Commission, December 6, 2016, No M.8124

Dear Sir/Madam,

(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 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 devices5 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.

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 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.

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 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 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 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 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 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 ofcustomers.

(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, 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. 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.

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 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

(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 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 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 platforms76 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 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 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 PSNs87 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 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 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 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.

<Table>

(114) Respondents provided mixed views on the possible supply-side substitutability between an enterprise social network and a PSN, that is, whether a company offering the former could easily develop and offer the latter. While a number of respondents explained that the basic features of the two platforms are similar, most respondents explained that this would be difficult, as it would require significant investments, time and effort to develop scale and connection.103 The submission mentioned in paragraph (113) above explains that an exterprise social network cannot easily turn into a PSN, as this would require tearing down the “walled-gardens” of the enterprise customer base. According to the submission, an enterprise social network lacks all the key features of a PSN, which would need to be built.

(115) Based on the resultats of the market investigation, the Commission therefore considers that PSNs may constitute a separate product market within social networks, given that they present different functionalities, features, and usage cases. In particular, users appear to rely on a PSN for different uses, content, goals and connections as opposed to a personal social network. Additionally, it appears that PSNs have a broader scope than specialized “vertical” social networks. While the latter also target and are used by professionnals, they a 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 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.

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.

(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 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 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 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

(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 2012122 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 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 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.

 (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 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 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

<Figure>

(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.

<Figure>

(190) Within the market for CRM software solutions, the largest segment is currently CSS, and is expected to remain as such at least until 2020, despite a small decline (from 39% of the overall market in 2014 to 37% by 2020). CSS is followed by Sales (24% in 2014 and decreasing to 21% by 2020), Marketing (21% in 2014 and increasing to 27% by 2020) and DC (constant at 16%), as shown in the below Figure 3.

<Figure>

(191) Considering the further sub-segmentation between B2B and B2C CRM software solutions, in 2015 slightly below than 50% of the market for CRM software solutions consisted in the sales of B2B solutions, the remaining being sale of B2B solutions, and these proportions are expected to be constant in the future. In 2015 CSS accounted for the major part of the sales of B2B CRM software solution and corresponded to approximately 20% of the market for CRM software solutions (decreasing to 18% by 2020), followed by Marketing (15% of the market for CRM software solutions in 2015, increasing to 17% by 2020) and Sales (14% of the market for CRM software solutions in 2015, decreasing to 12% by 2020), as shown in the below Figure 4

<Figure>

(192) According to respondents to the market investigation, the market for CRM software solutions is competitive152 and the drivers to the adoption of a CRM software solution for customers, and therefore the parameter of competition in the market for CRM software solutions, are functionality, perceived “ease of use” of the system, price, provider reputation and support.153

(193) The market for CRM software solutions is also fragmented and characterized by a large number of different providers, as illustrated in Figure 5 below.

<Figure>

(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>

 (196) In 2010 Salesforce acquired Jigsaw, rebranded Data.com, an online business directory of companies and business professionals. In September 2016, 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 incorporatedML 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 customersresponding 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 itsoffering in EAS, including CRM software solutions, Oracle announced "Adaptive Intelligent Applications", providingML 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 B2B163profiles 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 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, 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>

(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.177Zoominfo 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,180as 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

(207) However, the adoption of sales intelligence solutions is accelerating. In particular, respondents to the market investifation noted that the importance of sales intelligence solutions for CRM customers will increase over the next two to three years.189 In this respect, IBM highlighted that sellers need insight and analytics to help guide them through the selling process,190 while Salesforce stated that “there has […] been a clear shift [toward increasing importance of sales intelligence solutions] with the increasing demand for artificial intelligence and machine learning functionalities.”191

4.2.2.2 Market shares

(208) According to the Non-Horizontal Guidelines, non-horizontal mergers do not give rise to competition concerns unless the merged entity has a significant degree of market power in at least one of the markets concerned.192 In this respect, market shared and concentration levels provide useful first indications of the market power and the competitive importance of both the merging parties and their competitors.193

(209) Moreover, pursuant to the Non-Horizontal Guidelines, the Commission will not extensively investigate the non-horizontal mergers where the market share post-merger of the merged entity in each of the markets concerned is below 30% and the post-merger concentration level is below 2000, as it is unlikely that competition concerns would be found, except where special circumstances arise, such as the imminent expansion of a company because of a recent innovation.

