customers Q18 of 28 October 2016, question 13.
251
Oracle's response to questionnaire to CRM Competitors Q5 of 14 October 2016, question 46; Agreed
non-confidential minutes of the conference call with a CRM provider held on 28 October 2016;
Agreed non-confidential minutes of the conference call with SAP held on 3 November 2011; Agreed
non-confidential minutes of the conference call with another CRM provider held on 3 November 2011,
point 6.
252
Response to questionnaire to CRM Competitors Q5 of 14 October 2016, question 46.
253
See paragraph (257) above.
58
(ii) Incentive to foreclose
(265) As regards Microsoft's incentive to foreclose CRM competitors by restricting
access to LinkedIn full data, the Commission notes the following.
(266) The results of the market investigation were not conclusive, with half of the
responding CRM competitors claiming that Microsoft would have such
incentive and another half that could not express a view.
254
(267) The Commission also notes that it is not clear to what extent any such
foreclosing strategy would be profitable for Microsoft. Indeed, as LinkedIn full
data are not yet made available in the market, it is not possible to estimate what
could be the actual profits that LinkedIn, and post-Transaction Microsoft, would
derive from the licensing of that dataset. However, Sales Navigator is a tool
which allows for the display of a subset of LinkedIn full data and could
therefore be considered a useful proxy to estimate the losses that the merged
entity would incur if it were to engage in a foreclosure strategy. In this regard,
the Commission notes that […].
255
If, as in the complainants' view, LinkedIn full
data are an unique input for ML in CRM software solutions, the profits that
LinkedIn and the merged entity could derive from its licensing should be at least
on par, or more likely greater, than those generated today from Sales Navigator.
(268) Furthermore, as noted in paragraph (261), LinkedIn full data are likely to be
relevant, if at all, for ML in the CRM B2B Marketing and B2B Sales sub-
segments. As shown in Table 2, Microsoft has a marginal presence with a
market share at most of [0-5]% in the EEA CRM Marketing segment overall.
While Microsoft has a larger presence in the CRM Sales segment, it is second to
Salesforce whose market share is more than twice as high in the EEA and more
than three times higher worldwide, and it is closely followed by SAP and
Oracle, having similar segment shares to Microsoft in the EEA. As noted in
paragraph (216), Microsoft's position would not be different if considering the
B2B segment and sub-segments. Should LinkedIn full data be relevant for ML
in the overall CRM software solutions market, quod non, Microsoft's market
share would again be quite limited if compared to the other players active in that
market.
(269) In this context, considering the unwillingness of CRM customers to switch
providers of CRM software solutions,
256
as well as the fact that today all of
Microsoft's larger competitors offer, or are going to offer in the near future, ML,
it appears that any strategy restricting access to LinkedIn full data to competing
CRM software solutions risks translating into significant losses which may not
be compensated by the gains from expanding market shares in the CRM
software solutions market.
(270) In line with this, the Commission notes that internal documents of Microsoft
demonstrate its incentive to continue its collaboration with other providers of
CRM software solutions post-Transaction.
257
(271) Indirect evidence of the likely incentives of Microsoft is also provided by the
strategic behaviour adopted by Microsoft with respect to other products in its
254
See responses to questionnaire to CRM competitors Q5 of 14 October 2016, question 44 and 45.
255
See paragraphs (233) and (235) above. This is the case, not only looking at CRM Sales and B2B Sales,
but also with respect to CRM Marketing and B2B Marketing.
256
See paragraph (201) above.
257
See paragraph (237) above.
59
portfolio and their integration with CRM software solutions of competitors.
Indeed, first, the Commission notes that certain of the data input currently used
or envisaged to be used by Microsoft's CRM competitors are emails and
calendar activities: yet Microsoft makes available to third parties, including
CRM vendors, Outlook APIs
258
and more broadly Office. Furthermore, the
Commission notes that, contrary to Salesforce (whose ML is included in its
CRM software), Microsoft had plans, absent the Transaction, to offer […].
259
(272) In light of the above, the Commission considers that, post-Transaction, it is at
least unclear whether Microsoft may have the incentive to foreclose competing
providers of CRM software solutions by restricting access to LinkedIn full data.
(iii) Overall likely impact on effective competition
(273) As regards the impact of a foreclosure strategy on the CRM software solutions
market, in the market investigation the overwhelming majority of CRM
customers and all providers of sales intelligence solutions considered that the
Transaction would not have any impact on their company or on the CRM
software solutions market.
260
The views of CRM competitors were mixed, with
half of them considering the impact of the Transaction as negative, and the other
half not raising any concern.
261
(274) The Commission considers that the Transaction is unlikely to have an overall
negative impact on effective competition in the market for CRM software
solutions, and any potential restriction of access to LinkedIn full data, or subset
thereof, is unlikely to lead to consumer harm.
(275) In this respect the Commission notes, first, that it is unlikely that if LinkedIn full
data, of a subset thereof, were to be used for ML only in Microsoft's CRM
software solution, this would affect a sufficiently important proportion of
Microsoft's competitors to result in a significant price increase or reduction of
market incentives to innovate. Indeed, as noted in paragraph (261), LinkedIn full
data are likely to be relevant, if at all, for ML in the CRM B2B Marketing and
B2B Sales sub-segments. As explained in paragraph (191), these sub-segments
together account for less than 30% of the entire CRM software solutions and are
expected to account for the same percentage of the CRM software solutions by
2020.
(276) Second, the Commission recalls that ML in CRM software solutions requires
access to multiple data sources to provide customers useful insights. LinkedIn is
only one of such data sources
262
which is unlikely to be essential. Therefore it is
unlikely that competing CRM software providers to whom access to LinkedIn
full data would be restricted would be hampered in their ability to compete and
innovate in the CRM software solutions market or that the mere likelihood that
the merged entity would carry out a foreclosure strategy would raise barriers to
entry to potential competitors.
258
On Outlook APIs, see paragraph (322) below.
259
See Microsoft's response to RFI n. 21, Annex Q.1.
260
Responses to questionnaire to CRM Customers Q7 of 14 October 2016, questions 30 and 31, responses
to questionnaire to additional CRM Customers Q18 of 28 October 2016, questions 14 and 15, and
responses to questionnaire to Sales intelligence solutions Competitors Q6 of 14 October 2016,
question 24, 25 and 26.
261
Responses to questionnaire to CRM Competitors Q5 of 14 October 2016, questions 47 and 48.
262
See paragraph (259) above.
60
(iv) Conclusion
(277) In light of the above, the Commission considers that the Transaction does not
raise serious doubts with regard to its compatibility with the internal market as a
result of input foreclosure effects to the detriment of providers of CRM software
solutions.
263
4.2.3 Foreclosure of competing PSN service providers
(278) Some of the products and services offered by Microsoft (such as OS,
productivity software and online communications services) are common IT
products or services that can potentially be used by any consumer or company
employee. Likewise, LinkedIn's PSN services are potentially available to any
consumer or company employee who has access to an Internet-enabled device.
As a result, some of Microsoft's products can be considered to be
complementary or at least closely related to LinkedIn's PSN services within the
meaning of paragraph 91 of the Non-horizontal Guidelines.
(279) Accordingly, the Commission analysed whether the Transaction could give rise
to conglomerate non-coordinated effects through (i) foreclosure of providers of
PSN services that compete with LinkedIn; or (ii) foreclosure of providers of
certain IT products that compete with Microsoft.
(280) In the present Section, the Commission analyses conglomerate non-coordinated
effects through the possible foreclosure of competing providers of PSN services.
The possible foreclosure of Microsoft's competitors in productivity software and
in online communications services is analysed in Sections 4.2.5 and 4.2.6 below.
(281) The Commission also assessed a possible conduct of the merged entity
consisting of bundling Microsoft products with LinkedIn's online recruitment
services. As this conduct could potentially foreclose not only PSNs but also
other providers of online recruitment services, it will be analysed separately in
Section 4.2.4 below.
4.2.3.1 Market shares
(282) This Section sets out available market share data with respect to (i) PSN
services; (ii) OSs; and (iii) productivity software.
(a) PSN services
(283) The Notifying Party has provided the Commission with market share data for
PSN services based on desktop traffic in July 2016.
264
According to the
Notifying Party's data, LinkedIn would hold a market share below [20-30]% in a
potential EEA-wide market for PSN services, being the second largest player
after Facebook (with a market share of over [50-60]%). At national level,
LinkedIn would account for a share of the PSN market between [5-10]% and
263
For the sake of clarity, antitrust rules, in particular article 102 TFEU will continue to apply to the
merged entity after the closing of the Transaction, regardless of the outcome of the present assessment
under the Merger Regulation.
264
The market shares are based on data from Similar Web. Each website's share is based on its total
traffic comprising all visits to the website in July 2016, with the exception of Facebook, which is
included with a share of 12.4% of its total traffic, since this is the share of Facebook's daily users that
match "work" characteristics in all industries (Form CO, Annex 9, Document 2).
64
Excel and PowerPoint) were taken into account separately. Furthermore, while
the market shares in Table 5 and 6 above do not cover email and calendaring
software (such as Outlook), other sets of data available suggest that Microsoft's
market share is likely to be significantly high in this segment too. In particular,
the data provided by the Notifying Party regarding email deployment indicate
that Outlook was deployed in [50-60]% of organisations in Western Europe in
December 2015.
277
In addition, according to estimates provided by the Notifying
Party, Office as a whole has been installed on [70-80]% of all PCs in Europe in
the course of 2016.
278
(294) As regards other productivity software products that are not included in the
definition of Office,
279
Microsoft's presence appears to be significantly smaller
than for Office products.
280
4.2.3.2 Conglomerate non-coordinated effects
(295) In light of Microsoft's very high market shares in the markets for (i) OSs for PCs
in the EEA, and (ii) productivity software for PCs in the EEA, as well as some
of its possible segments as set out in Section 4.2.3.1 above, the Commission
considers that Microsoft has at least a strong market position within the meaning
of paragraph 93 of the Non-Horizontal Guidelines in each of those markets.
Accordingly, the Commission investigated if, after the Transaction, the merged
entity could have the ability and the incentive to leverage its strong market
position from those markets to the market for PSN services by means of
exclusionary practices.
(296) By contrast, the Commission considers that Microsoft does not have a strong
market position within the meaning of paragraph 93 of the Non-Horizontal
Guidelines in other markets which could potentially be considered to be related
to the market for PSN services. This applies to the markets for OSs for tablets
and mobile phones (Windows and Windows Phone),
281
consumer
communications services (Skype Consumer),
282
online search services (Bing)
283
and online search advertising services (Bing) discussed in Sections 3.5 and
paragraph (150) and for which Microsoft could provide product bundles or ties
that include PSN services. In those markets, the Commission considers that
Microsoft's relatively limited market position would prevent it from foreclosing
competing PSN service providers by means of tying or bundling practices.
284
277
Annex 9 to Form CO, Document 6, p. 5. The data provided relate to Microsoft Exchange and
Microsoft Exchange Online, which can be used in this context as a proxy for Outlook.
278
Annex 9 to Form CO, Document 6, p. 9.
279
See footnote 276 above.
280
For example, with respect to Sharepoint, Microsoft accounts for [5-10]% of revenues in the segment of
content management software, [10-20]% in file synchronisation and sharing software, [40-50]% in
enterprise portals and [40-50]% in team collaborative applications. Moreover, Microsoft has a market
share of [10-20]% in publishing products (with Microsoft Publisher) and of [0-5]% in diagramming
products (with Microsoft Visio); Notifying Party's response to RFI n. 24, question 3.
281
See footnotes 270 and 272 above.
282
Skype Consumer's worldwide monthly active users in December 2015 were 272 million (lower than
WeChat's 650 million, Messenger's 800 million and WhatsApp's 990 million (see Form CO, Figure
77).
283
According to the information provided by the Notifying Party, Microsoft's position in online search is
limited, and in any event not above [10-20]% in Europe.
284
Microsoft's market shares in the online search advertising markets in Member States are between [0-
5]% and [5-10]% (Form CO, Annex 9, Document 8, "Online Search Advertising – Microsoft Share of
65
(a) Notifying Party's view
(297) According to the Notifying Party, the Transaction would not lead to foreclosure
of competing SN services, including those that focus on professionals.
(298) The Notifying Party argues, first, that, with or without the Transaction, LinkedIn
would have no ability to foreclose competitors due to the specific features of the
market where LinkedIn operates. In particular, according to the Notifying Party,
barriers to entry into SN services are low (as services can be hosted on cloud
platforms without the need to invest in expensive hardware) and successful
services can be developed by small teams of developers (as illustrated by the
entry of Instagram, Snapchat and beBee). Furthermore, the Notifying Party
claims that multi-homing among users of SN services is high, and provides data
aimed at supporting its claim. In the Notifying Party's view, users can and do
easily join other social networking services (as all that is required for this
purpose is to create a profile or to port it from LinkedIn) and they have the
incentive to do so to maximise the number of professional connections and the
visibility in the business community. The Notifying Party also provides data
aimed at showing that multi-homing users tend to be more engaged than single-
homing users and spend more time on each of the services that they use. In
addition, in the Notifying Part's view, as LinkedIn reaches only a limited share
of addressable demand in terms of active users and time spent, addressable
demand is large, leaving competitors ample scope for entry and expansion.
