Antitrust treatment of online advertising markets

As online advertising continues its rapid growth, online ad revenues become an ever more important opportunity (and/or threat) for many TMT businesses. How these growing markets are treated by antitrust agencies can be critical to strategy – either in terms of M&A or defending business models from the actions of other market participants. This article looks at some of the issues that online advertising markets raise for antitrust agencies and how these issues have been considered to date.

Online markets: issues for antitrust agencies

Many online services are offered free of charge. These services are either funded directly by advertising or in return for more diffuse benefits such as supporting growth of a platform or operating system. Since antitrust agencies traditionally attempt to analyse markets via effects on pricing; free services make it more difficult to define markets and assess market power. They make it even more difficult to assess effects of a deal or commercial practice on diffuse dynamic factors such as platform growth.

Second, speed of innovation/growth. Online markets have evolved enormously in recent years and their development continues at a rapid pace. This speed exacerbates the risk associated with the traditional ‘error types’ facing antitrust agencies. Type 1 error refers to intervention when the perceived problem may not exist; or the market left to its own devices may fix the perceived problem quickly. Type 2 error refers to intervention that comes not at all or too late and the market has irreparably changed (eg ‘tipped’ in favour of a particular network or platform).

Third, antitrust decisions are legal decisions – they require evidence and create precedent. Even if antitrust laws are flexible enough (and most courts/agencies think they are), cases themselves are necessarily backward-looking since the evidence used to support them reflects the past. A prediction of the future can play a part in assessing evidence – but a prediction isn’t enough to issue a fine or block a deal. Precedent is also an issue. Despite the fact that market assessments are not supposed to carry precedential value – in practice, what has happened before has a big effect on how the individuals involved treat similar issues in future.

Market definition

The familiar starting point for antitrust assessment is market definition. This exercise sets the frame of reference in which most antitrust issues are analysed.
The Commission has found a number of product market distinctions in online ad industries. It long ago distinguished between online and offline advertising. This approach was recently reaffirmed in Viacom/Channel 5 Broadcasting [2014] where the Commission stated that the specificity of online advertising continues to set it apart from less targeted forms of advertising used offline. In the offline world, separate markets tend to be found for each advertising medium (print, TV etc – see News Corp/BSkyB [2010]).

Within online advertising, there is a distinction between search-based and non-search-based display advertising. The justification for this is that search-based allows targeting of adverts on a real-time expression of interest (Google/DoubleClick [2008]). Within search, separate markets have been found for general internet searches and vertical internet searches (which focus on specific segments of online content, ie legal, medical etc – see Microsoft/Yahoo! [2010]). A product market has also been identified for intermediary services related to online advertising (Google/DoubleClick). A further distinction between mobile (search and non-search) and static (PC) online advertising has been considered but not yet formally recognised as separate markets (Facebook/WhatsApp [2014]).

Despite the Commission’s continued separation between online and offline advertising, there is a recognition that convergence is happening. For example, the German antitrust authority has recognised a competitively significant convergence between in-video-streaming online advertising with TV advertising (GU ProSiebenSat.1 Media AG/RTL interactive Gmbh [2011]).

The current subdivision of advertising into number of antitrust markets has a number of consequences. It makes mergers and agreements between players in traditional advertising markets (TV, print, radio etc) more difficult – despite the fact that ad revenues in all these media have been disappearing online for years now. Within online advertising, dividing into narrow product markets may also have hindered agencies’ ability to see the effects of a transaction on the overall online advertising system.

Network effects

Beyond market definition, network effects have been a common feature of antitrust assessment in ad-funded online markets. Network effects occur when the value to the user of the product increases based on the number of other users of the product. The Commission has found network effects to exist in social media platforms and mobile operating systems (Facebook/WhatsApp, para 115; Microsoft/Skype, para 91); but has not found strong network effects in search engines (Google/DoubleClick).

The existence of network effects affects antitrust agencies’ perception of the risks of intervention. In markets with strong network effects, there is often an argument that antitrust intervention must come early – before the network effect causes markets to ‘tip’ in favour of one company/platform etc. However, even here, the fast pace of change and risk of disruption has been used as a successful defence (Facebook/WhatsApp, paras 130-132). In Microsoft/Skype [2011] the Commission found usage shares as high as 80-90%, even with a recognisable network effect, these were not indicative of market power because the voice over IP (VoIP) services market was growing so rapidly and there was space for further innovation and disruption.

All roads lead to Google

Ultimately, many of today’s antitrust issues in online advertising lead to Google and its role at the heart of the complex online advertising and online ad tech industry.

The Commission is currently investigating Google for allegedly abusing its dominant position in the market for general internet search services by favouring its own comparison shopping products in its general search results pages (Google Search). The Commission needs to find that Google holds a dominant position in the market for general internet search services. Google’s large market share would traditionally make this straightforward but search services are offered free of charge to those users and users are free to switch to an alternative search provider (Bing, Yahoo! etc) with no cost and with minimal effort. Google argues that it invests heavily in the technology that sits behind its search engine. This functionality is what drives the high number of users, if that declined consumers would quickly and easily switch away.

Google’s dominance is sometimes said to derive from the fact that Google is able to collect and store a large amount of data on its users. This enables Google to offer highly targeted advertising capabilities to its advertising customers leading to a higher advertising pricing/revenues – and is something that it is difficult or impossible for competitors to match.

Once dominance is established the Commission must show abusive practice. On this, a recent UK High Court decision seems to support Google. Preferential treatment of Google Maps at the expense of a competitor, Street Maps, was deemed to help competition rather than hinder it and did not constitute abuse ( Ltd v Google Inc & ors [2016]). Similarly, a German court came to the conclusion that the display of Google’s ‘Weather OneBox’ at the top of weather forecast searches in preference to other results was a legitimate practice encouraging innovation.

Watch this (online) space

The outcome of the Google investigation will have a substantial effect on both Google and how antitrust agencies consider online markets in future cases. Whichever way the Google cases go, however, the antitrust assessment of these markets will continue to develop and change – if always lagging the changes in the underlying industry.