(a) CRM software solutions

(210) Table 2 below shows the market shares of Microsoft and its main competitors in the market for the provision of CRM software solutions and in potential narrower segments by functionality at worldwide EEA levels in 2015.

<Tableau>

(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 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

(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 […]%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.

(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 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).

(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 2016213 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, 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 solutions219 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 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. 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 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 solutions232 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, 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, 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 LinkedIn244 and the submissions of the Notifying Parties245 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 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

(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 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 APIs258 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 sources262 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.

(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 slightly above [30-40]% depending on the EEA country at issue and would be the second largest player after Facebook in most EEA countries.265

(284) The Commission notes that the market shares provided by the Notifying Party above are likely to significantly underestimate LinkedIn’s position in the PSN market. These market shares have been calculated on a total market that also includes Facebook and a large number of small players which the Notifying Party qualifies as “specialized vertical social networks” as they appear to focus on specific professions and niches.266

(285) In order to derive market share data for PSN services within EEA, the Commission has only taken into account those providers that are considered to be active in the market for PSN services based on the responses to the market investifation and on public information, thereby excluding the providers of SN services (such as Facebook) and of vertical (or industry-specific) PSN services.267 On this basis, the following competing providers of PSN services have been identified in individual countries within the EEA: Viadeo (active in France), Xing (active in Austria and Germany) and GoldenLine (active in Poland).

(286) Accordingly, as shown in Table 3 below, LinkedIn appears to be either substantially the sole provider of PSN services (in most countries within the EEA), or the largest or second largest provider in a market with only two competitors (in Austria, Germany, France and Poland). In a potential EEA-wide market, LinkedIn would be a distant leader.

<tableau>

(b) Oss

(287) The data provided by the Notifying Party indicate that Microsoft with its Windows branded Oss accounts for the large majority of shipments of Oss for desktops and laptops both at the worldwide and at the EEA level.

(288) At a worldwide level, as shown in Table 4 below, in the period from Q4 2014 to Q3 2015, Windows accounted for [80-90]% of shipments of desktops/laptops.270

<tableau>

(289) The Notifying Party has not provided market share data for Oss at the EEA level, but it has submitted aggregated market share data on shipments for France, Germany and the United Kingdom, which in the Notifying Party’s view do not materially differ from market shares on an EEA-wide market. As shown in Table 5 below, in the period from Q4 2014 to Q3 2015, shipments of Windows desktops/laptops accounted for [80-90]% of total aggregated shipments of desktops/laptops in Germany, France and the United Kingdom.272

<tableau>

(c) Productivity software

(290) The data submitted by the Notifying Party indicate that Microsoft with Office accounts for the large majority of revenues from productivity software for PCs both at the worldwide and at the EEA level.

(291) At a worldwide level, as shown in Table 6 below, Microsoft had a market share of above [90-100]% in terms of revenue in the overall market for productivity software for PCs in 2015, including in its consumer and enterprise segments.

<tableau>

(292) In Western Europe, as shown in Table 7 below, Microsoft has a market share of above [90-100]% in terms of revenue in the overall market for productivity software in 2015, as well as in its consumer and enterprise segments. In the Notifying Party’s view, market shares for the EEA would not materially differ from market shares for Western Europe.

<tableau>

(293) There are no indications that Microsoft’s market share would be materially different if the main individual products constituting Office276 (notably, Word, 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

(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 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 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 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 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 factors298 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 area299 take place via PCs running the Windows OS.300 As regards PSN user engagement, the data submitted by the Notifying Party suggest that, in Q2 2016, […] of LinkedIn micro-sessions301 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 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 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 engagement314 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. 

(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 "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 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 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 them327 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 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.