(299) Second, the Notifying Party argues that the integration of LinkedIn with
Windows and Office that Microsoft intends to carry out post-Transaction would
not foreclose competitors. In particular, the Notifying Party submits that
Microsoft does not control access points to social networking services as
Windows and Office are not "must-have" distribution platforms for social
networking services. More than 60% of LinkedIn's traffic is currently on mobile
devices (a share that is growing), where Window has no meaningful presence
(with a share below […]%). Even on PCs, the Notifying Party submits that most
([…]%) of the social networking traffic does not come though the operating
system or productivity software, but through search engines and web browsers,
where Microsoft's presence is limited. In addition, the Notifying Party submits
that pre-installation and integration do not drive substantial sign-ups and
engagement, as demonstrated by the fact that Skype's pre-installation on
Windows and integration with Office did not improve Skype's market share,
which remained flat (while new entrants such as WhatsApp grew rapidly). Also,
according to the Notifying Party, new sign-ups to LinkedIn are mainly driven by
online search, invitations from existing members and word-of-mouth, while only
[…]% of sign-ups originate from LinkedIn partnerships. In particular,
LinkedIn's partnerships with Samsung and Sony mobile devices, with Yahoo!
and Gmail concerning integration features did not significantly drive LinkedIn
uptake and engagement, as such features were activated by less than […]% of
Samsung/Sony smartphone holders and Yahoo!Mail or Gmail users.
(300) Third, the Notifying Party argues that the new services envisaged by Microsoft
arising from the integration of Microsoft products with LinkedIn would benefit
Supply"). Skype's worldwide monthly active users in December 2015 were 272 million (lower than
WeChat's 650 million, Messenger's 800 million and WhatsApp's 990 million (Form CO, Figure 77).
Skype for Business' market share was [10-20]% worldwide in 2015 in terms of revenues (Form CO,
Table 30).
66
consumers and cause no harm to competition. The Notifying Party argues that
such services would not involve forcing any products on customers, who would
be free to decide whether they want these enriched experiences. Moreover, it
argues that these services would not involve any degradation of competing
products. This is because Microsoft currently provides APIs that enable third
parties to develop software add-ins for Outlook and actively encourages third
parties to do so because add-ins improve the Outlook user experience and
increase the value of Microsoft products. The Notifying Party submits that, post-
Transaction, Microsoft intends to continue making these APIs available and to
further develop the add-in APIs to provide third parties with additional ways in
which to integrate with Office.
(b) Commission's assessment
(301) In this Section, the Commission assesses the concern that the combination of
LinkedIn's PSN services with Microsoft's PC-based OSs and productivity
software could lead the merged entity to leverage its strong market position
from the markets for PC-based OSs and for productivity software to the market
for PSN services, where LinkedIn's position is already strong, thereby
foreclosing competitors in that market and harming competition.
(302) On the basis of the submissions from respondents to the market investigation,
the Commission has identified two key sets of practices through which the
merged entity may be able to foreclose competing providers of PSN services by
leveraging its position in the markets for PC-based OSs and for productivity
software in the EEA. These practices consist of:
developing and pre-installing a LinkedIn branded application on PCs
running Windows OS ("pre-installation of a LinkedIn application on
Windows PCs"); and
integrating LinkedIn features within Microsoft's productivity software (in
particular, Outlook and other Office products) in various forms, while at
the same time denying similar levels of integration to competing providers
of PSN services ("integration of LinkedIn features into Office" and "denial
of access to Microsoft APIs").
(303) In addition, some respondents to the market investigation argued that the
merged entity could decide to offer a LinkedIn branded application as a tie or
bundle with Microsoft's productivity software (Office) and foreclose
competition for PSNs as a result of such conduct.
(304) The Commission considers that the concerns regarding possible tying or
bundling of Office with a LinkedIn application are unlikely to be well-founded.
First, unlike the two key sets of practices mentioned in paragraph (302) (pre-
installation of a LinkedIn application on Windows PCs and integration of
LinkedIn features into Office), Microsoft stated that it has no current plans to
include any LinkedIn branded standalone application with Office or Office
365.
285
Indeed, no such plan features in Microsoft's internal documents reviewed
by the Commission. Second, Microsoft typically makes available different
versions of Office, some of which include additional applications while others
do not, and it may therefore continue to offer post-Transaction versions of
Office or Office 365 with and without the LinkedIn application. Moreover, users
285
Notifying Party's response to RFI n. 21, question 12.
67
can typically pick and choose the apps to install when they acquire perpetual
licenses to Office.
286
All these elements, taken together, suggest that it is
unlikely that conglomerate effects would arise from a tying or bundling of
Microsoft's Office with a LinkedIn application. In any event, competition rules,
in particular Article 102 TFEU and Article 54 of the EEA Agreement will
continue to apply to the merged entity post-Transaction, regardless of the
outcome of the present assessment under the Merger Regulation. Accordingly,
potential tying or bundling of Office with a LinkedIn application will not be
assessed further in this Decision.
(305) In contrast, the pre-installation of a LinkedIn application on Windows PCs, the
integration of LinkedIn features into Office and their potential to foreclose
competing PSN service providers will be examined in detail below. For this
purpose, consistent with paragraph 94 of the Non-Horizontal Guidelines, the
Commission will examine: (i) whether the merged entity would have the ability
to foreclose competing PSN service providers; (ii) whether it would have the
economic incentive to do so; and (iii) what overall impact such a foreclosure
strategy would have on effective competition.
(i) Ability to foreclose
(306) Concerns have been raised that, post-Transaction, (i) the pre-installation of a
LinkedIn application on Windows PCs and (ii) the integration of LinkedIn
features into Office, in combination with the denial of access to Microsoft APIs,
would enable the merged entity to foreclose competing providers of PSN
services in the EEA.
(307) In the following paragraphs, the Commission will, for each of those two sets of
practices, analyse their technical feasibility, their likely impact on LinkedIn's
user base and activity, and ultimately on competing PSN providers.
Pre-installation of a LinkedIn application on Windows PCs
(308) As regards technical feasibility, Microsoft would first have to develop a
LinkedIn application for Windows and then pre-install it on Windows PCs on
the basis of its cooperation agreements with original equipment manufacturers
(“OEMs”).
287
Typically, OEMs are contractually required to install those
applications determined by Microsoft on the Windows PCs that they distribute.
Neither of these steps appears to raise any particular technical difficulty for
Microsoft. In fact, Microsoft has already identified a number of use case
scenarios that involve pre-installation of a LinkedIn application on Windows
PCs, specifically Windows 10 PCs.
288
In particular, Microsoft envisages
possibly developing a LinkedIn application and including it as a "tile" in the
Windows 10 start menu and on the desktop.
289
In terms of timing, according to
information submitted by the Notifying Party, the development of a LinkedIn
286
Notifying Party's response to RFI n. 21, question 12.
287
Microsoft traditionally sells the OSs it develops via three distribution channels: (i) to OEMs which
preinstall Windows on the devices they sell; (ii) upgrade versions in bulk to business customers; and
(iii) as individual copies through the retail channel (Form CO, paragraph 744). The information
provided by the Notifying Party indicates that sales through OEMs account for the large majority of
sales of the Windows 10 OS (Notifying Party's submission of 13 November 2016 at 15:07).
288
Form CO, Section 6, paragraph 270 and Table 14.
289
Form CO, Table 14.
68
application and its pre-installation on Windows 10 PCs could be implemented in
a relatively short timeframe, that is to say […].
290
(309) Regarding the impact of pre-installation of a LinkedIn application on Windows
PCs, the Commission notes that, in principle, software pre-installation can make
switching more difficult, in view of users' inertia which leads to the so-called
"status quo bias."
291
(310) In the present case, competing providers of PSN services who replied to the
market investigation confirmed that pre-installation of a LinkedIn application on
Windows PCs would be likely to significantly increase the number of LinkedIn's
members and their engagement and eventually foreclose competing PSN service
providers.
(311) According to XING, pre-installation of an application on a device can
significantly drive an application's distribution and, therefore, its usage. In
support of its claim, XING submits that its experience with the pre-installation
of the XING application on Windows mobile devices for business customers
was successful (also thanks to users' ability to access the XING application
using log-in credentials from the mobile OS), leading to an increase in user
activity.
292
In XING's view, given that Windows is omnipresent in a work
environment (with a 90% market share in Germany) and is the most important
touchpoint for PSNs, the pre-installation of a LinkedIn application on Windows
PCs would bring LinkedIn to a very large user base every day.
293
According to
GoldenLine, the impact of such pre-installation would be "enormous" since,
given that most PCs and laptops run the Windows OS, the LinkedIn application
would become available to all business PC users.
294
(312) While the Notifying Party argues that the pre-installation of a LinkedIn
application on Windows PCs would not drive significant growth in members
and usage, it has been unable to provide any precise estimates in that respect. In
addition, given that LinkedIn has not previously been installed on Windows PCs
or other devices, there are no data available to determine with a sufficient degree
of certainty which impact such a pre-installation would have.
(313) The Notifying Party's argument that only […]% of sign-ups to LinkedIn are
driven by LinkedIn partnerships cannot provide a reliable estimate of the effects
of a potential pre-installation of a LinkedIn application on Windows PCs
because those partnerships involved forms of cooperation different from pre-
installation of the LinkedIn application.
295
In addition, the Notifying Party's
argument that pre-installation of the Skype application on Windows PCs did not
290
Form CO, Table 14.
291
See Case C-3/39.530 - Microsoft (Tying), Commission decision of 16 December 2009.
292
In particular, XING submitted that, in the months following the marketing of Windows Phone devices
having the XING application pre-installed, it experienced a 400% uplift in installations of the XING
application, the large majority of which were due to pre-installation. Moreover, the opportunity for
Windows Phone owners to use log-in credentials from the OS to use the XING application translated
into an increase in directly logged-in traffic and in repeat usage. XING's response to questionnaire to
social networks Q3 of 17 October 2016, question 34.3, and XING's response to RFI n. 17, pages 6-7.
293
XING's response to RFI n. 1, pages 40 and 45.
294
GoldenLine's response to questionnaire to social networks Q3 of 17 October 2016, question 34.1.
295
The information provided by the Notifying Party indicates that the main partnerships at issue
concerned LinkedIn's agreements with Sony, Samsung, Yahoo! and Gmail, none of which involved
pre-installation of the LinkedIn application. See Form CO, footnote 665 and Parties' responseto RFI n.
19, question 1.a.
69
significantly drive new Skype member uptake (either before or after its
acquisition by Microsoft) does not appear to be convincing. In that case, Skype
was preinstalled on more than half of Windows PCs sold to consumers prior to
its acquisition by Microsoft and, therefore, the transaction would have had a
relatively limited impact in terms of additional installed basis for Skype.
296
In
addition, this meant that precise data were available to the Commission as to the
impact of Skype's pre-installation for the purpose of reviewing the proposed
merger.
297
No such data are available in the present case, as LinkedIn has not
been previously installed on Windows PCs or on other devices. Moreover, the
data on users' reactions to Skype's pre-installation and on the impact of Skype's
pre-installation on Skype's market position (whether before or after its
acquisition by Microsoft) cannot be used to reliably predict the impact of
LinkedIn's pre-installation, as LinkedIn and Skype belong to different product
markets, which may differ in terms of user behaviour and other factors
298
in
ways that may influence the impact of pre-installation. As a result, it is not
possible to rely on those data to exclude any meaningful impact of pre-
installation of a LinkedIn application on Windows PCs.
(314) Nevertheless, on the basis of the limited information available, the Commission
notes the following.
(315) First, according to Microsoft's envisaged pre-installation use case scenarios
described in paragraph (308) above, a pre-installed LinkedIn application would
be particularly prominent to Windows PC users, as it would be included both as
a "tile" to the Windows 10 Start menu and as an icon on the desktop. As set out
above, PCs running Windows OS account for more than […]% of new PCs
distributed in the EEA and arguably for a similar share of PCs that currently
exist in the EEA. Accordingly, if Microsoft were to pre-install a LinkedIn
application on all or even some Windows PCs, this would be likely to enhance
LinkedIn's visibility to a very large number of users compared to the pre-
Transaction scenario in which there is no pre-installation of a LinkedIn
application on Windows PCs.
(316) Second, the users towards whom LinkedIn would gain increased visibility are
those that so far have accounted for the largest share of LinkedIn's growth, both
in terms of user base and of usage. Indeed, the majority of users who sign-up to
LinkedIn currently do so on Windows PCs as opposed to other PCs or mobile
devices. This can probably be explained, aside from Windows' very large share
on PCs, by the fact that users find it more convenient to carry out the tasks
associated with signing up to a PSN platform (i.e. typing and pasting career
related information) on a PC than on a mobile device.
(317) According to data submitted by the Notifying Party, the share of LinkedIn
acquisitions occurring on Windows PCs is […] on an EEA-wide basis and even
higher in certain EEA countries, including in Germany ([…]), France ([…]) and
Poland ([…]). These data are consistent with those provided by XING,
according to which 74% of new acquisitions to its PSN in the German-speaking
area
299
take place via PCs running the Windows OS.
300
As regards PSN user
296
Case M.6281 – Microsoft/Skype, Commission decision of 7 October 2011, paragraph 161.
297
Case M.6281 – Microsoft/Skype, Commission decision of 7 October 2011, paragraph 162.
298
See e.g. the explanation provided in footnote 326 below.
