On Jan. 16, four app development companies filed a class action against Facebook in the Northern District of California, alleging that Facebook monopolized the social data and social advertising markets in the U.S. and globally between May 24, 2010 until April 30, 2015. Reveal Chat HoldCo LLC v. Facebook, Inc., 3:20-CV-00363 (N.D. Cal. Jan. 16, 2020). The complaint alleges two types of anticompetitive conduct: (1) denying competing third-party app developers access to Facebook's platform, and (2) acquiring and integrating nascent competitors WhatsApp and Instagram.
The Plaintiffs' First Set of Claims
Plaintiffs allege that Facebook violated Sherman Act Sections 1 and 2 by using market dominance in the "social networking market" to erect social data barriers to entry and monopolize "social advertising" and "social data" markets. Facebook allegedly excluded competing third-party app developers and direct competitors from these markets when it denied access to its platform by: removing Friends and News Feed APIs; refusing to sell social data; using whitelists and data sharing agreements with competing platforms to collect user data; and blacklisting rivals who refused data-sharing agreements (i.e., blocking them from access to Facebook's platform) and "scraping" their social data without consent.
These claims raise several legal questions that Facebook undoubtedly will raise in the case. Below we have discussed a few of them.
I. Are "social data" and "social advertising" cognizable, sufficiently-pled antitrust markets?
Did the plaintiffs plead sufficiently plausible markets?
Prior decisions in the Northern District of California suggest that judges there may view allegations of very narrow online advertising markets skeptically. In two cases against Google, the court rejected a proposed online search advertising market definition as implausible, holding that "search advertising is reasonably interchangeable with other forms of Internet advertising" because "a website may choose to advertise via search-based advertising or by posting advertisements independently of any search." Kinderstart.com LLC v. Google, Inc., C 06-2057 JF (RS), 2007 WL 831806, at *6 (not reported) and Person v. Google, Inc., C 06-7297 JF (RS), 2007 WL 832941 (N.D. Cal. Mar. 16, 2007) (not reported), aff'd, 2007 WL 1831111 (N.D. Cal. Jun. 25, 2007) (not reported), aff'd, 346 Fed.Appx. 230 (9th Cir. 2009).
Here, one question is whether it is plausible that social advertising is not reasonably interchangeable with other online advertising formats from an advertiser's and app developer's perspective.
Likewise, judges may view plaintiffs' "social data" market definition skeptically. They describe how Facebook and other social networks collect different types of data, in many different contexts, which may be used by advertisers to target users. Compl. ¶ 316-20. But they do not appear to provide any "social data" market definition, beyond stating that "social data arises from engagement within a social network among its users." Compl. ¶ 321. Many judges may consider that insufficient, since it does not explain why social data is not interchangeable with other online user data for purposes of targeting ads, though there are examples of judges that have allowed claims of such narrow markets to proceed. See, e.g., hiQ Labs, Inc. v. LinkedIn Corp., 273 F.Supp.3d 1099, 1118 (N.D. Cal. 2017), aff'd, 938 F.3d 985 (9th Cir. 2019).
Did the plaintiffs sufficiently plead effects on all sides of the market?
The Reveal Chat complaint also presents another market definition question: whether social advertising qualifies as a multi-sided "transactional" platform under the Supreme Court's Amex decision. Ohio v. Am. Express Co., 138 S. Ct. 2274 (2018). The Amex case suggests that if the platform at issue is a transactional platform, a plaintiff must in their market definition account for dynamic responses to the alleged conduct by all customers of the platform (not just one side) and establish that the specifically alleged conduct on the whole harms all customers of the platform (not just one set of customers). In Amex, the court indicated that newspaper advertising is not a transactional platform, because the relationship between advertisers and readers on the platform is not a one to one transaction like credit cards. Id. at 2286. Its Amex ruling would not apply in such markets. In targeted social media advertising, which seems to be the Reveal Chat plaintiffs' focus, the relationship between the advertiser or app developer and the user seems closer to a one-to-one transaction when a user clicks on a targeted ad, though perhaps still not quite like a credit card transaction between merchants and cardholders.
Amex was about a Section 1 Sherman Act claim regarding agreements, not a Section 2 Sherman Act monopolization claim like the one at issue in Reveal Chat. In a recent case, one federal judge has suggested that this means Amex does not apply to Sherman Act Section 2 cases. See Mem. Op., Fed. Trade Comm'n v. Surescripts, LLC, 1:19-cv-01080-JDB at 16-17 (D.D.C. 2019). It will be interesting to see if other federal judges agree with that. It is not immediately clear why there would be a distinction between Section 1 and 2 cases for purposes of applying Amex, since both cases require market definition.