4.2.4 Foreclosure of competing online recruitment services providers

4.2.4.1 Market shares

(353) The Notifying Party has provided the Commission with market shared for online recruitment services based on desktop traffic shares (based on the total number of visits) as reported by SimilarWeb for the month of July 2016.333

(354) According to the Notifying Party’s estimates the market is highly fragmented with several market players including LinkedIn and many other local providers with a market share of below [5-10]% each in the respective EEA countries. The Notifying Party’s estimated market shares of the three largest providers for online recruitment in terms of number of web visits and of LinkedIn in the EEA and in six EEA countries are presented in the Table 7 below.

(355) The responses to the market investigation confirm that the market for online recruitment is highly fragmented with a high number of providers including LinkedIn, Monster, Indeed and many local players (like e.g. Xing in Germany) 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 Office336 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.

(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 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 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 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  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 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

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.

(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.353In 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.

(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").

(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 APIs358 and the associated Outlook Add-In Program359 to entities that are not controlled by or related to Microsoft or LinkedIn and which design, operate, and offer PSN Services360 ("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

(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 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 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 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 Excel387 ("LinkedIn Features for Office");

(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 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 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.

(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 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 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.

(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.

NOTES :

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.

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.

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.

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.

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.

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.  

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.

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.

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.

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 11.

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. 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.

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. 61 Responses to questionnaire to Sales Intelligence Solutions Competitors Q6 of 14 October 2016, 14 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. 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. 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.

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.

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.

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.

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.

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.

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.

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.

103 Responses to questionnaire to Social Networks Q3 of 17 October 2016, question 15.

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.

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.

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.

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. 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.

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. have adjusted their ad intermediation tools to enable advertisers to manage campaigns across different platforms, such as desktop and mobile. 122 Case M.6314 – Telefonica UK/Vodafone UK/Everything Everywhere/JV, Commission decision of 4 31 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.

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. 

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.

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.

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. 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.

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.

150 Notifying Party’s response to RFI n. 24, question 5.

151 Notifying Party’s response to RFI n. 24, question 5.

152 In the market investigation, CRM customers overwhelmingly stated that the CRM market is competitive or highly competitive: see responses to questionnaire to CRM Customers Q7 of 14 October 2016, question 20, questionnaire to additional CRM Customers Q18 of 28 October 2016, question 4. See Form CO, Section 6, paragraphs 479 and following. See also Salesforce’s submission “BRG White Paper on MSFT-LinkedIn”, paragraph 20.

153 See Form CO, Section 6, paragraph 496, and Salesforce, BRG White paper, paragraph 65.

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.html165 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.

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

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, 44 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 customer base continuing to be based in the United States. See Form CO, Annex 5, Document 46, spreadsheet “Standalone revenue detail”.

189 Responses to questionnaire to CRM Customers Q7 of 14 October 2016, question §22, questionnaire to CRM Competitors Q5 of 14 October 2016, question 30, questionnaire to additional CRM Customers Q18 of 28 October 2016, question 7, and questionnaire to Sales Intelligence Solutions Competitors Q6 of 14 October 2016, question 14.

190 IBM’s response to questionnaire to CRM Customers Q7 of 14 October 2016, question 22.

191 Salesforce’s response to questionnaire to CRM Competitors Q5 of 14 October 2016, question 30.

192 See Non-Horizontal Guidelines, paragraph 23

192 See Non-Horizontal Guidelines, paragraph 24.

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.

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. solutions and any foreclosure strategy would negatively impact Microsoft's overall revenue.

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.

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

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, 50 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.

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".

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.

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.

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.

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. 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.

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.

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.

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.

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).

265 In the two EEA countries LinkedIn would be the third largest player after Facebook and Github (in Austria), or after Facebook and XING (in Germany).

266 These include angel.co, houzz.com, houzz.co.uk, houzz.fr, houzz.it, houzz.de, 99designs.com, 99designs.de, 99designs.co.uk, 99 designs.it, 99designs.fr, avvo.com, Zillow.com, behance.net, styleseat.com, weddingwire.com, envato.com, envato.de, seekingalpha.com, wyzant.com, angielist.com, doximity.com, militaty.com, stackoverflow.com, stackerxchange.com, sermo.com, etoro.com, etoro.fr, stocktwits.com, researchgate.net, researchgate.com, academia.edu, spiceworks.com, github.com, praticefusion.com, rocketlawyer.com, rocketlawyer.co.uk, Edmodo.com, techrepublic.com, dxy.com, 99designs.es, seekingalpha.de, seekingalpha.co.uk.