299
Including Germany, Austria and Switzerland.
300
XING's response to RFI n. 1, question A.1.a (page 10).
70
engagement, the data submitted by the Notifying Party suggest that, in Q2 2016,
[…] of LinkedIn micro-sessions
301
and […] of LinkedIn page views in the EEA
occurred on PCs, while the remaining micro-sessions and page views occurred
on mobile devices.
302
(318) Thus, Windows PCs constitute the most important channel for PSNs to acquire
new customers, accounting for more than half of total LinkedIn sign-ups. In
addition, Windows PCs represent an important channel to ensure user
engagement, accounting for more than 30% of LinkedIn usage. While these
shares may be partly eroded in the next years due to the further switching of
some users to smart mobile devices, they are likely to remain significant.
Moreover, there is no indication that Windows will not continue to be installed
on the vast majority of new PCs that are distributed in the EEA.
(319) Accordingly, the Commission considers that the pre-installation of a LinkedIn
application on Windows PCs could potentially lead to a meaningful increase in
LinkedIn membership and user activity.
(320) Third, it is doubtful whether competing providers of PSN services may be able
to deploy effective counterstrategies to offset the effects of pre-installation of a
LinkedIn application on Windows PCs. This is because OEMs may lack
incentives to pre-install a second PSN application which would, in essence,
duplicate the same functionalities as the LinkedIn application. According to
PSN respondents to the market investigation, users would be unlikely to decide
spontaneously to download an application that is not already pre-installed.
303
As
competing PSN providers would not be afforded equivalent access to customers
via pre-installation, they would eventually be foreclosed from the market.
(321) In light of the above, it is likely that the merged entity would be able to pre-
install a LinkedIn application on Windows PCs and that this may foreclose
competing providers of PSN services in the EEA or in some EEA countries.
Integration of LinkedIn features into Office, including denial of access to
Microsoft APIs
(322) Currently, Microsoft enables providers of third-party applications to build add-
ins for Outlook and for other core Office products by making specific sets of
APIs publicly accessible for this purpose (the "Outlook APIs" and, together with
the APIs for Word, Excel and Power Point, the "Office APIs").
304
For example,
XING currently offers an add-in for Outlook.
305
In addition, Microsoft makes
available the "Microsoft Graph", a unified set of API that includes APIs from
301
As explained by the Notifying Party, a microsession is a collection of full page views made by a single
user on a single browser on a single portal. A session is closed and a new one created when (i) there is
>30 minute of (full page) inactivity; (ii) the midnight PST hard cut-off; (iii) a different member signs
in on the same browser; or (iv) the same member starts using a different browser or browsing different
portal instead. (In some cases, members can be in 2+ concurrent sessions if they have multiple open
browsers logged into LinkedIn). The microsession statistics reflect members logged-in on the
LinkedIn.com platform. Form CO, footnote 131.
302
Form CO, Figures 16 and 17.
303
Responses to questionnaire to social networks Q3 of 17 October 2016, question 34.2.
304
According to the Notifying Party, there are currently over 200 add-ins in the Microsoft Office Store
supporting diverse capabilities including social networking integration, CRM, package tracking, digital
signature and fax processing, data encryption, and others.
305
See https://store.office.com/en-001/app.aspx?assetid=WA104379526&ui=en-US&rs=en-
001&ad=US&appredirect=false.
71
several Microsoft services (the "Microsoft Graph APIs").
306
The Office APIs
and the Microsoft Graph APIs will be referred to below as "Microsoft APIs".
(323) As regards technical feasibility, integration of LinkedIn features into Office and
denial of access to Microsoft APIs to competing PSN service providers do not
appear to raise any particular technical difficulty for the merged entity. In fact,
Microsoft has already identified a number of use case scenarios that involve
integration of LinkedIn features into Outlook and some other Office products.
307
According to the Notifying Party, the main use cases that allow for the
integration of LinkedIn features into Office include, for example: (i) seeing
information on existing LinkedIn contacts within Outlook (and other core Office
365 products, such as Word and Excel); (ii) adding new LinkedIn contacts from
Outlook (and other core Office 365 products, such as Word and Excel);
(iii) making suggestions for new LinkedIn contacts on the basis of the Outlook
address book (subject to user consent).
308
In terms of timing, according to
information submitted by the Notifying Party, these use case scenarios could be
implemented in a relatively short timeframe, […].
309
With respect to the merged
entity's possible decision to deny competing PSN service providers access to
Microsoft APIs, such decision could likely be implemented at any moment.
(324) Regarding the impact of integration of LinkedIn features into Office, competing
providers of PSN services who replied to the market investigation confirmed
that such integration would be likely to lead to a significant increase in the
number of LinkedIn's members and their engagement. In support of their claims,
PSN respondents submitted that displaying LinkedIn profiles within Outlook
would be "very useful" for users, while displaying LinkedIn profiles of
authors/processors of Excel, PowerPoint or Word documents would be
somewhat useful or not useful.
310
Moreover, PSN respondents submitted that
integration would provide LinkedIn with valuable insights on social interactions
within other Microsoft products based on the matching (also by means of
introducing a unified login and/or address book function) of the LinkedIn user
data with the user data available to Microsoft through its other services,
including a user's address book, connections, email correspondents, meeting
attendees, etc.
311
On this basis, LinkedIn would be able to map a user's network
and recommend with a high degree of precision new relevant LinkedIn
306
The “Microsoft Graph” is a branded API that enables applications to access user data and content
stored within a given cloud application to enable that user to obtain functionality and services using
that data. Microsoft and third-party developers can use the Microsoft Graph to access Microsoft
services, to make “calls” on data and content created by users of these services and to make such data
and content accessible to the tenant’s users.
307
Form CO, Annex 27.
308
Form CO, Annex 27.
309
Form CO, Annex 27.
310
Responses to questionnaire to social networks Q3 of 17 October 2016, question 36.
311
For completeness, the Commission notes that the use by Microsoft of the user data available to it to
this end would in any event always be subject to applicable data protection rules. Moreover, Microsoft
submits that it only has access to user data to the extent to which such data is stored in the Microsoft
cloud, not if user data that is stored on-premise (that is to say on an end user's device). In addition, to
the extent to which Microsoft has access to certain user data as a result of providing professional cloud
services to business customers (for example, the user data concerning the employees of a given
company using Microsoft's cloud services), Microsoft would also require the agreement of the relevant
business customer. Given the importance of data integrity and confidentiality for these customers and
the intense competition in the provision of professional cloud services, it is at best uncertain whether
Microsoft would be able to obtain any such additional consent.
72
connections, thereby increasing the size of LinkedIn's network and user
activity.
312
Moreover, according to PSN respondents, if combined with the
merged entity's denial of access to Microsoft APIs, such integration could
eventually foreclose competing PSN service providers. In particular, one
respondent submitted that Microsoft APIs need to be published/specified for
third-party developers to be able to use them and that Microsoft controls APIs'
authentication lists and could selectively block APIs' users.
313
(325) While the Notifying Party argues that the integration of LinkedIn features into
Office would not drive significant growth in members and usage, it has been
unable to provide any precise estimates in that respect. In addition, given that
LinkedIn features have not previously been integrated into Office products in
the way envisaged by Microsoft post-Transaction and that there is currently no
LinkedIn add-in for Office, there are no data available to determine with a
sufficient degree of certainty which impact the integration of LinkedIn features
into Office would have on LinkedIn's position.
(326) The Notifying Party's argument that LinkedIn's partnerships with Samsung,
Sony, Yahoo and Gmail did not significantly drive LinkedIn users' uptake or
engagement
314
cannot be used to reliably predict the impact of the integration of
LinkedIn features into Office. Indeed, there appears to be substantial differences
between LinkedIn's previous partnerships and Microsoft's envisaged integration
of LinkedIn features into Office. The partnerships with Samsung and Sony
enable the display of LinkedIn information on mobile devices, not on PCs.
315
Moreover, Yahoo! and Gmail are email services that tend to be predominantly
used for private purposes, while Outlook is very widespread among the
categories of users for whom integrated LinkedIn features would likely be most
valuable (i.e. professionals). Finally, using the LinkedIn features on
Sony/Samsung phones and within Gmail and Yahoo! requires some particular
form of activation on the part of the user.
316
This does not appear to be
necessarily the case for the integration envisaged by Microsoft.
317
(327) Nevertheless, on the basis of the information available, the Commission notes
the following.
312
Responses to questionnaire to social networks Q3 of 17 October 2016, question 35.1.
313
XING's response to RFI n. 2, page 4.
314
According to the Notifying Party, the LinkedIn integration features have been activated by less than
[…] of Yahoo! Mail and Gmail users and by less than […] of Samsung and Sony smartphone holders.
315
LinkedIn's partnership with Samsung allows users, when they email or have a calendar appointment
with another LinkedIn member, to view those members' LinkedIn information on their smartphone
screen. LinkedIn's partnership with Sony concerns LinkedIn for Info-eye, a plug-in that allows users to
scan a contact's business card and automatically access their LinkedIn profile on Sony Xperia
smartphones. Notifying Party's submission "Second Briefing Paper – No competition concerns in
social networking services" of 3 November 2016.
316
In order to display the LinkedIn features, Samsung and Sony smartphone holders must first download
the LinkedIn application and then actively syncronise their LinkedIn account with their phone's
applications. Similarly, activating LinkedIn features within Gmail and Yahoo! mail services requires
users to either actively synchronise their accounts (in the case of Yahoo!Mail), or to download a
browser extension (in the case of Gmail). Parties' response to RFI n. 19, question 1.a and 1.c.
317
According to the Notifying Party, Microsoft is still at the early stages of discussing and planning
integration use case scenarios, and the Parties have not yet designed a particular flow or user interface.
Accordingly, the Notifying Party submits that it remains to be determined exactly what steps a user
will need to take to activate the integration features. Parties' response to RFI n. 19, question 1.c.
73
(328) First, according to Microsoft's envisaged integration of LinkedIn features into
Office, the LinkedIn features would be particularly prominent to users of
Outlook (and potentially other flagship Office products). As set out in
paragraphs (290)-(293) above, Office products account for more than […]% of
productivity software in the EEA. Accordingly, if Microsoft were to integrate
LinkedIn features into Outlook (and other flagship Office products), this would
likely enhance LinkedIn's visibility to a very large number of users compared to
the pre-Transaction scenario in which there is no such integration. In addition,
the ability (subject to user consent) to access Outlook users' address books and
suggest new LinkedIn connections on this basis may enable the merged entity to
significantly expand the size of its PSN.
(329) Second, if the merged entity were to deny competing providers of PSN services
access to the Outlook API (and potentially other Microsoft APIs), such
providers would likely have no counterstrategy at their disposal to sufficiently
counter the merged entity's actions. Given that no alternative productivity
software suite enjoys a degree of user penetration comparable to Outlook's (and
Office's in general), competing PSN service providers would not be able to
replicate the experience that they would otherwise be able to create with
Outlook APIs (and potentially other Office APIs).
(330) In light of the above, it is likely that the merged entity would be able to integrate
LinkedIn features into Office, while denying competing PSN service providers
access to Microsoft APIs, and that this may foreclose competing providers of
PSN services in the EEA or in some EEA countries.
(ii) Incentive to foreclose
(331) Both the pre-installation of a LinkedIn application on Windows PCs and the
integration of LinkedIn features into Office are specifically envisaged by
Microsoft post-Transaction. As a result, the Commission considers that, post-
Transaction, Microsoft would have an incentive to engage both in the pre-
installation of a LinkedIn application on Windows PCs and in the integration of
LinkedIn features into Office, without it being necessary to conduct a detailed
analysis of Microsoft's incentive to engage in those practices.
(332) Indeed, Microsoft appears to be well aware of the benefits that those practices
would create in terms of increased user base and activity of LinkedIn and, as a
result, of increased monetisation opportunities. This is reflected in Microsoft's
internal documents relating to the Transaction.
318
For example, an internal
document of Microsoft titled "How Mumbai can help London" mentions several
"integration opportunities", including with Office 365 […], leading to benefits
such as "signups, connection density".
319
According to another internal
document of Microsoft, which sets out LinkedIn's expected synergies in terms of
new members and engagement, Microsoft expects LinkedIn to gain […]
(compared to LinkedIn's situation absent the Transaction).
320
(333) Moreover, Microsoft's envisaged integration of LinkedIn features into Office
may create an incentive to deny competing PSN service providers access to
Microsoft APIs to prevent them from achieving similar levels of integration.
Internal documents of Microsoft show indeed that Microsoft considers
318
See also Form CO, Section 6, paragraph 238 and following, paragraph 270 and following, Annex 27.
319
Form CO, Annex 5, document 48, page 3.
320
Form CO, Annex 5, document 4, page 7.
74
"competition" among the current "issues" for countries such as Germany (where
there is currently a PSN service provider competing with LinkedIn) and that
"drive growth" in Germany is among the "highest priority needs" regarding
Europe.
321
(334) In addition, there do not appear to be any factors that are likely to reduce or
eliminate the incentive of the merged entity to engage in such practices.
(335) First, it is unlikely that the merged entity would incur any significant losses by
engaging in those practices. Even assuming that having a LinkedIn application
pre-installed on their Windows PCs or having LinkedIn features integrated into
Office would dissatisfy or frustrate certain users of Windows PCs or Office, it is
unlikely that this would prompt a sufficient number of users to abandon
Windows or Office and make such pre-installation and integration unprofitable
for Microsoft.