If a judge were to conclude that Amex applies here, that would suggest that the plaintiffs must allege facts regarding the impact of the alleged conduct on all customers of the platform -- users, advertisers and app developers -- to appropriately plead the relevant market and anticompetitive effect. The Reveal Chat plaintiffs recognize the different sides of the Facebook platform. They point to how a lack of "critical mass of social data" would impair advertising retargeting, thus reducing revenues for market participants in the Social Data Market's advertising sales channel and the Social Advertising Market. Compl. ¶¶ 348-352. They also allege that "without adequate social data and engagement with the social network, market participants cannot display content to users that would provide enough value to generate engagement and additional social data. ... Because of network effects, users will not use a social network that lacks enough social data to provide targeted content or to provide valuable connections to other users." Id. But it is not clear that such general observations about network effects really suffice to allege facts relevant to market definition and establishment of anticompetitive effects.
II. Did plaintiffs adequately state a cognizable antitrust violation where they complain about Facebook's restriction of access to its own platform?
Where the plaintiffs take issue with Facebook's alleged refusal to grant them access to its platform, whether through APIs or otherwise, they essentially assert a refusal to deal claim. The Supreme Court has held that the antitrust laws typically do not impose a duty to deal with rivals on any firm, monopolist or not, explaining that refusal to deal claims are at the outer antitrust boundaries. Verizon Comms., Inc. v. Law Offices of Curtis V. Trinko, LLP, 540 U.S. 398, 409 (2004). It therefore is difficult to prevail on such claims. One thing that a plaintiff would have to allege and show in order to state an actionable refusal to deal antitrust claim is that a monopolist sacrificed short-term profits from dealing with rivals to gain long-term monopoly profits by excluding rivals from the relevant market and thus acquiring or maintaining a monopoly. Many courts also require that a plaintiff allege that there was a previously profitable course of dealing between the monopolist and the foreclosed rivals that the monopolist then terminated to gain its monopoly.
To try to meet this high legal standard, the Reveal Chat plaintiffs have asserted that Facebook used to profitably provide API and other access to its platform for app developers, and then terminated that access later, sacrificing the profits it earned from it to foreclose rivals. They allege that was part of a plan to stifle competition. They may face challenges in prosecuting that claim.
First, several courts in the 9th Circuit have held that the owner of a website or app, no matter how large, can decide unilaterally what it allows third parties to do on its platform, whether for privacy or economic reasons.
For example, in LiveUniverse, Inc. v. MySpace, Inc., 304 Fed.Appx. 554, 557 (9th Cir. 2008) (non-precedential), the 9th Circuit dismissed the plaintiffs' Section 2 claims that MySpace had allegedly redesigned its site to block LiveUniverse's social networking site, VidiLife.com, including preventing MySpace users from embedding LiveUniverse-hosted videos into their MySpace pages or even mentioning VidiLife on MySpace. The court held that MySpace users' ability to link to plaintiffs' content before the MySpace redesign "could indicate a prior course of dealing between MySpace and its users," but not any agreement or implicit understanding between the parties. Id. The court held that MySpace could break links to competitive sites if it wanted and that such behavior did not constitute an actionable "refusal to deal." Id. (without addressing justifications for MySpace's design changes).
Likewise, in Sambreel Holdings LLC v. Facebook, Inc., 906 F.Supp.2d 1070, 1075-76 (S.D.Cal. 2012), the court dismissed plaintiffs' Section 1 and 2 claims, concluding that Facebook was not required to make its popular social-networking platform available to others as a venue for their business. Id. at 1075-76, 1080. The court also expressed views that antitrust law cannot interfere with Facebook's rights "to control its own product, and to establish the terms with which its users, application developers, and advertisers must comply in order to utilize this product," and "to protect the infrastructure it has developed, and the manner in which its website will be viewed." Id. at 1079.
Where a plaintiff is seeking to scrape publicly available information on a website, 9th Circuit courts might be more receptive. For example, in hiQ Labs, Inc. v. LinkedIn Corp., 273 F.Supp.3d 1099 (N.D. Cal. 2017), aff'd and remanded, 938 F.3d 985 (9th Cir. 2019), the 9th Circuit affirmed a preliminary injunction based on such a claim, though that was under California state unfair competition law, not federal antitrust law. But it does not seem that the Reveal Chat plaintiffs are seeking to scrape publicly available data from Facebook. Indeed, in hiQ, the 9th Circuit specifically pointed out that "Facebook data, by contrast, is not generally accessible, and therefore is not an equivalent alternative source of data [to LinkedIn data]." Id. at 993.