267 See paragraph (108) above.

268 For the purpose of calculating the market shares, only the desktop traffic of LindekIn and the other companies identified as providers of PSN services in each given EEA country was taken into account (XING ins Austria) and Germany, Viadeo in France and GoldenLine in Poland). While the market shares in this table relate to the month of July 2016, the Notifying Party estimates the market shares for 2014 and 2015 would not be materially different from those provided in this table (Notifying Pary’s response to RFI n.16, question 2).

269 In EEA countries other than Austria, France, Germany and Poland, LinkedIn would have a market share of [90-100]% or close to [90-100]%

270 By contrast, according to the Notifying party, Windows accounted for only respectively [5-10]% and [0(-5]% of the worldwide shipments of tablets and mobile phones. Notifying Party’s response to the Commission’s questions of 29 August 2016, page 13, IDC date on shipments for Q4 2014 to Q3 2015.

271 While the market shares in this table relate to Q4 2014-Q3 2015, the Notifying Party estimates the market shares of 2013, 2014 and 2015 would not be materially different from those privded in this table (Notifying Party’s response to RFI, n. 26, question 1).

272 By contrast, according to the Notifying Pary, Windows accounted for only respectively [10-20]% and [5-10]% of aggregated shipments of tablets and mobile phones in those Member States. Notifying Party’s response to the Commission’s questions of 29 August 2016, page 14, IDC data on shipments for Q4 2014 to Q3 2015.

273 While the market shares in this table relate to Q4 2014-Q3 2015, the Notifying Party estimates that market shares for 2013, 2014 and 2015 would not be materially different from those provided in this table (Notifying Party’s response to RFI n. 326, question 1)

274 The market shares in Table 5, which are based on Gartner data, refer to “Office suites”, a term which, according to the Notifying Party’s, includes word processing, spreadsheet manipulation, presentation graphics, document creation, management and collaboration. (Notifying Party’s response to RFI n. 24, question 4). While the market shares in this table relate to 2015, the information submitted by the Notifying Party indicates that market shares for 2013 and 2014 would not be materially different rom those provided in this table.

275 Regarding the meaning of “Office Suites”, see footnote 274 above. While the market shares in this table relate to 2015, the information submitted by the Notifying Party indicates that market shares for 2013 and 2014 would not be materially different from those provided in this table.

276 As explained in paragraph (20) and footnote 13 above, for the purpose of this Decision, unless indicated otherwise, the term “Office” refers to the Microsoft productivity suites consisting of Word, Excel, Powerpoint and Oulook (including both on-premises and cloud versions).

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 64Supply"). 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).

285 Notifying Party's response to RFI n. 21, question 12.

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.

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.

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).

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.

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.

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.

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.

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.

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.

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.

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.

333 According to the Notifying party’s methodology for the calculation of the share data, traffic shared are based on the total traffic of all websites categorized under “Jobs and Employment” as well as LinkedIn, XING and Viadeo. Each website’s share is based on its total traffic comprising all visits to the website in July 2016. In contrast to “Jobs and Employment” sites, the majority of the traffic on LinkedIn estimates that only […]% of its European desktop traffic is directly attributable to recruitment activity. Therefore, LinkedIn has multiplied the total national (within the EEA) traffic of social networks LinkedIn, XING and Viadeo by […]% in order to estimate these websites’ recruitment traffic. The Notifying Party submitted that market shares based on desktop traffic of LinkedIn, XING, GlodenLine and Viadeo in the EEA, Germany, Austria, Poland, France and other EEA countries in 2014 and 2015 were not materially different from those indicated in Table 8 (Notifying Party’s response to RFI n. 26, question 2).

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.

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.

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.

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.

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.

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.

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).

356 Case M.6281 – Microsoft/Skype, Commission decision of 7 October 2011, paragraphs 92-94.

357 Form CO, Section 6, paragraph 595 and following

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. terms and policies applicable to Office Store developers, without discrimination against Third-Party PSN Service Providers.

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.

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.

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.

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.

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.

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 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.

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). 

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.

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.