(336) Second, it appears that Microsoft would not abstain from engaging in those
practices out of fear of violating antitrust rules or other legal provisions. As
explained in paragraphs (308) and (323) above, Microsoft has already expressed
a positive intention internally to engage in those forms of conduct, as reflected
in a large number of its internal documents.
322
It also cannot be excluded that
Microsoft would be discouraged from denying competing PSN providers access
to Microsoft APIs on related legal grounds.
(337) In light of the above, it is likely that the merged entity would have the incentive
to pre-install a LinkedIn application on Windows PCs and to integrate LinkedIn
features into Office while denying access to Microsoft APIs, thereby foreclosing
competing providers of PSN services in the EEA or in some EEA countries.
(iii) Overall likely impact on effective competition
(338) As explained in Sections (i) and (ii) above, post- Transaction, the merged entity
is likely to have both the ability and the incentive to foreclose competing
providers of PSN services both by (i) pre-installing a LinkedIn application on
Windows PCs; and by (ii) integrating LinkedIn features into Office, while
denying access to Microsoft APIs. As also explained above, each of those
practices would likely lead to an increase in the size and usage of LinkedIn's
PSN platform in a way that competing providers of PSN services would be
unable to match.
(339) In the market investigation, concerns were raised that, once LinkedIn's PSN
would grow to the point that the balance would "tip" in LinkedIn's favour,
competing PSN service providers would be unable to compete effectively
321
Form CO, Annex 5, document 12, page 11.
322
The Court of Justice has held in connection with a merger involving non-horizontal conglomerate
effects that the Commission should comprehensively examine the likelihood of the adoption of a
certain conduct by the merged entity taking into account both the incentives to adopt such conduct and
the factors liable to reduce those incentives, including the possibility that the conduct is unlawful.
However, according to the Court of Justice, the Commission is not required to carry out a detailed
assessment in this respect. According to the Court, "an assessment intended to establish whether an
infringement of Article [102 TFEU] is likely and to ascertain that it will be penalised in several legal
orders would be too speculative and would not allow the Commission to base its assessment on all of
the relevant facts". See Case C-12/03 P, Commission v Tetra Laval [2005] ECR I-987, paragraphs 74-
75.
75
against LinkedIn and new entrants would lack the ability or the incentive to
enter the markets for PSN services within the EEA.
323
(340) On the basis of the information available, the Commission considers that
LinkedIn's growth and the negative impact on competing PSN providers could
be further enhanced by virtue of the network effects that characterise the market
for PSN services.
(341) Network effects occur when the value of a product or service for a customer
increases when the number of other customers also using it increases.
324
In the
present case, network effects are likely to play an important role in light of the
nature of PSN services. Indeed, professionals tend to benefit as more
professionals join the network and use it actively, as this is likely to translate
into a higher number of professional contacts, of profile views and of
recruitment opportunities. The majority of respondents to the market
investigation confirmed the importance of network effects for PSN services and
submitted that the size of the user base is a very important parameter of
competition in PSN services.
325
(342) The existence of network effects as such does not a priori indicate a competition
problem in the market affected by a merger. Such effects may however raise
competition concerns in particular if they allow the merged entity to foreclose
competitors and make more difficult for competing providers to expand their
customer base. Network effects have to be assessed on a case-by-case basis.
(343) In the present case, it appears likely that network effects could potentially
strengthen the foreclosure of competing providers of PSN services that currently
exist in certain EEA countries or of potential new entrants. According to the
concerns put forward by certain respondents to the market investigation, as the
number of members of LinkedIn would grow, additional users would be induced
to join LinkedIn and to generate activity on its platform. By contrast,
increasingly fewer users may be induced to join competing PSN service
providers, as those providers would become less attractive in terms of size of
their networks and of recruitment opportunities. As a result, this could lead to a
slowing down in the growth of competing PSNs’ member bases and eventually
to a decline in the activity of competing PSNs’ actual members who would
instead become active on LinkedIn. This trend could continue up to the point
where the market would "tip" in favour of LinkedIn's network and LinkedIn's
already strong position would become entrenched.
(344) Furthermore, it is doubtful whether the impact of network effects could be
sufficiently mitigated by multi-homing (i.e. by the fact that users would choose
to be active on more than one PSN platform) or by the entry of potential new
PSN service providers.
(345) As regards multi-homing, actively engaging on a PSN platform usually requires
users to curate and update their profiles as well as to build and interact with new
contacts. As this can require significant time on the part PSN users, it can in
some cases act as a disincentive to multi-homing between PSN platforms. This
323
Responses to questionnaire to social networks Q3 of 17 October 2016, questions 32, 33 and 35.
324
Non-Horizontal Guidelines, paragraph 62 and footnote 64. Examples include communication devices,
specific software programmes, products requiring standardisation, and platforms bringing together
buyers and sellers.
325
Responses to questionnaire to social networks Q3 of 17 October 2016, questions 21, 26 and 30.
76
factor distinguishes PSN services from consumer communications services.
326
While the data submitted by the Notifying Party indicate that a meaningful share
of LinkedIn members multi-home with another PSN platform (in those EEA
countries where such platform is present), this is not entirely confirmed by the
results of the market investigation. Already today, according to the PSN
respondents to the market investigation, although many users have accounts on
two PSNs, they actively use only one of them
327
or, in any event, they view one
of them as their "main network".
328
After the Transaction, should the growth of
LinkedIn's platform make competing PSNs less attractive to users, such users
may not consider it worth investing the effort associated with updating a profile
on a competing PSN. Hence, multi-homing may decrease as LinkedIn's market
position is strengthened.
(346) As regards potential entry of new PSN service providers, although launching a
new PSN service may not necessarily entail significant difficulties, the results of
the market investigation indicate that achieving a sizeable user base may
constitute a high, sometimes unsurmountable, barrier to entry.
329
Moreover, it is
doubtful whether providers active in neighbouring markets would have both the
ability and the incentive to enter the market for PSN services. Indeed, as
explained in paragraph (110) above, the market investigation indicates that it
may be difficult for providers of SN services without a professional focus to
transform into a PSN service provider. Also, as explained in paragraph (114)
above, the conversion of an enterprise social network into a PSN would not
necessarily be immediate and straightforward.
(347) Therefore, neither multi-homing nor potential entry may be sufficient to prevent
the market for PSN services from "tipping" in favour of LinkedIn's PSN.
Moreover, once the market has "tipped", it may become even more difficult for
actual competing providers of PSN services to regain their ability to compete
and for potential competitors to enter the market, given the increased barriers to
entry and expansion.
(348) As regards the ultimate impact on consumers and choice, the Commission notes
the following. Should the market for PSN services reach such "tipping point",
LinkedIn's platform would remain the only PSN service provider in the EEA
today and potentially in the coming years. The possible detrimental effect on
consumers would be twofold.
(349) First, this would entail a substantial reduction of consumer choice, as LinkedIn's
platform would remain the only PSN service provider available to users in the
EEA, with no or limited prospects of entry by new PSN service providers.
(350) Second, to the extent that these foreclosure effects would lead to the
marginalisation of an existing competitor which offers a greater degree of
326
In Facebook/WhatsApp, the Commission found, in relation to consumer communications apps, that
multi-homing is facilitated by the ease of downloading such apps, which is generally free, easy to
access and does not take up much capacity on a smartphone. Also, the Commission found that using
multiple consumer communications apps is easy, since a user does not have to log in each time, when
switching an app, and the messages are "pushed" (that is, delivered automatically) onto a user's device.
See Case M.7217 – Facebook/WhatsApp, Commission decision of 3 October 2014, paragraph133.
327
Responses to questionnaire to social networks Q3 of 17 October 2016, question 28.2.
328
XING's presentation of 30 August 2016 "Merger Microsoft/LinkedIn – Competitive Considerations,
XING Presentation to DG COMP", page 37.
329
Responses to questionnaire to social networks Q3 of 17 October 2016, question 30.
77
privacy protection to users than LinkedIn (or make the entry of any such
competitor more difficult), the Transaction would also restrict consumer choice
in relation to this important parameter of competition when choosing a PSN.
330
By way of example, the results of the Commission's investigation revealed that,
today, in Germany and Austria, Xing seems to offer a greater degree of privacy
protection than LinkedIn.
331
For instance, during the registration process, XING
asks users to actively accept XING's privacy policy and Terms & Conditions by
ticking a box, whereas LinkedIn users accept LinkedIn's privacy policy
automatically when they press the button "join now". Moreover, when XING
introduces new services which have an implication on how it collects and/or
uses its members' data, it explicitly seeks active consent from the members. In
addition, regardless of whether members give their consent in such specific
cases or not, they will be able to continue to use XING as such without losing
any of the functions to which they previously had access. In contrast, when
LinkedIn makes changes to its collection, storing, processing or usage of
personal data, LinkedIn only informs the members of those changes and
considers that LinkedIn members agree with those changes, if they continue to
use LinkedIn's services after they have been notified of the changes.
332
(iv) Conclusion
(351) For all the above reasons, the Commission concludes that the possible
foreclosure of PSN service providers stemming from each of (i) the pre-
installation of a LinkedIn application on Windows PCs and (ii) the integration of
LinkedIn features into Office, in combination with the denial of access to Office
APIs, will have a negative impact on effective competition in the markets for
PSN services in the EEA.
(352) In light of the above, the Commission considers that the Transaction raises
serious doubts as to its compatibility with the internal market as regards the
market for PSN services in the EEA.
330
The results of the market investigation have indeed revealed that privacy is an important parameter of
competition and driver of customer choice in the market for PSN services (Responses to questionnaire
to social networks Q3 of 17 October 2016, questions 18 and 21). The finding of the importance of
privacy as parameter of competition is consistent with the Commission's findings in
Facebook/WhatsApp (Case M.7217 – Facebook/WhatsApp, Commission decision of 3 October 2014,
paragraphs 87 and 102 and footnote 79) in relation to consumer communication services.
331
LinkedIn's response to RFI n. 12, XING's response to RFI n. 13.
332
For completeness, the Commission notes that, in the course of the market investigation, concerns were
raised that companies based outside the EEA may be able to try to circumvent European data
protection rules (either by not fully complying with such rules or by interpreting them in their favour),
while European competitors need to fully comply with EU data protection rules. In this respect, it
suffices to say that LinkedIn post-Transaction remains subject to European data protection rules, in
particular to Irish data protection law with respect to the data collection, processing and usage of
European users. Moreover, in the case where Irish data protection rules were less restrictive than those
in other Member States, the Commission notes that the newly adopted General Data Protection
Regulation is directly applicable and therefore the scope for divergence between Member States'
national data protection laws will be reduced, including in their enforcement. Finally, if Microsoft
were to transfer and process personal data of LinkedIn' members outside of the EEA, the rules of the
national laws implementing the Data Protection Directive and, as of 25 May 2018, the General Data
Protection Regulation will still apply.
79
including national job agencies.
334
Contrary to the Notifying Party's estimates,
many respondents (16 out of 33) to the market investigation named LinkedIn as
one of the five if not the strongest provider of online recruitment services in the
EEA or in their respective EEA country.
335
4.2.4.2 Conglomerate non-coordinated effects
(356) In light of Microsoft's high market shares in the markets for (i) OSs for PCs and
(ii) productivity software (see paragraph (278) above), the Commission
considers that Microsoft has a strong market position within the meaning of
paragraph 93 of the Non-Horizontal Guidelines on those markets. Accordingly,
the Commission investigated if, after the Transaction, the merged entity would
have the ability and the incentive to leverage its strong market position from
those markets to the market for online recruitment services by means of tying,
bundling or other exclusionary practices.
(a) Notifying Party's view
(357) According to the Notifying Party, the Transaction would not lead to foreclosure
of competing online recruitment services providers.
(358) First, the Notifying Party argues that there are no current plans to bundle
LinkedIn’s recruiting solutions with Office or Office 365 products or product
suites and it would not make business sense to do so. The Notifying Party claims
that because of (i) the very significant mismatch in terms of the seats needed for
LinkedIn’s Talent Solutions as compared to Office
336
and (ii) the ratio of
productivity software seats to recruiting solution seats is likely to vary
significantly across organisations, it would be practically and financially not
attractive for Microsoft to bundle LinkedIn's Talent Solutions with Microsoft's
high-volume Office productivity software. To avoid losing the Office business
of organisations with below-average demand for Talent Solution licenses,
Microsoft would have to reduce the bundle price, thereby losing revenue on
sales to organizations with above-average demand for Talent Solution licenses.
(359) Second, the Notifying Party argues that even if Microsoft was to include Talent
Solutions with Office no competition concerns would arise since the market for
recruitment solutions is fragmented and LinkedIn’s estimated share of
recruitment traffic is only approximately [0-5]% at the EEA level. Further, job
seekers and recruiters multi-home and customers do not choose the recruiting
services that they want to use based on whether they form part on an Office suite
of products or integrate with Office.
(360) Finally, the Notifying Party argues that even if Microsoft were to be considered
dominant in productivity software, the integration scenarios planned by
Microsoft only relate to the cloud-based variant of Office, i.e. Office 365, which
to date has only […] monthly active enterprise users.
334
Responses to questionnaire to recruiters and providers of online recruiting services Q4 of 21 October
2016, question 22.2.
