Online platforms cater their services to at least two separate customer groups by facilitating an interaction. Accordingly, when assessing the market power of an online platform, it is essential to establish whether those customer groups are part of a single relevant market or multiple relevant markets.
The purpose of this article, available here, is to provide practical guidance on the market definition process for online platforms in light of their distinctive characteristics. It does so through three sections:
The first section shortly discusses the importance of the market definition for the application of Article 102 TFEU in practice.
Findings of dominance rely on an adequate measurement of economic market power. Whether economic market power amounts to legal dominance is a determination that depends on a properly defined relevant market. Market definition is also required for evaluating any possible efficiency arguments that would justify the prima facie anticompetitive practices of the concerned undertaking.
The second section develops an approach to market definition for online platforms based on the interactions these customers facilitate between participating customer groups.
Market definition in the context of online platforms requires taking into account their two- or multi-sided nature. The literature is full of suggestions to the effect that whether one or more markets will have to be defined depends on the nature of a platform or its business model. Online platforms, however, do not have a predetermined number of types or business models. Instead, a framework applicable to all platforms must focus on something that all platforms have in common: i.e. the nature of the interaction between the sides of the platform under competition law scrutiny.
The author suggests that two or more customer groups should be found to be part of the same relevant market only when the intermediary matching service by the platform is indispensable for meeting the demands of each separate customer group. Cases concerning bi- or multilateral matching functionalities that do not meet this indispensability requirement would entail looking at each customer group of the platform as separate, albeit related, markets.
The third and final section looks at key considerations in the assessment of substitutability between online platforms and their closest online and offline competitors.
The number of relevant markets that need to be defined in each case cannot be solely determined based on the type of matching interaction facilitated by the platform. It also entails determining whether the demand of each customer group is met by other (non-platform) undertakings.
As regards business-to-consumer platforms, defining the relevant market with regard to merchants will usually require assessing whether they will be willing to switch to a competitor (i.e. other (possibly non-platform) sales or distribution channels) in the event of a participation fee raise by the online platform. From a consumer perspective, the assessment of demand substitution for online platforms when used as a sales channel will depend on numerous variables that will influence the substitutability assessment in each case. These variables include the implications of using different sales channels for the products or services in question, as well as the scope of the products or services offered by online platforms.
This is another interesting paper on market definition and platforms (a topic that was already addressed by the OECD in a piece that was justly nominated from an award).
While I understand the paper’s criticism of typological approaches, I always had the impression that such typologies were merely meant to be used as first approximations, and that market definitions should reflect the strength and direction of indirect network effects (see, for example, here and here). At the same time, I think that the paper develops its own typology – one based on unilateral or multisided interactions, which seems to hark back to the distinction between transaction and non-transaction platforms with a twist. As such, I understand the paper to argue that this typology is more accurate than other typologies, more than to provide guidance on how to conduct market definitions in multisided markets – a valuable and challenging endeavour nonetheless.
The most interesting contention of the paper – but one which is not really addressed explicitly – is that market definition should be pursued by reference to the conduct which is suspected to be anticompetitive. This was a point which was already made in what is, for me, still the best paper on market definition for platforms, and the one with the best advice on how to pursue market definitions for online platforms – the contribution to the OECD work on the topic which I reviewed here.