Epic v. Google

Why did Epic win against Google when they lost against Apple?

and

Dec 13, 2023

We’ve received various questions about the Epic v. Google decision and thought we would share our thoughts. As an initial matter, while the case would seem similar to Epic v. Apple, the ruling in one does not apply to the other. Other than that, the main question is: Why the opposite outcomes on seemingly similar facts?

Market Definition

The first step in analyzing anti-competitive behavior is determining the relevant markets. A relevant market, for antitrust purposes, contains both a geographic component and a product or service component. This is a question of fact, which is relevant because the Apple case was a bench trial where the finder of fact was the judge, whereas the Google trial was a jury trial. Juries are often less predictable than a judge.

In both cases, Epic tried to define the relevant market as the respective app distribution and in-app payment systems (iOS in the Apple case and Android in the Google case), as well as the respective in-app billing systems for digital goods.

In the Apple case, the judge (serving as the fact-finder) determined that the market was broader than Epic’s proposal. She found that the relevant product market was the mobile game transactions market as a whole. She also found that the relevant geographic market was global, excluding China. For the relevant product market, the district court pointed to several factors that made her decide on a broader market than Epic argued for:

Apple didn’t sell or license iOS;
Epic didn’t present evidence that consumers unknowingly locked themselves into the Apple ecosystems when buying iOS devices; and
Epic’s market definition expert was “weakly probative” (judge-speak for an untrustworthy hired gun) because their analysis was flawed along several major factors.

In the Google case, Epic was able to learn from the failures in the Apple case. Epic could present evidence of consumer lock-in as well as changing experts and refining the expert analysis. Ultimately, the jury agreed with Epic’s market definition (which Judge Rogers had rejected in the Apple case).

Because the product market definition in Google (Android App and Android payments market) was much narrower than that in Apple (video game mobile transactions market), it was easier to show abuse of that market in Google than in Apple.

Jury Instructions & Findings

The jury in Google found that various Google agreements, like Project Hug, Developer Agreements, and OEM agreements were unreasonable restraints of trade. These included agreements with major developers and publishers. Other than the Developer Agreement, the Apple case did not appear to have similar agreements in evidence.

In addition, the jury in Google received an instruction from the judge allowing them to make a “permissive inference” because of Google’s bad behavior related to discovery issues. Specifically, the judge instructed:

You have seen evidence that Google Chat communications were deleted with the intent to prevent their use in litigation. You may infer that the deleted Chat messages contained evidence that would have been unfavorable to Google in this case.

In other words, the jury was primed to conclude that Google had acted inappropriately, making it more likely that the jury would ultimately side with Epic, at least on evidentiary issues.

Attacking the Verdict

Google has stated that it will seek to overturn the jury’s verdict. It has several options, including attacking the verdict at the trial court level and/or by filing an appeal. It may do both. Unfortunately for Google, the relevant market determination is a question of fact that is reversible only if the evidence compels a conclusion contrary to the factfinder’s verdict. This is also known as a “clear error” standard and is difficult to overcome on appeal. Here, Google would need to convince the Ninth Circuit that the jury clearly erred in their determination that the relevant product market was (1) the Android app distribution market and (2) the in-app billing services for digital goods and services transactions.

Reconciling the Outcomes

One of the hardest things about this verdict is that it seems irreconcilable, conceptually, with the Apple decision. But this is how the law works. Plaintiffs have to prove their cases with evidence. In the Apple case, Epic failed to do that; whereas in front of the jury in the Google case, it learned from its past mistakes, had some extra help from Google’s bad behavior, and ultimately succeeded.

We predict this case, like Apple’s, will get appealed all the way to the Supreme Court. There is just too much at stake. However, Epic’s appeal in the Apple case is fairly narrowly drawn to a standard used in the anti-competitive analysis and the ruling is unlikely to apply to the Google case. Similarly, the issues at stake in Google are probably not going to be entirely relevant to the Apple case.

What’s Next?

For many, the big question is: What will the remedy be? This will determine the practical effect for companies other than Epic. Stay tuned -- that will come early next year.