335
Responses to questionnaire to recruiters and providers of online recruiting services Q4 of 21 October
2016, question 22.2.
336
According to the Notifying Party, customers pay according to the number of “seats” that they
subscribe to. The great majority of LinkedIn's customers are companies that use Talent Solutions
within their HR department. Whereas recruiting solutions are purchased by HR departments for the
teams involved in recruiting activity only, productivity software is sourced by IT departments for the
organization as a whole covering the often very large number of employees.
80
(b) Commission's assessment
(361) The Commission considers that the Transaction would not result in the
foreclosure of competing online recruitment services providers.
(362) First, the Commission considers that there are no reasons why the merged entity
should not be able to bundle or tie LinkedIn’s online recruitment services with
Microsoft's Office or Office 365 products post-Transaction. The Commission
notes, however, second, that the merged entity's incentive to do so appears rather
limited.
(363) This is because private customers would not be interested in purchasing a
bundle of Office with LinkedIn’s Talent Solutions that are by their nature only
of interest for recruiters or recruitment agencies. With regard to business
customers, the merged entity's incentive to bundle or tie Office 365 with
LinkedIn’s Talent Solutions is likely to be rather limited due to the significant
difference in terms of the seats needed by corporate customers for LinkedIn’s
Talent Solutions (only needed by employees in the HR department) as compared
to Office 365 (needed by almost all employees). However, the Commission
considers that at least for recruiting agency customers the merged entity would
have an incentive to apply a bundling/tying strategy since recruiting agencies
would need a similar number of seats for LinkedIn’s Talent Solutions and Office
365. Consequently, the Commission cannot exclude the merged entity's
incentive to apply a bundling strategy for such customers.
(364) Third, the Commission considers that foreclosure effects would be unlikely.
Based on the Parties' claims (see paragraph (358)) the Commission takes the
view that any tying or bundling strategy by the merged entity with regard to
LinkedIn's recruiting and hiring tools and services and Office/Office 365 would
be limited to recruitment agencies and, hence, would not target a large enough
fraction of the market in order to significantly impede effective competition.
337
In addition, the market for online recruitment services is highly fragmented with
many players present. Further, online recruitment services are differentiated
services since the functionalities offered may vary between different providers.
(365) Fourth, the majority of respondents to the market test considered that multi-
homing and switching between online recruitment services providers is easy for
recruiters and job seekers.
338
Recruiters as well as job seekers have an incentive
to use several online recruitment provider platforms to increase their chances of
finding the suitable job-candidate or suitable job. In addition, the majority of
respondents to the market test considered that entry barriers are low and post-
Transaction there will remain a sufficient number of online recruitment services
providers.
339
(366) That said, it cannot be excluded that the pre-installation of a LinkedIn
application on Windows PCs and/or the integration of LinkedIn features into
Office as identified in Section 4.2.3 above could also have a negative impact on
the market for online recruitment services by strengthening the position of
337
According to the Notifying Party, the large majority of customers of LinkedIn's recruitment solutions
are companies that only use these solutions in its human resources department (see Notifying Party's
response to RFI n.21, question 6 and Form CO, Annex 19).
338
Responses to questionnaire to recruiters and providers of online recruiting services Q4 of 21 October
2016, question 26 to 29.
339
Responses to questionnaire to recruiters and providers of online recruiting services Q4 of 21 October
2016, question 32 to 35.
81
LinkedIn as the only PSN service provider in an EEA country or by potentially
foreclosing competing PSNs.
340
Those forms of conduct may indeed strongly
increase LinkedIn's base of job-candidates and their engagement which in turn
would make LinkedIn more attractive to recruiters. As a result of these network
effects, competing online recruitment services provider would likely be harmed
as they would gradually see their subscriptions to their online recruiting services
decrease or not sufficiently increase.
(367) However, the Commission considers that the commitments offered by Microsoft
(see Section 5.1.3) would also eliminate any such possible anti-competitive
impact (by ensuring that LinkedIn's position as a PSN service provider is not
strengthened as a result of the Transaction). In other words, in so far as
competition on B2C side (market for PSN services) can be preserved,
competition on the B2B side (market for online recruitment services) remains
equally preserved because recruiters would have an incentive to use competing
online recruitment service insofar as a sizeable base of job candidates remain on
those services.
(368) In light of the above, the Commission considers that with regard to a possible
bundling or tying of the LinkedIn's recruiting and hiring tools with Microsoft's
Office or Office 365 products, the Transaction does not raise serious doubts as
to its compatibility with the internal market as regards the market for online
recruitment services in the EEA area. With regard to the effects that the pre-
installation of a LinkedIn application on Windows PCs and the integration of
LinkedIn features into Office may have on effective competition in the market
for online recruitment services in the EEA, any competition concerns that the
Transaction may raise would be addressed by the Microsoft's commitments.
4.2.5 Foreclosure of competing productivity software providers
4.2.5.1 Market shares
(369) As discussed in Section 4.2.3.1. (c), Microsoft, with Office, is by far the largest
provider of productivity software in the EEA under a number of parameters (see
Table 7 above). Therefore, the Commission considers that Microsoft has at least
a strong market position within the meaning of paragraph 93 of the Non-
Horizontal Guidelines in the productivity software market in the EEA.
4.2.5.2 Vertical non-coordinated effects (input foreclosure)
(370) During the market investigation a concern has been raised that Microsoft could
further increase its dominant position in the market for productivity software by
using LinkedIn full data. Specifically, Microsoft would be able to combine
Microsoft's data with LinkedIn full data which would constitute in the near
future an important input within the meaning of paragraph 31 and 34 of the
Non-Horizontal Guidelines for the provision of ML in productivity software
solutions.
(371) On this basis a concern has been raised that post-Transaction Microsoft could
restrict access to LinkedIn full data for the purposes of ML in competing
productivity software solutions, thereby making it harder for other providers of
340
In any event, antitrust rules, in particular Article 102 TFEU and Article 54 of the EEA Agreement will
continue to apply to the merged entity after the closing of the Transaction.
82
productivity software solutions to compete and to bring innovation in the
market.
341
(a) Notifying Party's view
(372) The Notifying Party submits that post-Transaction the merged entity would have
no ability or incentive to foreclose competition in the market for the provision
productivity software solutions because the LinkedIn full data is not a must-have
for productivity software solutions and its added value remains unproven at this
stage.
342
Moreover, there are alternative data sources available in the market
from other vendors.
343
(b) Commission's assessment
(i) Ability and incentive to foreclose
(373) While it is unclear whether Microsoft may have the incentive to foreclose
competing providers of productivity software, the Commission considers that
post-Transaction the merged entity would not have the ability to foreclose
competing providers of productivity software as, in any event, by reducing
access to LinkedIn full data, it is unlikely to negatively affect the overall
availability of data for ML in productivity software solutions.
(374) First, LinkedIn does not appear to have a significant degree of market power
within the meaning of paragraph 35 of the Non-Horizontal Guidelines in any
potential relevant upstream market, which in this case would be a hypothetical
market or segment for provision of data for the purposes of ML in productivity
software solutions.
344
Indeed, as stated in paragraph (203), LinkedIn does not
currently license any data to any third party. Moreover, as stated in paragraph
(254), absent the Transaction, it was not planning to license its full data, or a
subset thereof, to any third party, including for ML purposes.
(375) Second, as described in paragraph (255), data protection rules in the EEA may
limit Microsoft's ability to have access to the LinkedIn full data.
(376) Third, the Commission considers that LinkedIn full data, or a subset thereof,
cannot be qualified as, and is not likely to become in the next two to three years,
an important input within the meaning of paragraph 34 of the Non-Horizontal
Guidelines with respect to the provision of ML in productivity software
solutions. As described in paragraph (258), Microsoft does not have plan from
its synergies documents to use LinkedIn full data. Therefore, Microsoft may not
have the incentive to use the entire dataset itself.
(377) Moreover, the respondent raising this concern did not explain how LinkedIn
data could become important for the future in relation to productivity software
and ML functionalities, nor the Commission's investigation shed any further
light on this point.
(378) In addition, the majority of respondents to the market investigation, including
the main competitors in this space, expect the effects of the Transaction on their
company, as well as on the market for the provision of productivity software, to
341
Salesforce's response to RFI n. 1, question 20.
342
Form CO, Section 6, paragraph 701.
343
Notifying Party's response to RFI n.21 and Annex Q.1 of 16 November 2016.
344
This would be the narrowest segment where LinkedIn, is providing its data for machine learning,
would be active and which is relevant for the assessment of the present case.
83
be neutral.
345
No other respondent to the market investigation raised similar
concern in relation to the potential use of LinkedIn full data as input for the
provision of ML in productivity software solutions.
(379) The Commission therefore concludes that the merged entity will likely not have
the ability to foreclose competing productivity software providers by not
providing access to LinkedIn full data.
(ii) Overall likely impact on effective competition
(380) The Commission considers that the Transaction is unlikely to have an overall
negative impact on effective competition in the market for productivity software
solutions, as any potential restriction of access to LinkedIn full data, or subset
thereof, is unlikely to lead to consumer harm. First, as described in paragraph
(378) above, in the market investigation the majority of respondents considered
that the Transaction would not have any impact on their company or on the
market for the provision of productivity software. Second, as described in
paragraph (276), ML requires access to multiple data sources to provide
customers useful insights. LinkedIn is only one of such data sources and there
are alternative third party data sources available in the market.
(iii) Conclusion
(381) In light of the above, the Commission considers that the Transaction does not
raise serious doubts as to its compatibility with the internal market as a result of
input foreclosure effects to the detriment of providers of productivity software
solutions.
346
4.2.5.3 Conglomerate non-coordinated effects
(382) The Commission also investigated whether the conduct discussed in Section
4.2.3, such as the bundling or tying of LinkedIn's features with Microsoft's
productivity software, could strengthen Microsoft's strong position in the market
for productivity software and its possible segments (see market shares in Section
4.2.3.1.(c) and following). The concern has been raised that the ultimate
outcome of such practices would be that Microsoft's productivity software
would be improved by the pre-installation and integration of LinkedIn in a way
that could not be matched by competing providers of productivity software.
347
(a) Notifying Party's view
(383) According to the Notifying Party, the integration of LinkedIn features into
Office will not provide the combined entity with the ability to foreclose rivals in
the market for productivity software.
(384) First, the Notifying Party argues that social networking services (including
LinkedIn’s functionalities) are not a “must-have” element of productivity
software. Productivity software like Word, PowerPoint, and Excel is about
creation of content while social networking services enable sharing of content –
a separate and logically subsequent step. In addition, the Notifying Party claims
345
Responses to questionnaire to productivity app providers Q9 of 21 October 2016, questions 2 and 3.
346
For the sake of clarity, antitrust rules, in particular article 102 TFEU will continue to apply to the
merged entity after the closing of the Transaction, regardless of the outcome of the present assessment
under the Merger Regulation.
347
SoftMaker’s response to questionnaire to productivity app providers Q9 of 21 October 2016, question
2.1 and 3.1.; Salesforce's responses to RFI n.1, question 22, and to market test questionnaire Q22 of 17
November 2016, question 4.1.
84
that customers do not choose productivity software solutions on the basis of
whether social networking features are integrated, but on the basis of price,
feature quality, reliability, service and the overall ability to enhance efficiency
and productivity in the workplace.
(385) Second, the Notifying Party argues that Microsoft’s competitors offer attractive
and competitive productivity solutions and will continue to do so post-
Transaction. For example, Google already integrates its social networking
services, Google Hangouts and Google+, in its productivity suite, Google Apps
at Work. In addition, LinkedIn has partnership agreements with productivity
software vendors other than Microsoft, e.g., Google (for Gmail) and Yahoo! (for
Yahoo! Mail). These agreements enable email users to see LinkedIn account
information for the people they email with from within their email client. These
functionalities are comparable to the integration of LinkedIn features that
Microsoft plans to introduce in Outlook post-transaction. LinkedIn will have
every incentive to maintain these relationships because of the strength of Gmail
and Yahoo!.
(386) Finally, the Notifying Party argues that users can and will be in the future able
to access LinkedIn services from any device regardless of the productivity
software they use. Like other social networking services, LinkedIn relies on
openness and easy access. Any strategy that limits accessibility to one channel
(e.g., Microsoft’s productivity solutions) would hamper the growth of
LinkedIn’s user base and limit potential traffic. It would thus run counter to one
of Microsoft’s key objectives underlying the Transaction, namely to grow
LinkedIn’s existing business.
(b) Commission's assessment
(387) The Commission considers that the Transaction would not result in the
foreclosure of competing productivity software providers through the conduct as
described in Section 4.2.3., such as the bundling or tying of LinkedIn's features
with Microsoft's productivity software.
(i) Ability and incentive to foreclose
(388) While it is unclear whether Microsoft may have the incentive to bundle or tie
LinkedIn's PSN features with Microsoft's productivity software, the
Commission considers that Microsoft would not have the ability to foreclose its
competitors.
(389) First, the possibility to offer additional social networking features like those
offered by LinkedIn does not appear to be essential for competing providers of
productivity software in order to compete effectively against Microsoft.
