Unlike recent merger cases where the Commission looked at the concentration of data within a merged entity, the Commission’s focus in Apple/Shazam was on vertical concerns, including as a result of access to Shazam customer data. The focus was more about the potential impact on Apple’s rivals rather than whether the acquisition of data would raise barriers to entry. In the end the Commission had no issues because:
- access to Shazam’s data would not materially increase Apple’s ability to target listeners and any conduct aimed at making customers switch would only have a negligible impact. Competition was for new customers not switching.
- the merged entity would not be able to shut out competing providers of digital music streaming services by restricting access to the Shazam app given its limited importance as an entry point to the music streaming services of Apple Music’s competitors
- the integration of Shazam’s and Apple’s datasets on user data would not confer a unique advantage to the merged entity
Because Apple/Shazam provides by far the most detailed assessment of big data competition concerns so far, here are five observation on the data related aspects:
1) A (new) market for licensing music data but no separate market for the data itself: one of horizontal overlaps was a market for the licensing of the data that each of Apple and Shazam use to create “music charts” – charts which show which tracks are trending in a geographic area or the most purchased etc. There have been some suggestions, at least from academics, that that data itself should be a relevant market but the Commission seems to look beyond internally generated/used data to data that is actually commercialised (in this case licensed) . That makes sense as what is a market without demand and supply. Of course the EU essential facility doctrine may not require a “market” for input as such but that’s a different issue.
2) Relevance of regulatory and contractual arrangements for foreclosure analysis: the Commission considered whether Apple would get access to commercially sensitive data of competing providers of music streaming services and encourage users to switch to Apple Music. That data was treated as business secrets (even though not forward looking price etc. – para 216). When looking at the ability to foreclose, the Commission paid attention to legal and contractual limitations on the use of the information by Apple post-transaction, such as (i) the application of the GDPR and the e-Privacy Directive, (iii) Shazam’s terms of service and privacy notice, and (iii) the potential for changes in Android’s Developer Guidelines and/or apps’ Application Programming Interfaces that could make it more difficult for Shazam to access data about apps installed on users’ phones (paras. 221-238). Mind you, the Commission’s press release did provide a bit of a GDPR compliance warning: “a merger decision does not release companies from respecting all relevant data protection laws”.
3) Data privacy as a parameter of competition?: when investigating Apple’s incentives to foreclose, the Commission seemed to treat as relevant Apple’s intention to “change Shazam’s data collection practices to bring them in line with Apple’s industry leading-positions on privacy”. That would mean that Shazam would not actually send Apple certain user info unless the music streaming service of that user agreed to allow this information to be sent to Apple. While there was no suggestion of competition in a ‘market for privacy’, the reference – and perhaps attention given – to “industry leading position on privacy” reminded me of this debate.
4) Internal documents were important again when exploring any incentive to foreclose. (paras. 239-245).There were apparently 50 RFIs in phase 2 which resulted in the submission of over 100,000 internal documents.
5) Four Vs (variety, velocity, volume, value) – the parameters of big data : “big” is of course a vague concept so all the Big Data literature tends to describe four parameters (Four Vs) for assessing the commercial and therefore competitive importance of a data set: variety, velocity, volume, value. For the first time (I think) the Commission adopts that approach when looking the role/importance of a data set. In this case it was the importance of Shazam user data which Apple might use to improve its own streaming service and which it might cut off access to rivals. The Commission footnotes the Franco-German paper on big data (para 234). This seems an intuitive and workable framework which could be applied when calibrating a claimed “data advantage” even in market power cases:
a) Variety of data: Similar data was collected by rivals in this case so data was available from different sources. Plus, Shazam’s data was more limited : only having access to music discovery data, and not to music consumption data.
b) Velocity of data collection: Shazam collected data at a lower speed and users spend less time on Shazam than on music streaming apps.
c) Volume of data set: Providers of music streaming apps have access to much larger data sets than that of Shazam.
d) Economic value of data set: The Shazam data is not a key asset and is not unique. It is not a key element of success of digital music streaming apps. In the music industry, the most valuable data relates to music consumption, which is held by music streaming providers.
Conclusion: the unconditional phase 2 clearance shows how the Commission’s is progressing its analysis of data-related mergers. The 4Vs in particular seem to have earned a place in the antitrust vernacular and are bound to feature regularly in future data related assessments.