CMA blocks the Facebook/Giphy merger: you can’t say they didn’t warn us

The UK Competition and Markets Authority (“CMA”) investigation into the acquisition of Giphy by Facebook (now Meta) has been watched by merger control practitioners with morbid fascination. It has included everything one could wish for in a merger case – a Big Tech killer acquisition (arguably), novel theories of harm (arguably), satellite litigation, the highest ever procedural fine, and ultimately a foreign regulator blocking a US transaction. The merger was formally prohibited on 30 November 2021.

This blog post will explain what the case means for UK merger control in the coming years.

Context

As readers of this blog will know, the CMA published a 2,000-page market study report in July 2020, which showed Facebook to have substantial market power and a strategic position as a gatekeeper in the UK economy. The CMA’s Digital Markets Taskforce recommended an enhanced merger control regime for Big Tech acquisitions (which just about made it into the UK Government’s White Paper in July 2021 – see here and here), and the CMA’s CEO, Andrea Coscelli, made his clearest statements on the issue in a speech in October 2021:

“In relation to mergers: of 400 acquisitions made by the largest digital firms between 2008 and 2018, none were blocked by competition authorities. There is now a general consensus that some of these acquisitions should not have gone ahead and that they allowed these firms to amass and reinforce their market power.”

and

“This does not just matter for individual consumers and businesses: it matters for innovation and growth, and therefore for the economy as a whole. This is visible in the acquisition strategy of these powerful firms, notably their tendency to acquire innovative new entrants and potential challengers. This dynamic is thwarting innovation and reducing incentives to develop businesses with real-world profitability. … These acquisitions, over time, represent a potentially enormous loss of consumer welfare.”

Any attempted acquisition of further market power by Big Tech firms must be viewed in this context. You are free to think the CMA is misguided, but you should not be surprised when the CMA gives the firm a rough time.

In addition, this particular case has been bad-tempered throughout. Facebook took the CMA to court mid-way through the investigation when the CMA did not grant it a derogation from the standard hold-separate order. Facebook lost in the Competition Appeal Tribunal and again in the Court of Appeal. The other recent example of such satellite litigation on procedural matters is Sainsbury’s/Asda, a large supermarket merger that was also blocked by the CMA.

Then, in October 2021, the CMA fined Facebook over GBP 50 million (nearly USD 70 million) for “consciously refusing” to comply with the hold-separate order. These types of considerations inevitably affect a competition authority’s deliberations on the substance of a case. 

Interesting though the procedural aspects of this case are, I will focus more on the substantive competition analysis in this blog post.

Why did the CMA block the deal?

According to the CMA, Facebook’s platforms (Facebook, WhatsApp, Instagram) account for 73% of user time spent on social media in the UK, and Giphy is “the world’s leading provider of free GIFs and GIF stickers”. The two parties operate mostly at different levels of the supply chain, with Giphy’s services being an input for social media platforms. In any market, that basic fact pattern would give rise to prima facie competition concerns.

There has been some commentary poking fun at the CMA for taking a stand over something as inconsequential as free GIFs, but Facebook is one of the most powerful companies in the world, and GIFs are central to how the world communicates nowadays, so I do not see why this market is any less important than any other market.

The CMA ultimately found a substantial lessening of competition (“SLC”) on two bases:

• horizontal effects resulting from the loss of potential competition in display advertising; and
• vertical effects on competition in the supply of social media arising from input foreclosure.

It is striking that the CMA based its findings almost entirely on the internal documents of the merging parties and of third parties, and on its discussions with a wide range of market participants. This is not a case where the CMA’s econometricians and data scientists ran the investigation. The CMA’s analysis in dynamic markets has increasingly relied on this type of qualitative evidence as it is often the best-available evidence on what the market expects will happen in the future. It is also easier to block mergers with this type of evidence, but I make no comment on the CMA’s motivations.

• Horizontal effects

Giphy had what the CMA describes as an “innovative” advertising product that was available in the US pre-merger, but not in the UK. Giphy’s internal documents showed that it “hoped” to develop its product and expand internationally, but it had not yet done so. Hence, the merger arguably resulted in a zero market share increment in the UK, although that fact does not tell the whole story.

Facebook had shut down Giphy’s advertising product once it took over Giphy, thus triggering “killer acquisition” alarm bells. In any case, there would have been a theory of harm based on the loss of potential competition. This is the concern that the merger would remove a business that had the potential to compete with Facebook in display advertising. (Of course, Facebook is the company that famously had, or perhaps still has, a “copy, acquire or kill” strategy when faced with competitive threats).