Productivity software solutions are intended to provide solutions for the creation
of documents rather than the sharing of content. Customers do not choose
productivity software solutions on the basis of whether social networking
features are integrated, but on the basis of price, feature quality, reliability,
service and the overall ability to enhance efficiency and productivity in the
workplace. With regard to email and calendar software (like Outlook) the
Commission notes that the lines between creating and sharing content are more
blurred. However, the Commission considers that the factors for consumers to
choose email and calendar software are predominantly security and reliability
and not (at least at this stage) social networking features. The Commission
therefore considers that access to social networking features are not a "must
have" and do not influence customer's choice such that it would provide the
85
merged entity with the ability to foreclose competing productivity software
providers.
(390) Second, even if social networking features were to be regarded as essential for
productivity software, the Commission notes that other providers of productivity
software already offer such software with social networking features. For
example, Google already integrates its social networking services, Google
Hangouts and Google+ in its productivity suite, Google Apps at Work.
348
In
addition, LinkedIn has partnership agreements with productivity software
vendors other than Microsoft, e.g., Google (for Gmail) and Yahoo! (for Yahoo!
Mail) and the probability that LinkedIn will end these partnership agreements
post-Transaction seems to be low against the background that this would hamper
the growth of LinkedIn’s user base and limit potential traffic and, hence, runs
counter Microsoft’s key objectives underlying the proposed merger, namely to
grow LinkedIn’s existing business.
(391) The Commission therefore concludes that the merged entity will likely not have
the ability to foreclose competing productivity software providers by bundling
or tying LinkedIn's features with Microsoft's productivity software.
(ii) Overall likely impact on effective competition
(392) The Commission considers that the Transaction is unlikely to have an overall
negative impact on effective competition in the market for productivity software
solutions. Apart from the fact that Microsoft does not have any ability to
foreclose competing providers of productivity software, any potential bundle of
LinkedIn features with Microsoft's productivity software, is unlikely to lead to
consumer harm in the form of higher prices or less choice. This is confirmed by
the market investigation, since the majority of competing providers of
productivity software considered that the Transaction would not have any
impact on their company or on the market for the provision of productivity
software.
349
(iii) Conclusion
(393) In light of the above, the Commission considers that the Transaction does not
raise serious doubts as to its compatibility with the internal market as regards the
market for productivity software in the EEA.
4.2.6 Foreclosure of competing enterprise communications service providers
(394) The Commission investigated the likelihood that post-Transaction Microsoft
will integrate LinkedIn features into its enterprise communications product,
Skype for Business ("SfB"), such as communicating instantaneously with
LinkedIn users or finding experts on LinkedIn for reviewing documents
produced with Microsoft tools. Such functionalities would not be offered to
competing providers of enterprise communications services. The concern has
been raised that the ultimate outcome of this conduct would be that Microsoft's
enterprise communications product, SfB, would be improved by the addition
and integration of LinkedIn features in a way that could not be matched by
competing providers of enterprise communications services since they do not
have access to LinkedIn.
350
348
Form CO, Section 6, paragraph 703.
349
Responses to questionnaire to productivity app providers Q9 of 21 November 2016, question 2 and 3.
350
Cisco's submission of 4 November 2016; Broadsoft's submission of 24 November 2016.
86
4.2.6.1 Market shares
(395) Microsoft is active in the enterprise communications market through its SfB
product, and its predecessors Microsoft Lync and Microsoft Office
Communicator. SfB is sold as part of various software suites that included other
products and services, such as Microsoft Office.
351
(396) Microsoft's worldwide market share in revenues, was around [10-20]% in 2015.
Microsoft's main competitors are Cisco ([10-20]%), Avaya ([5-10]%), IBM ([0-
5]%). The remainder of the market is very fragmented.
352
(397) Regarding the possible segmentation of enterprise communications services by
platform or by OS, the Notifying Party was unable to provide disaggregated
market shares, as customers typically purchase enterprise communications
services as an integrated solution which they use across devices and OSs.
Nonetheless, Microsoft does not expect its share based on such segmentations to
be significantly higher than its shares for the whole market.
353
In relation to the
segmentation by functionality, the Notifying Party was unable to estimate
market shares by each functionality, since its SfB offering consolidates all
communications functionalities and does not record or report revenues for each
functionality.
354
The Notifying Party though submits that, according to IDC, a
market research, analysis and advisory firm, in 2015 Microsoft's market share in
terms of revenues for the functionalities covered under "converged
conferencing", which includes virtually all enterprise communications
functions,
355
amounted to [20-30]% worldwide and [30-40]% in the EMEA
region (which constitutes a reasonably good proxy for the EEA market share).
4.2.6.2 Conglomerate non-coordinated effects
(a) Notifying Party's view
(398) The Notifying Party submits that, post-Transaction, Microsoft will lack the
ability to foreclose competitors in enterprise communications services because
(i) LinkedIn features are not a must-have element of enterprise communications
which, in a currently competitive market, will not drive consumer choice of
enterprise communication services providers, (ii) Microsoft does not hold
market power in enterprise communications services, (iii) the market for
enterprise communications services is very competitive and Microsoft faces
competition from both specialised vendors and players offering a broad suite of
services, (iv) users typically multi-home and (v) the market is highly innovative.
351
Form CO, Section 6, paragraph 601.
352
Form CO, Annex 15, document 1, "Synergy – UC Collaboration Market Tracker –Worldwide", 2015.
Market shares in 2014 were: Cisco ([10-20]%), Microsoft ([10-20]%), Avaya ([5-10]%), IBM ([0-
5]%).
353
Notifying Party's response to RFI n.25, question 2.b and 2.c.
354
Notifying Party's response to RFI n.25, question 2.c.
355
This includes application and screen sharing, including markup and annotation, instant messaging and
presence, live streaming video, polls and surveys, whiteboard, and video and audioconferencing
(Notifying Party response to RFI n.25, question 2.c).
87
(b) Commission's assessment
(i) Ability and incentive to foreclose
(399) The Commission considers that, while it is unclear whether Microsoft will have
the incentive to foreclose competing providers of enterprise communications
services, it will not have the ability to do so.
(400) First, social networking features, like those of LinkedIn, do not appear to be an
important input or "must-have" for the choice of enterprise communications
services providers. Rather, important factors for customers of enterprise
communications services are redundancy and robustness, security, reliability,
ancillary functionality, management and support requirements.
356
(401) Second, Microsoft faces strong competition in the market for enterprise
communications by a number of players, like Avaya, Adtran, Mitel, Audiocodes
and with Cisco being the market leader, as mentioned in paragraph (396).
357
These competing providers of enterprise communications services could also
engage in partnerships with other providers of social networking functionalities
in order to offer comparable features.
(402) Third, with the exception of one respondent, which however did not clearly
articulate the reasons underlying its concerns, none of the respondents to the
market investigation consider that the Transaction will have a negative impact
on the market for enterprise communications services.
(ii) Overall likely effect on effective competition
(403) The market for enterprise communications services is today a competitive one
and is characterised by the presence of a number of strong players, including the
current market leader Cisco. The Commission considers it therefore unlikely
that the addition of LinkedIn to Microsoft's existing enterprise communications
solutions would lead to the anti-competitive foreclosure of its competitors.
(404) The arguments set out in Section 4.2.6.2 (i)(ii) above are equally valid for any
specific market segment within enterprise communications services, such as for
"converged conferencing".
(405) Consequently, the Commission concludes that the integration of LinkedIn
features into Microsoft's enterprise communications services, while denying
competitors similar levels of integration will not foreclose competing providers
of enterprise communications services.
(iii) Conclusion
(406) In light of the above, the Commission considers that the Transaction does not
raise serious doubts as to its compatibility with the internal market with regard
to the market for enterprise communications in the EEA.
5 COMMITMENTS
(407) In order to remove the serious doubts arising from the Transaction described in
Section 4 in relation to the foreclosure of competing PSN service providers, on
15 November 2016 the Notifying Party submitted commitments pursuant to
Article 6(2) of the Merger Regulation (the "Initial Commitments").
356
Case M.6281 – Microsoft/Skype, Commission decision of 7 October 2011, paragraphs 92-94.
357
Form CO, Section 6, paragraph 595 and following.
88
(408) The Commission launched a market test of the Initial Commitments on 17
November 2016, seeking responses from SN services providers and from
OEMs. The Commission informed the Notifying Party of the results of the
market test on 24 November 2016. Following the feedback received from
market participants in the market test, the Notifying Party submitted a revised
set of commitments on 29 November 2016 and a further revised version on 30
November 2016 (the "Final Commitments").
5.1 Description of the proposed commitments
5.1.1 Initial Commitments
(409) The Initial Commitments consisted of two sets of commitments. One set of
commitments was meant to address the Commission's concerns relating to the
possible integration of LinkedIn features into Office and to the possible denial
of access to Microsoft APIs (the "Integration Commitments"). Another set of
commitments was meant to address the Commission's concerns relating to the
possible pre-installation of a LinkedIn application on Windows PCs (the "Pre-
installation Commitments").
(410) The Initial Commitments were intended to apply throughout the EEA and to
remain in effect for five years from the closing of the Transaction.
(411) The Initial Commitments provided for a monitoring trustee (to be proposed by
Microsoft with the Commission's approval) and for a fast-track dispute
resolution procedure.
(412) The two sets of commitments constituting the Initial Commitments are
described in detail below. For both sets of commitments, the term "User" is
intended to refer to any user of a given Microsoft product, including
organisations and end users (whether consumers or employees within
organisations), irrespective of the distribution channel (acquisition from OEMs
or download from the Internet).
5.1.1.1 Integration Commitments
(413) The Integration Commitments consisted of the following elements.
(414) First, Microsoft committed to make available the Outlook APIs
358
and the
associated Outlook Add-In Program
359
to entities that are not controlled by or
related to Microsoft or LinkedIn and which design, operate, and offer PSN
Services
360
("Third-Party PSN Service Providers"). Microsoft also committed to
apply the standard terms and policies applicable to the Outlook APIs and the
Outlook Add-In Program (including access to all software development kits,
code samples, developer tools, and support resources) without discrimination
against Third-Party PSN Service Providers.
(415) Second, Microsoft committed to make available the Office Store for distribution
and downloading of Outlook add-ins for PSN Services and to apply the standard
358
Outlook APIs are application programming interfaces, including any associated call-back interfaces,
that third parties can use to build Outlook Add-ins (including those available through the Microsoft’s
Outlook Dev Center at https://dev.outlook.com/restapi/reference).
359
Including any updates and successors during the term of the Initial Commitments.
360
PSN Services are defined as applications, websites, or services that are primarily targeted to
professionals wherein users create public or semi-public profiles based on their real identity, connect
with professional contacts, and interact with these contacts on the basis of professional content,
excluding enterprise social networks.
89
terms and policies applicable to Office Store developers, without discrimination
against Third-Party PSN Service Providers.
(416) Third, Outlook add-ins would be able to run independently of any new
functionality included in Outlook mail and calendar services that involves
LinkedIn profile and activity information being displayed in Outlook (the
"LinkedIn Features for Outlook") and be accessible to Users in Outlook.
(417) Finally, Microsoft committed to grant Users in the EEA the ability to disable
and re-enable the LinkedIn Features for Outlook.
5.1.1.2 Pre-installation Commitments
(418) The Pre-Installation Commitments consisted of the following elements.
(419) First, in the event that Microsoft builds a LinkedIn branded application for
Windows PC OS (the "LinkedIn Windows PC Application") and includes it in
the Windows PC OS, Microsoft committed to allow Windows PC OEMs not to
install that application on their PCs that are distributed in the EEA. The same
commitment applied to the scenario in which Microsoft would build a LinkedIn
branded Start tile or Taskbar button and include it in the Windows PC OS (the
"LinkedIn Windows PC Tile").
(420) Second, as an ancillary obligation, Microsoft committed not to retaliate in any
way against any OEM for developing, using, distributing, promoting or
supporting a Windows PC application and/or a Windows PC tile for Third-Party
PSN Service Providers, in particular by altering Microsoft's commercial
relations with that OEM or withholding any monetary payment or other value
("Consideration")
361
provided to that OEM. As an additional ancillary
obligation, Microsoft committed not to enter into any agreement with an OEM
that links or conditions the grant of any Consideration on the OEM's refraining
from developing, using, distributing, promoting or supporting a Windows PC
application and/or a Windows PC tile for Third-Party PSN Service Providers.
The two ancillary obligations described in this paragraph are referred to as the
"Ancillary Obligations against Retaliation and Exclusivity".
362
(421) Finally, Microsoft committed to grant Users in the EEA the ability to remove
the LinkedIn Windows PC Application entry point and the LinkedIn Windows
PC Tile entry point from the Windows PC OS taskbar or the portion of the
"Start" menu that supports application or tile pinning.
5.1.2 Results of the market test
(422) The majority of the respondents to the market test expressed the view that the
Initial Commitments, taken as a whole, would be appropriate to effectively
eliminate the competition concerns in the market for PSN services in the EEA
and/or in individual EEA countries.
363
Those respondents also viewed the Initial
361
"Consideration" means the value provided by a party entering into contractual obligations and includes
any monetary payment, discount or the provision of preferential licensing terms; technical, marketing,
and sales support; enabling programs; product information; information about future plans; developer
support; hardware or software certification or approval; or permission to display trademarks, icons or
logos or any other preferential treatment.
362
Without prejudice to the Ancillary Obligation against Exclusivity, as well as to the application of EU
competition law, the Initial Commitments did not prevent Microsoft from providing Consideration to
an OEM as part of a non-exclusive agreement with any OEM to promote LinkedIn (including through
the pre-installation of a LinkedIn Windows PC Application and/or a LinkedIn Windows PC Tile).
