I have spent 15 days reporting live from the Epic v. Google trial: an antitrust dispute over whether Google’s Android app store is an unfair monopoly. I’ve watched a parade of witnesses go by, including Epic CEO Tim Sweeney and Google CEO Sundar Pichai. We’re now in a weeklong break before both parties return on December 11th to make their closing arguments, after which a jury will decide who’s right. I’ve chronicled every major thrust, parry, and riposte leading up to that in our Verge StoryStream, writing nearly 600 dispatches from the courtroom so far.

But who’s got the time to dig through all that, am I right?

So here are straightforward versions of the 20 most interesting things we’ve learned — starting with the fact that Epic could win the whole thing.

1. Epic could win.

I thought Epic was tilting at a windmill by challenging Google after losing so thoroughly to Apple. If Epic couldn’t prove Apple’s walled garden is an illegal monopoly, how could it beat the company that made Android an open-source project from the start?

After just over a fortnight’s worth of cumulative evidence — loosely speaking — the answer has become clear. A decade after Skyhook, Google still uses a tangled web of contracts and deals to make partners fall in line — incentivizing phone makers and the biggest app developers to share, rather than compete for, an ever-growing pile of app store cash.

Judge James Donato, who’s presiding over this jury trial, thinks Epic absolutely has a chance of winning. On Friday, he declined to hand Google a default victory, saying there’s “more than enough evidence for the jury to find for plaintiff on each of their claims.”

Mind you, it would still be a high bar to clear! A jury would have to decide not only that Google has an illegal monopoly on “Android app distribution” or “in-app billing services on Android devices,” but also that the harms outweighed any competitive benefits.

So let’s start with something easier. Was Google scared for its profits?

2. Google was running scared — of Epic specifically.

I had some doubts on day two, but not anymore: Google internally saw Epic Games as a “contagion” that would infect Android’s biggest game and app developers, turning them into “agitators” that would demand lower fees and ultimately defect from the Play Store. Google’s “biggest Play priority” for Q1 2019 was to address agitators like Epic, and Epic’s name appears many times in internal Google strategy documents we saw.

Google feared as many as 100 percent of top game developers would follow Fortnite’s lead by launching outside of Google Play, denying Google a 30 percent cut of revenue, and that fear doesn’t seem to have been misplaced. Not only were other “agitators” in direct contact with Epic, but many either launched or were internally planning to launch a competing game store on Android — before Google approached them with a deal.

Google called this the “contagion effect” and forecast it might lose $2 billion in revenue by the end of 2022 — perhaps even $3.6 billion. The company estimated it would miss out on $250 million from Fortnite alone, and $243 million a year if Activision Blizzard King (ABK) removed its games like Candy Crush from Google Play.

And so, Google formulated a plan. It would save money by spending fewer billions — perhaps $1.78 billion? — on “Play Risk Mitigation.” It came up with multiple strategies it would pursue in parallel, including “Project Hug,” “Project Banyan,” and “Project Elektra.”

3. Project Hug targeted a who’s who of mobile game developers — and 20 of them signed.

Project Hug was a plan to invest money in game developers that were at risk of “contagion,” we learned in court. Google approached 22 of them in total, including Activision, Aniplex, Bandai Namco, Bethesda, Blizzard, Com2uS, EA, King, Mixi, Niantic, NCSoft, Netmarble, NetEase, Nexon, Nintendo, Pearl Abyss, The Pokémon Company, Riot, Square Enix, Supercell, Tencent, and Ubisoft.

At one point, the Google Play biz dev team asked for $575 million through 2022 to make those deals, though it’s likely the final numbers are far higher — Activision alone was offered over $360 million in a deal that’s now apparently worth billions.

Twenty game developers signed. But not Supercell — and not Epic, which got a special $147 million offer designed to convince the company to launch Fortnite on Google Play.

Project Hug was formalized as the Google Games Velocity Program.

4. Epic failed to show Project Hug was a direct “block” or “bribe.”

In Epic’s opening arguments, it promised to show the jury that Google literally pays competitors not to compete — and it seemed like Project Hug would be the way.

But over the course of the trial, we learned that all but one of Google’s “bribes” weren’t direct payments. (It does have a long-standing deal to pay Motorola and may have paid LG.)

Instead, as you can see in this breakdown of a $90 million Project Hug deal with Riot Games, the company generally promises millions upon millions of dollars in support, credits, gift card programs, promotions, and dedicated access to Google staff. And in exchange, it doesn’t block a game developer from shipping on a different store, as long as it arrives on Google Play the same day and at the same quality.

Epic’s economic experts suggested that Google’s deals are effectively the same as giving out money, but Google’s experts suggest it’s more like a discount.