The CMA’s Merger Assessment Guidelines (“Guidelines”) were updated earlier this year specifically to improve the CMA’s ability to analyze mergers in dynamic markets such as this one. Digital platforms are cited as an example of where losses of potential competition are particularly relevant to a merger assessment (Guidelines, para 5.4). The Guidelines (para 5.15) also state that “the impact is likely to be more significant … where the other merger firm would already have market power absent the merger (with greater market power being associated with a greater likelihood of an entrant having a bigger impact on competition)”, and “where one merger firm has a strong position in the market and there are few significant potential competitors, even small increments in market power may give rise to competition concerns. Therefore, the acquisition by any such firm of a potential entrant may be concerning even if that potential entrant is expected to be small.”

The CMA seems to have simply followed its own published guidelines. It found that Facebook has significant market power in display advertising, so the CMA’s starting point is that any further lessening of competition (or threat of competition) is potentially concerning. It found that Giphy itself also had a strong position for social media engagement tools. It also found that barriers to entry were high. The CMA had evidence from Giphy’s customers who were positive about its product, internal documents from Giphy discussing significant interest from customers regarding international expansion, and the CMA itself assessed the product and found it had some advantages. Added to that, the CMA found evidence that Facebook and other market participants were interested in monetizing the same or similar social media features as Giphy’s product did.

Overall, one can certainly argue about whether such a concern can plausibly constitute a “substantial” lessening of competition, but the theory of loss of potential competition is clearly set out.

• Vertical effects

The CMA also found vertical effects concerns based on a fairly traditional input foreclosure analysis. SLCs based on vertical effects are less common, but not unheard of (see for example the merger between Intercontinental Exchange and Trayport which was blocked on the basis of vertical theories of harm only).

Giphy had only one close competitor (Google’s Tenor) in searchable GIF libraries, which is an important input for social media platforms. Facebook has significant market power in social media and it competes with those other social media platforms. The CMA therefore assessed whether Facebook would have the post-merger ability and incentive to limit other platforms’ access to Giphy, and whether this foreclosure would have an effect on the ability of those platforms to compete with Facebook.

The CMA found that Facebook would deny or limit other platforms’ access to Giphy GIFs, driving more traffic to Facebook’s platforms. Facebook could also change the terms of access by requiring other platforms to provide more user data in order to access Giphy GIFs.

Once the CMA felt it had sufficient evidence to show other platforms’ reliance on Giphy, a vertical effects SLC was very likely to be found. If Giphy’s services were straightforward to replicate, or if other platforms could operate successfully without that type of service, it would have been more difficult to find an SLC.

Remedies

Facebook had already completed the transaction, as it is allowed to do under UK merger control rules. The CMA decided that behavioural remedies or a partial divestment would not comprehensively solve its competition concerns, so Facebook is being forced to sell the entirety of what it has bought. It takes some exceptionally good arguments to move the CMA away from a full divestment remedy nowadays.

Conclusion

The aggressive approach of the CMA in both digital markets and merger control were brought together in this case, with predictable outcomes. For both the horizontal and the vertical SLCs, I would argue that the theories of harm are not actually particularly novel or unexpected. The decisive factor is that the threshold for finding an SLC and therefore intervening in a merger is now so extremely low, particularly where one party already has market power. That low threshold is a very difficult thing for merging parties to challenge in court because the UK uses a judicial review standard that is slow to interfere in the CMA’s discretionary judgment calls on the meaning of “substantial” in the SLC test.

The new Guidelines have now been employed in earnest, and the CMA will defend its decision to the death. The CMA has made it clear to everyone who will listen that it intends to prevent the accumulation of further market power through acquisition by a company that already has a strong position.  Facebook is one of a handful of unusually powerful platforms, but the principles in this decision can readily be applied in other contexts too. 

Merging parties in markets with few (even up to 5 or 6) main players should take note. If they shout at the CMA about a low market share increment, they are simply not paying attention to the realities of the new world. They need to supplement their arguments with a more sophisticated analysis of closeness of competition and evidence about how the relevant markets are developing.

We assume an appeal against the CMA’s decision is currently being drafted, and it would need to be filed at the end of the Christmas period. The morbid fascination of this case may not be over yet.

[Disclosure: The author is a former legal director at the CMA and helped to write the new merger assessment guidelines referred to above, but he was not involved in this case.]