363
Responses to market test questionnaire Q22 of 17 November 2016, question 31.
90
Commitments as appropriate to effectively remove any competition concerns
stemming from the Transaction.
364
(423) In addition, the majority of the respondents considered that the provisions of the
Initial Commitments would be sufficiently clear to be capable of being
effectively implemented and that the implementation of the Initial Commitments
could be effectively monitored.
365
(424) With respect to duration, the majority of the respondents considered five years
to be appropriate for both the Integration Commitments and the Pre-Installation
Commitments, while only a few respondents requested a longer period of
time.
366
(425) Finally, most respondents considered that the fast-track dispute resolution
procedure would be appropriate to ensure the proper implementation of the
Initial Commitments.
367
5.1.2.1 Integration Commitments
(426) According to the majority of the respondents to the market test, conceptually,
the Integration Commitments would be sufficient to preserve the ability of
Third-Party PSN Service Providers to compete by building add-ins for
Outlook.
368
(427) Moreover, most respondents considered the definition of PSN Services used for
the purposes of the Initial Commitments to be appropriate.
369
(428) However, as regards the scope of the API access, most respondents submitted
that, for the Integration Commitments to be effective, they should also ensure
the continued availability to Third-Party PSN Service Providers of add-in
programs and APIs for products of the Office suite other than Outlook, such as
Word, Excel and PowerPoint.
370
Moreover, a limited number of respondents
took the view that access to the "public" Office APIs would not be sufficient.
According to these respondents, Third-Party PSN Service Providers should be
afforded equal treatment with LinkedIn, meaning either that they should obtain
access to the "internal" APIs that Microsoft will use to achieve integration with
LinkedIn, or that, vice versa, LinkedIn should only be granted access to
Microsoft's "external" APIs accessible to Third-Party PSN Service Providers.
371
(429) Furthermore, as regards the ability for Users to disable the LinkedIn Features for
Outlook, some respondents considered that it would be sufficiently easy in
practice for Users to do so, while others expressed the concern that the disabling
functionality would not be sufficiently visible to Users.
372
A limited number of
respondents also argued that Users' ability to disable the LinkedIn Features for
Outlook would be insufficient and that such features should not be present by
default within Office products.
373
As regards the scope of the disabling
364
Responses to market test questionnaire Q22 of 17 November 2016, question 15 and 29.
365
Responses to market test questionnaire Q22 of 17 November 2016, questions 13, 14, 27 and 28.
366
Responses to market test questionnaire Q22 of 17 November 2016, questions 12 and 26.
367
Responses to market test questionnaire Q22 of 17 November 2016, question 30.
368
Responses to market test questionnaire Q22 of 17 November 2016, questions 4 and 6.
369
Responses to market test questionnaire Q22 of 17 November 2016, questions 3 and 8.
370
Responses to market test questionnaire Q22 of 17 November 2016, questions 5 and 7.
371
Responses to market test questionnaire Q22 of 17 November 2016, question 4.
372
Responses to market test questionnaire Q22 of 17 November 2016, question 9.
373
Responses to market test questionnaire Q22 of 17 November 2016, question 9.1.
91
functionality, most respondents took the view that Users should also be allowed
to disable LinkedIn features that Microsoft may implement with respect to
Office products other than Outlook.
374
In addition, some respondents indicated
that OEMs should also be granted the ability to enable and disable the LinkedIn
Features for Outlook in those cases in which OEMs pre-install Outlook as part
of Office on their PCs.
375
(430) Finally, a number of respondents raised the concern that access to data generated
by users of Microsoft products (e.g. Outlook contacts) would provide the
merged entity with a significant competitive advantage compared to other PSN
service providers, as it could help LinkedIn make recommendations for new
connections with a view to expanding the membership of its PSN.
376
5.1.2.2 Pre-Installation Commitments
(431) As regards the commitment allowing OEMs not to install a LinkedIn Windows
PC Application or a LinkedIn Windows PC Tile, all major OEMs who replied to
the market test confirmed that they are currently neither free, nor technically
able, to install a specific application or tile that is included in the Windows PC
OS image, unless specifically authorised by Microsoft.
377
Most OEM
respondents did not raise any particular concerns regarding the Pre-Installation
Commitments. In particular, the OEMs did not raise any concerns as regards the
effectiveness of the Ancillary Obligations against Retaliation and Exclusivity.
378
However, a number of OEM respondents took the view that OEMs should also
be allowed not to install the LinkedIn Windows PC Application and the
LinkedIn Windows PC Tile on Windows PCs in the event that, post-
Transaction, Microsoft would include such an application or tile as part of the
Office suite of products for pre-installation by OEMs on Windows PCs.
379
(432) Respondents to the market test expressed mixed views as to whether, under the
terms of the Pre-Installation Commitments, it would be sufficiently easy for
Users to remove the LinkedIn Windows PC Application or the LinkedIn
Windows PC Tile entry points from the Windows PC OS taskbar or the relevant
portion of the "Start" menu.
380
According to several respondents, actual ease of
removal would be dependent on how Microsoft constructs future versions of its
Windows OS.
381
A number of respondents submitted that simply allowing Users
to remove a shortcut to the LinkedIn Windows PC Application (as opposed to a
comprehensive uninstallation) would not be sufficient, as the LinkedIn
Windows PC Application may continue running in the background.
382
In
addition, a limited number of respondents argued that the LinkedIn Windows
PC Application should not be pre-installed at all, as Users should be prompted
374
Responses to market test questionnaire Q22 of 17 November 2016, question 10.
375
Responses to market test questionnaire Q22 of 17 November 2016, question 11.
376
Responses to market test questionnaire Q22 of 17 November 2016, question 31.1.
377
Responses to market test questionnaire Q22 of 17 November 2016, questions 16-19. Likewise, those
respondents explained that it would be technically or contractually impossible for them not to install
the LinkedIn Windows PC Application or LinkedIn Windows PC Tile, once Microsoft has included
such application or tile in the Windows PC OS, unless Microsoft would allow them to do so and
provide the technical means.
378
Responses to market test questionnaire Q22 of 17 November 2016, questions 24 and 25.
379
Responses to market test questionnaire Q22 of 17 November 2016, question 21.
380
Responses to market test questionnaire Q22 of 17 November 2016, questions 22 and 23.
381
Responses to market test questionnaire Q22 of 17 November 2016, questions 22 and 23.
382
Responses to market test questionnaire Q22 of 17 November 2016, question 22.1.
92
to exercise a positive choice as to which PSN application(s) they want to install
on their Windows PCs.
383
(433) Finally, a number of respondents to the market test took the view that the Pre-
Installation Commitments should also apply to mobile devices running the
Windows OS.
384
5.1.3 Final Commitments
(434) The Final Commitments consist of a revised version of the Integration
Commitments (the "Revised Integration Commitments") and of the Pre-
Installation Commitments (the "Revised Pre-installation Commitments").
(435) The geographic and temporal scope of the Final Commitments is the same as
that of the Initial Commitments (see paragraph (410) above). The Initial
Commitments have been modified to provide that, during the term of the Final
Commitments, the Commission may request all information from the Parties
that is necessary to monitor the effective implementation of the Final
Commitments.
(436) The modifications included in the Revised Integration Commitments and in the
Revised Pre-installation Commitments are described in detail below.
5.1.3.1 Revised Integration Commitments
(437) The Revised Integration Commitments entail the following modifications to the
Integration Commitments:
(a) Access to Microsoft's APIs is no longer limited to Outlook APIs (and the
Outlook Add-In Program), but has been extended to cover all Office APIs,
namely APIs for all core Office products, including Outlook, Word,
PowerPoint and Excel (and the overall Office Add-In Program);
385
(b) The commitment ensuring access to Office APIs has been complemented
by a comparable commitment to ensure the availability to Third-Party
PSN Service Providers of the Microsoft Graph, a unified gateway that
enables developers to build applications and services that can, subject to
User consent, access data (such as contact information, calendar
information, email and files) from Microsoft's cloud services;
386
(c) The commitment enabling Users to disable LinkedIn features is no longer
limited to the LinkedIn Features for Outlook, but has been extended to
cover LinkedIn features for the entire Office suite of products, including
Outlook, Word, PowerPoint, and Excel
387
("LinkedIn Features for
Office");
383
Responses to market test questionnaire Q22 of 17 November 2016, question 22.1.
384
Responses to market test questionnaire Q22 of 17 November 2016, question 31.1.
385
In particular, under the Final Commitments, Office includes Microsoft’s Outlook 2013, Word 2013,
PowerPoint 2013 and Excel 2013 desktop clients for Windows PC OS licensed as part of an Office
Suite or via an Office 365 subscription, including any updates or successors thereto made available or
released before or during the term of the Final Commitments.
386
As described at https://graph.microsoft.io/en-us/. This includes updates to the Microsoft Graph and
successors thereto. In particular, Microsoft commits to make available the Microsoft Graph to Third-
Party PSN Service Providers and to apply the standard Microsoft Graph terms and policies applicable
to the Microsoft Graph without discrimination against Third-Party PSN Service Providers.
387
See footnote 385 above.
93
(d) The ability to disable LinkedIn Features for Office has been specified to
ensure that the mechanisms available to Users to disable such features are
as accessible and prominent as those typically used to disable (and re-
enable) other connected services to Office.
5.1.3.2 Revised Pre-Installation Commitments
(438) The Revised Pre-installation Commitments entail the following modifications to
the Pre-installation Commitments:
(a) The scope of the Pre-installation Commitments (including the Ancillary
Obligations against Retaliation and Exclusivity) is no longer limited to
OEMs, but has been extended to cover any distribution partner who vis-à-
vis a third-party OEM has a right to determine which applications are
installed on the PCs manufactured by that OEM that run the Windows PC
OS (together with OEMs, "Microsoft PC Distribution Partners");
388
(b) The Users' ability to remove the LinkedIn Windows PC Application or the
LinkedIn Windows PC Tile is extended to cover the entire removal of that
application or tile from the Users' Windows PC OS;
(c) Microsoft further commits not to offer the LinkedIn Windows PC
Application or the LinkedIn Windows PC Tile to Users through Windows,
prompt Users to install such application or tile, or include such application
or tile in updates of Windows PC OS. This is without prejudice to
Microsoft's ability to promote the LinkedIn Windows PC Application, for
instance through the Windows Store or on its own or third-party websites.
5.2 Commission's assessment
5.2.1 Principles
(439) Where a concentration raises serious doubts as to its compatibility with the
internal market, the parties may undertake to modify the concentration so as to
remove the grounds for the serious doubts identified by the Commission and
thereby gain clearance of their merger in Phase I.
389
(440) It is for the parties to the concentration to put forward commitments.
390
The
Commission only has power to accept commitments that are deemed capable of
rendering the concentration compatible with the internal market.
391
In Phase I,
commitments can only be accepted where the competition problem is readily
identifiable and can easily be remedied. The competition problem therefore
needs to be so straightforward and the remedies so clear-cut that it is not
necessary to enter into an in-depth investigation and that the commitments are
sufficient to clearly rule out serious doubts within the meaning of Article 6(1)(c)
of the Merger Regulation. Where the assessment confirms that the proposed
388
Moreover, the term "Consideration" has been extended to also explicitly include "products". The
Ancillary Obligations against Retaliation and Exclusivity have been extended to also explicitly refer to
the possibility of an OEM (or, where applicable, distribution partner) "installing" a Windows PC
application or tile of a Third-Party PSN Service Provider.
389
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.
390
Remedies Notice, paragraph 6.
391
Remedies Notice, paragraph 9.
94
commitments remove the grounds for serious doubts on this basis, the
Commission clears the merger in Phase I.
392
(441) As concerns the form of acceptable commitments, the Merger Regulation leaves
discretion to the Commission as long as the commitments meet the requisite
standard.
393
(442) While divestiture commitments are generally the best way to eliminate
competition concerns resulting from horizontal overlaps, other structural
commitments, such as access remedies, or other non-divestiture remedies may
be suitable to resolve concerns if they are equivalent to divestitures in their
effects.
394
(443) In the ultimate assessment of proposed commitments, the Commission considers
all relevant factors including inter alia the type, scale and scope of the proposed
commitments, judged by reference to the structure and particular characteristics
of the market concerned, including the position of the parties and other
participants on the market.
395
The commitments must be capable of being
implemented effectively within a short period of time.
396
(444) It is against this background that the Commission analysed the Final
Commitments in this case.
5.2.2 Assessment of Final Commitments
(445) The Commission considers that the Final Commitments are sufficient to remove
the serious doubts regarding the compatibility of the Transaction with the
internal market in relation to the markets for PSN services within the EEA.
5.2.2.1 Revised Integration Commitments
(446) The Commission considers that the Revised Integration Commitments address
the serious doubts as to the compatibility of the Transaction with the internal
market in relation to the possible foreclosure of competing PSN service
providers stemming from the integration of LinkedIn features into Office and
the denial of access to Microsoft APIs by competing PSN service providers after
the Transaction.
(447) First, the Revised Integration Commitments ensure that Third-Party PSN
Service Providers will continue to have access to the Office Add-in Program and
the associated Office APIs based on the standard terms and policies (including
as regards access to related resources)
397
and without discrimination. After this
commitment has been extended in scope to reflect the concerns expressed in the
market test, Third-Party PSN Service Providers will be afforded the same type
of access also in relation to other Office products for which add-in programs
exist, namely Word, PowerPoint and Excel.