III. Have plaintiffs sufficiently alleged antitrust injury where their claim is about conduct that only affects what they can do on defendant's platforms, not on other platforms?
The bulk of plaintiffs' claims are about what Facebook is precluding them from doing on Facebook's web properties and apps, not what they might do on other popular web properties and apps like Snap, LinkedIn, Twitter, TikTok, Pinterest, YouTube etc. In some of the same opinions as discussed above, some 9th Circuit courts have concluded that a website owner's decision to bar a third party from its site did not amount to antitrust injury -- that is harm to competition in the market as a whole -- because a network's actions on its own website cannot negatively affect what occurs on other websites. LiveUniverse, 304 Fed.Appx. at 557 (concluding plaintiff did not allege antitrust injury because it was unclear how "MySpace's actions on its own website can reduce consumer's choice or diminish the quality of their experience on other social networking websites, which is the relevant market.")
Another hurdle plaintiffs might face, on a motion to dismiss or, if they survive that, at class certification or summary judgment, is that a number of them seem to not be engaged in services that compete with Facebook, such that it is not clear that they are foreclosed rivals. Nor is it clear that they can all be classified as (direct) purchasers of Facebook services, though perhaps their claim is that they were "purchasers" of social data. In that respect, each representative plaintiff may be differently situated. This could lead to a number of legal questions, including whether the class representatives lack antitrust standing or injury, whether they have a plausible claim of exclusionary conduct as required for a Section 2 monopolization claim, and whether they are sufficiently similarly situated to all class members to represent them.
The Plaintiffs' Second Set of Claims
Plaintiffs also allege Facebook violated Sherman Section 2 and Clayton Section 7 by using "secret mobile surveillance" to identify and "aggressively'' acquire nascent rival social media platforms competitors WhatsApp (chat app) and Instagram (photo-sharing). Plaintiffs seem to be latching on reports of the FTC and several state attorneys general investigating those transactions by Facebook. Plaintiffs' claims challenging these acquisitions raise a number of much-debated legal and policy questions.
The Reveal Chat plaintiffs mainly seem to allege that Facebook used software to identify WhatsApp and Instagram as popular upstart social networks or messaging apps that could in the future become significant competitors, acquired those two companies, after which they grow into very popular apps.
Is that enough to state a claim under the antitrust laws? If so, that seems to allow a lot of speculation, including about cause and effect. As Plaintiffs allege, "Instagram had not at the time of the merger meaningfully monetized its user engagement and social data." Compl. ¶ 263. And, in 2012, when Facebook acquired Instagram, it only had 27 million users, relatively limited data, and was valued at $25 million. Who knows if Instagram would have grown into a significant competitor on its own - was its initial product that good, or was its success and growth in some or significant part due to acquisition by Facebook?
Conversely, if one assumes that Instagram would have become a significant competitor to Facebook on its own, creating a more competitive market, by starting from scratch without a large-scale user base and user data, does that not undermine the theory of Plaintiffs' claim that a firm needs a large user base and set of social data to be able to challenge the likes of Facebook? Examples of startups growing into highly popular social media apps like SnapChat, Twitch (now owned by Amazon) and TikTok also beg that question. Plaintiffs also highlight in their complaint that established players like Google+ and MySpace, which had a relatively large user base and corresponding user data, did not survive. That, too, begs the question how plausible plaintiffs' allegations are that commercial success or failure is tied to how much access to social data one possesses.
Another question Facebook may raise is how plausible a claim of monopoly power is when the plaintiffs identify a number of well-known and popular competitors in their complaint, such as Linkedin, YouTube, Twitter, SnapChat, Pinterest, and others. Most judges will be very familiar with those and other social networks as other household-name online companies that compete for advertisers' dollars.
Unusual Request for Relief
In addition to a damages claim, the Reveal Chat plaintiffs apparently seek to have Facebook divest WhatsApp and Instagram, as well as to have Mark Zuckerberg divest his personal control over Facebook as a stockholder. The latter request, in particular, is unusual in an antitrust case. It is unclear how it is tied to the antitrust harms that plaintiffs assert, much less how it would resolve them.