General Court confirms self-preferencing abuse of Google Shopping

On 10 November 2021, the General Court (GC) largely confirmed the European Commission’s (EC) Decision of 27 June 2017, which imposed a fine of €2.42bn on Google for giving favourable treatment to its comparison shopping service in its search results. According to the EC, this practice had resulted in increased traffic to Google’s comparison shopping service, to the detriment of competing comparison shopping services that would have otherwise benefited from this traffic. The EC concluded that Google had therefore abused its dominant position by (i) leveraging its dominant position on the general search markets to the comparison shopping services markets; and (ii) protecting its dominant position on the general search markets.

On appeal, Google argued that the alleged abuse concerned merely the implementation of quality improvements and that the favouring mechanism constituted competition on the merits and could not therefore be qualified as an abuse. This argument was rejected by the GC. It ruled in respect of this that:

1. favouring/leveraging practices are indeed not necessarily prohibited (but they can be in certain circumstances);

2. the criticised practice not only concerned self-preferencing but also the simultaneous demotion of results from competing comparison services (by using an adjustment algorithm).

   The GC rejected Google’s assertion that it did not discriminate between its own comparison shopping services and that of its competitors. The GC considered that the differing treatment of services was not based on an objective difference between the services, and even ruled that the demotion and more generic display of competitors’ results was not necessarily proportionate to the claimed lesser degree of relevance.

   It also rejected Google’s argument that the Google shopping results should have been compared with the treatment of (paid) advertisement and not with free generic results. It clarified that the issue of paid advertisements was not as such put to question as they are not part of Google’s comparison shopping service. Moreover, the point was made in this context that competing comparison shopping services were not allowed to make use of Google’s paid advertisement services unless they transformed themselves into merchant platforms, a move which would turn these suppliers into customers (as opposed to competitors), of Google’s comparison shopping service.

   Google’s argument that the EC had unjustly found that Google favoured its specialised Google Shopping page was equally rejected on the grounds that the EC targeted not only the Google Shopping page but also Google’s comparison shopping service as a whole (whereby the position and display of Google Shopping was considered to be only one way of Google favouring itself).

3. any practice – including quality improvements and favouring – capable of having anticompetitive effects can deviate from competition on the merits. The GC referred in this respect to the impact on traffic on both Google’s comparison shopping services and the competitors’ comparison shopping services, and on the behaviour of internet users, who typically concentrate on the first results.

   The GC clarified that the effects needed to be assessed taking into account the favouring and the demoting practices in combination (as opposed to an assessment for each practice separately). It further ruled that the EC was not required to compare the situation with a counterfactual scenario but could instead (i) identify a correlation between the alleged abuse and internet traffic; and (ii) subsequently find causation based on additional information such as market participants’ assessments. The GC confirmed that potential anticompetitive effects are also relevant; it found however that these were not sufficiently demonstrated to be present in respect of the markets for general search services, qualifying the EC’s assessment as “too imprecise” and “purely speculative”.

4. Google's general search pages are designed to be open and universal – ie all websites are displayed – making self-preferencing in such context “a certain form of abnormality” and therefore “not necessarily rational” (unless in case of dominance and entry barriers that do not allow a sufficiently quick competitor’s response); and

5. in such context the burden of proof was on Google to justify the difference in treatment. The GC made the analogy with the net neutrality obligation of internet access providers. It concluded that “certain differences in treatment” by a dominant undertaking in the internet sector may therefore constitute an abuse of a dominant position, emphasizing in this respect Google’s “special responsibility” as a result of its “undisputed ultra-dominant position”.

Google also argued that the EC was wrong (i) to qualify the alleged abuse as ‘leveraging’ rather than as a ‘refusal to supply’; and (ii) consequently to not apply the constituent conditions of such abuse (in particular the indispensability requirement). The GC disagreed with both the EC and Google and ruled instead that the abuse concerned the conditions of supply. The GC reasoned that a refusal to supply requires an express access request and a subsequent refusal and that it does not include implicit refusals of access. The GC concluded that the abuse constituted a discriminatory practice. On this basis, the GC ruled that the EC was not required to meet the indispensability requirement (but nevertheless found that the EC had in practice done so).

Google’s argument that its practices were aimed at improving the quality of its general search service – ie to compete on the merits – and not to drive traffic to its own comparison shopping services, was equally rejected. The GC referred in this respect to the established case-law that an abuse of a dominant position is an objective concept whereby intentions are relevant yet not sufficient by itself or necessary. Moreover, the GC found that the EC had in fact not considered any anticompetitive strategy or objectives possibly pursued by Google.

Finally, the GC rejected Google’s argument regarding the existence of an objective justification for the practices. It reasoned that while the experience of some internet users may have improved, such improvement was not likely to “counteract the harmful effects” of the practices concerned. This suggests that the GC applied a balancing test – whereby Google had the burden of proving the pro-competitive effects – instead of a usual objective justification test similar to that of Article 101(3) TFEU.

The GC concluded by confirming Google’s fine, rejecting Google’s assertion that the legality principle (nulla poena sine lege certa) should have prevented the EC from imposing any fine on Google for what constitutes a novel abuse. The GC ruled in this respect that in such case, the relevant legal criterion is whether the infringement was committed intentionally or negligently. It found in this context that Google “must have known that its conduct undermined equality of opportunity (…) and that that conduct was capable of foreclosing its competitors or restricting competition”, concluding that Google’s abuse was “pursued intentionally”.

This long-awaited Judgment has confirmed, for now at least, that self-preferencing can be added to the string of potential abuses open to undertakings deemed dominant, in particular in the digital sector, but also potentially in other sectors. It is unclear whether Google will appeal this case to the EU Court of Justice.

Should you need assistance, have any further questions regarding this client alert or competition law generally, please do not hesitate to contact any of the individuals listed or your usual contact at Simmons & Simmons.