398
Thanks to the continued access to
392
Remedies Notice, paragraph 81.
393
Case T-177/04 easyJet v Commission [2006] ECR II-1913, paragraph 197.
394
Remedies Notice, paragraph 19.
395
Remedies Notice, paragraph 12.
396
Remedies Notice, paragraph 9.
397
This includes access to all software development kits, code samples, developer tools, and support
resources.
398
While add-in programs currently also exist for OneNote, Access and Project, as explained in footnote
13 above, these products are either not present in all or most versions of Office or, in any event, they
account for a very limited usage of the Office suite compared to Word, Excel, PowerPoint
and Outlook. Moreover, according to information submitted by the Notifying Party, Microsoft has no
95
Office Add-in Program and the associated Office APIs, Third-Party PSN
Service Providers will be able to build add-ins for the different Office products
that offer similar functionalities as those that Microsoft is envisaging to
introduce in relation to LinkedIn.
(448) While Office add-ins will not necessarily enable Third-Party PSN Service
Providers to obtain "equal treatment", i.e. to fully replicate the integration
scenarios contemplated by Microsoft (as those presuppose deep end-to-end
integration) and to be potentially present as default within Office, the
Commission considers that, under the Revised Integration Commitments and in
the specific context of this case, those providers will be afforded sufficient scope
and means to effectively compete with LinkedIn. The information provided by
the Notifying Party suggests that many of Microsoft's use case scenarios (e.g.
involving the display of LinkedIn contact information in Outlook or a Calendar
syncronisation feature),
399
can be replicated in comparable forms through
Outlook add-ins.
400
Furthermore, the Revised Integration Commitments ensure
that the Office add-ins built by Third-Party PSN Service Providers will be
available to Users in the Office Store and will be able to run independently of
any LinkedIn features integrated into Office.
(449) Second, consistent with the input received from the market test, Microsoft will
not be able to block Third-Party PSN Service Providers from gaining access to
the Microsoft Graph in the event that Users wish to give consent to such Third-
Party PSN Service Providers accessing their data. Accordingly, Third-Party PSN
Service Providers will be able, subject to User consent, to access data from
Office 365 or other Microsoft cloud services and use them to suggest new User
connections and potentially increase their membership, along similar ways as
the merged entity will be able to do post-Transaction.
401
(450) Third, the Revised Integration Commitments ensure Users' ability to disable the
LinkedIn Features for Outlook and, following the modifications to the
commitments introduced by Microsoft, for all other major Office products. This
option will be made available to any User of Office, including not only
consumers, but also organisations (e.g. corporate IT departments) and
employees of those organisations. Also, to reflect the concerns raised in the
market test that the mechanisms to disable the LinkedIn Features would be
hidden or not sufficiently visible to Users, the Revised Integration Commitments
explicitly require that such mechanisms be made adequately accessible and
prominent.
current plans to integrate LinkedIn features with any of those products (Notifying Party's submission
of 2 December 2016 at 11:58). As regards other Office-branded products such as Sharepoint, as
explained in footnote 280 above, Microsoft's presence in this area is much more limited compared to
Word, Excel, PowerPoint and Outlook and, moreover, Sharepoint does not currently support any add-
in programs (Notifying Party's submissions of 26 November at 22:46 and of 2 December 2016 at
11:58).
399
See Form CO, Annex 27 and Parties' response to RFI n. 19, question 1.c.
400
Moreover, add-ins may even have certain advantages over deep end-to-end integration when the
products and services concerned belong to distinct companies. According to information provided by
the Notifying Party (Form CO, paragraph 696), add-ins enable PSN providers to remain in control of
their data, which they would otherwise have to relinquish to achieve deep end-to-end integration with
Office.
401
This is without prejudice to applicable data protection rules.
96
(451) With respect to the comments expressed by some respondents to the market test
as to the need to allow OEMs to disable the LinkedIn Features for Office, the
Commission notes that a commitment to this effect would likely not have any
material impact on the visibility of LinkedIn to Users. Indeed, the OEM
distribution channel currently accounts for a very limited share of supply of
Office products.
402
5.2.2.2 Revised Pre-Installation Commitments
(452) The Commission considers that the Revised Pre-Installation Commitments
address the serious doubts as to the compatibility of the Transaction with the
internal market in relation to the possible foreclosure of competing PSN service
providers stemming from the pre-installation of a LinkedIn application on
Windows PCs after the Transaction.
(453) First, the Revised Pre-Installation Commitments ensure that OEMs will be free
to decide not to install the LinkedIn Windows PC Application or the LinkedIn
Windows PC Tile on the Windows PCs that they distribute in the EEA. In
particular, following the modifications introduced to reflect the feedback from
the market test, this freedom will apply not only to OEMs but also, where
relevant, to distribution partners who may be entitled to decide which
applications OEMs should install on their Windows PCs. Furthermore, the
Ancillary Obligations against Exclusivity and Retaliation will preserve the
effective freedom of any OEM (or, where applicable, distribution partner) to
enter into arrangements with Third-Party PSN Service Providers to pre-install or
otherwise support those providers' Windows applications without undue
pressure from Microsoft.
(454) Second, the Revised Pre-Installation Commitments ensure that, in the event that
OEMs (or, where applicable, distribution partners) would actually decide to
install the LinkedIn Windows PC Application or the LinkedIn Windows PC Tile
on their Windows PCs, Users would be able to remove such application or tile.
This option will be made available to any User of Windows PCs, including not
only consumers, but also organisations (e.g. corporate IT departments) and
employees of those organisations. Moreover, following the modifications
introduced in response to the market test, Users will not only be able to remove
the LinkedIn application or tile entry point from the prominent areas of their
Windows PCs (i.e. the Windows taskbar or the relevant portion of the “Start”
menu that supports application or tile pinning). Instead, Users will be able to
remove the LinkedIn Windows PC Application or the LinkedIn Windows PC
Tile entirely from their Windows PCs if they so wish.
(455) Third, as part of the improvements introduced after the market test, the Revised
Pre-Installation Commitments seek to prevent Microsoft from circumventing the
previous provisions by "pushing" the LinkedIn Windows PC Application or the
LinkedIn Windows PC Tile to Users after they have purchased their Windows
PC. In particular, the Revised Pre-Installation Commitments address a scenario
where no LinkedIn application or tile is installed on a User's PC (either because
Microsoft has not offered OEMs to pre-install it, or because OEMs have decided
not to install it, or because the User has decided to remove it). In such a
402
According to data provided by the Notifying Party, the Office products sold through OEMs accounted
for only […] of revenue and […] of units of total sales of Office products (Notifying Party's response
to RFI n. 21, question 9).
97
scenario, Microsoft will not be allowed through the Windows OS to offer Users
the LinkedIn Windows PC Application or the LinkedIn Windows PC Tile or to
prompt them to install them. Similarly, in such a scenario, Microsoft will be
prevented from including the LinkedIn Windows PC Application or the
LinkedIn Windows PC Tile (or any prompts or offers to Users to install them) as
part of any updates or upgrades to the Windows OS that it may release. At the
same time, the Revised Pre-Installation Commitments preserve Microsoft's
freedom to offer and promote the LinkedIn Windows PC Application through
the Windows store standard promotional mechanisms (which are also available
to third-party applications) or through other means.
(456) In sum, the Revised Pre-Installation Commitments ensure that effective choice
whether or not to have the LinkedIn Windows PC Application or the LinkedIn
Windows PC Tile installed is preserved both at the OEM and at the Windows
User level. In particular, by allowing OEMs (or, where applicable, distribution
partners) not to install the LinkedIn Windows PC Application or the LinkedIn
Windows PC Tile, the Revised Pre-Installation Commitments will prevent such
application or tile from being automatically present on all the Windows PCs to
be distributed in the course the next five years, and, as a result, from gaining
increased exposure to a large number of potential customers in the EEA to a
degree that would not be afforded to competing PSN applications or tiles. At the
same time, by ensuring that OEMs are able to enter into cooperation agreements
with Third-Party PSN Service Providers, the Revised Pre-Installation
Commitments will allow competing PSN applications or tiles to potentially gain
visibility to the large number of new Windows EEA Users within the next five
years and to compete with LinkedIn on a substantially equal basis in that regard.
(457) In relation to the criticism expressed by some respondents whereby the LinkedIn
Windows PC Application should be prevented from being pre-installed at all on
Windows PCs, the Commission notes that such a requirement appears to be
disproportionate in the context of these merger proceedings as it would unduly
impinge on Microsoft's ability to market its products going forward.
(458) The Commission considers that the point raised by certain respondents to the
market test that the Pre-Installation Commitments should extend to the
Windows mobile OS would go beyond the concerns identified by the
Commission in this case. Indeed, as explained in footnotes 270 and 272
above,
403
Microsoft's presence in the mobile OS space is currently very limited
and the Commission's investigation has not revealed foreclosure concerns
stemming from possible pre-installation of a LinkedIn application on Windows
mobile devices.
(459) Furthermore, as regards the comments expressed by some respondents to the
market test as to the need to allow OEMs not to install the LinkedIn Windows
PC Application or the LinkedIn Windows PC Tile in the event that Microsoft
would include such an application or tile as part of the Microsoft Office suite of
products, the Commission notes that a commitment to this effect would likely
not have any material impact on the visibility of LinkedIn to Users. Indeed, as
403
According to the data presented in footnotes 270 and 272 above, mobile phones running Windows OS
(Windows Phone) accounted for, respectively, less than [5-10]% and less than [10-20]% of mobile
phones shipped worldwide and in selected European countries in the period between Q4 2014 and Q3
2015.
98
explained in paragraph (451) above, the OEM distribution channel currently
accounts for a very limited share of supply of Office products.
5.2.2.3 Overall assessment of the Final Commitments
(460) As explained in the previous Sections, in light of the overall positive feedback
from the market test and of the improvements subsequently included in the Final
Commitments, the Commission considers that the Final Commitments taken as a
whole are capable of addressing the serious doubts as to the compatibility of the
Transaction with the internal market in relation to the possible foreclosure of
competing PSN service providers.
(461) Regarding the temporal scope of the Commitments, the Commission notes, first,
that the market test only revealed limited concerns that a five-year duration
would be insufficient. Second, the information available indicates that user sign-
ups and activity on PSN services is expected to gradually shift from PCs to
smart mobile devices in the next years,
404
while Microsoft's position in mobile
OSs is relatively limited.
405
The Commission therefore considers that a period of
time of five years is sufficient and appropriate in the present case.
(462) In addition, the Commission considers on the basis of the feedback from the
market test that the monitoring trustee and the arbitration procedure provide for
appropriate compliance and enforcement mechanisms in the present case.
(463) In the context of the assessment of the Revised Commitments, the Commission
also notes that, on 24 November 2016, Microsoft entered into a cooperation
agreement with XING. This agreement provides for enhanced commercial and
technical cooperation between the two companies, including as regards the
XING application for Windows, XING add-ins for Office and XING's use of the
Microsoft Graph APIs. This partnership further corroborates the Commission's
view that, in light of the Final Commitments, Microsoft will not have the ability
to foreclose competing providers of PSN services from competing against
LinkedIn in the EEA.
(464) Finally, and for the avoidance of doubt, the Commission notes that Microsoft's
conduct remains subject to the applicability of EU competition rules, and
notably of Articles 101 and 102 TFEU.
5.2.3 Conclusion
(465) In the light of the above, the Commission considers that the Final Commitments
are sufficient to eliminate the serious doubts as to the compatibility of the
Transaction with the internal market and with the EEA Agreement.
5.3 Conditions and obligations
(466) Under the first sentence of the second subparagraph of Article 6(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.
404
In particular, according to data provided by the Notifying Party, by 2020, […] of all page views on
LinkedIn worldwide will be on mobile devices. Moreover, while […] of new member sign-ups in the
EEA were made on mobile devices in 2013, this percentage now represents […] of the total new
member sign-ups. LinkedIn anticipates that this general trend will continue in the coming years. Form
CO, paragraphs 171 and 173.
405
See footnote 403 above.
99
(467) The achievement of the measure that gives rise to the 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 no longer stands. Where the undertakings concerned commit a
breach of an obligation, the Commission may revoke the clearance decision in
accordance with Article 6(3)(b) 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.
(468) In accordance with the basic distinction between conditions and obligations
described above, all requirements set out in the commitments are considered to
constitute obligations.
(469) The full text of the Final Commitments is annexed to this Decision as Annex 1
and forms an integral part thereof.
6 CONCLUSION
(470) For the above reasons, the Commission has decided not to oppose the notified
operation as modified by the Final Commitments and to declare it compatible
with the internal market and with the functioning of the EEA Agreement,
subject to full compliance with the obligations contained in the Final
Commitments. This Decision is adopted in application of Article 6(1)(b) in
conjunction with Article 6(2) of the Merger Regulation and Article 57 of the
EEA Agreement.
For the Commission
(signed)
Margrethe VESTAGER
Member of the Commission
MICROSOFT / LINKEDIN
Case No. COMP M.8124
13
Signature: _____________________________
Name: …….
Position: …….
Company: …….
Duly authorized and acting for and on behalf of
Microsoft
Date:…….