5. Project Banyan would have made Samsung’s Galaxy Store a false choice, but it never happened.

The proposed 2019 deal would have seen Google Play and the Galaxy Store side by side on the default homescreen on Samsung, just like today, only Google would host the games, provide the billing, security, and updates in the Galaxy Store, too. But it never happened.

Epic tried to suggest that Google pushed Samsung into nerfing the Galaxy Store anyhow, raising the specter of handshake deals, but Epic never really provided enough evidence of either to be convincing. All I know is that Samsung did sign three deals valued at $8 billion in 2020, and the Galaxy Store apparently accounts for just 1 percent of Android downloads despite being on an estimated 39.5 percent of Android phones (as of mid-2021).

5. Project Elektra would have seen Google purchase a controlling stake in Epic — but it didn’t happen either.

Over two days in July 2018, Google cooked up a business justification for purchasing enough of Epic Games to “sway them on Epic’s approach to Android.” However, it’s not entirely clear whether a couple of Google execs seriously wanted to prop up a monopoly or were just casually brainstorming an opportunity.

The language certainly does seem suspect, e.g. “investment was the only way people could realistically think of to sway them on Epic’s approach to Android — because it’s easy to imagine us investing billions at some ridiculous valuation.” But there were certainly other reasons Google might have been interested in the Fortnite maker, as the person who wrote that email suggested in court.

Either way, I’m guessing Google quickly abandoned the idea when it found out Epic CEO Tim Sweeney is also the controlling shareholder.

6. Google execs clearly wanted to block other app stores.

While multiple Google executives testified they only cared whether games made it to Google Play — not rival app stores — internal documents told a different story.

  • One Google Play exec boasted she got Riot to halt development on its own app store by promising the company $10M in marketing.
  • Another exec wrote that one of Google’s two objectives with Riot was “have Riot choose Play vs. launching their own Android store.”
  • Despite its own stated goal of launching “on as many platforms as possible” to maximize its margins and deciding it would be fairly easy, Riot clearly felt Google didn’t want that and doing so would jeopardize a deal.
  • Another Google exec suggested that sharing revenue with Samsung would be “better at disincentivizing other app stores from being preloaded.”
  • What Google thought it would get from a proposed deal with Samsung: “No Galaxy Store.”
  • Like Riot, Samsung understood that Google wanted it to “get out of the store business,” too.
  • Google wanted to offer a 16 percent Google Play revenue share to key Chinese OEMs to “secure Play exclusivity” — and wound up giving them 20 percent.

7. And on some phones, Google arguably did.

During the trial, we learned that Google’s RSA 3.0 — those Revenue Sharing Agreements it signed with key Chinese OEMs among others — have a “Premier Tier” that does in fact keep phone makers from preinstalling alternate app stores without getting a special exemption from Google.

OnePlus, for example, gets:

  • 20 percent of “net basic ad revenue”
  • 10 percent of “net optimized ad revenue”
  • 5 percent of “net optimized Play transaction revenue”
  • 15 percent of “net premier ad revenue”
  • 20 percent of “net Play transaction revenue”

Epic was not able to prove this was a “block,” because we learned that phone makers like OnePlus are completely free to decide how many of its phones fall into each “tier” of the agreement. One Google exec suggested fewer than 25 percent of new Android phones are RSA 3.0 Premier, and only 0.4 percent of Fortnite-capable phones were RSA 3.0 Premier as of November 2021, according to a Google expert witness.

But Epic astutely suggested that an alternate app store would have to offer similar dollars to phone makers to get traction — and Google’s expert conceded that point.

8. Google’s contracts arguably had some pro-competitive benefits.

We at The Verge have long decried Android fragmentation and longed for closer-to-stock Android experiences. It’s historically been easier for OEMs and carriers to get their phones updated and patched if they’re not trying to reinvent the wheel with their own platforms and apps.

Google pointed out that its contracts do things like that, helping Android phones compete with the iPhone, and a Motorola exec highlighted its “Pure Android strategy” as a competitive strength and differentiator from other Android phones. Google CEO Sundar Pichai himself pointed out there’s not a single OEM selling phones without MADA contracts (outside of China, presumably) because people want Google apps and services. I can’t disagree; I’m one of them!

9. This trial destroyed any notion that Google treats developers fairly and equally.

I’ll just leave these headlines right here.

The first one is particularly important: Spotify can pay 0 percent (yes, zero) to get all the benefits of Google’s Play Store save payments.

Not only did Epic show Google cuts sweetheart deals with the biggest app and game developers, Google itself revealed that it wanted to keep things unfair to keep profits high. While the fact that Spotify had a deal with Google was public, Google’s lawyers fought hard to keep the numbers involved secret because they’d be “very, very detrimental for the negotiation we’d be having with […] other parties.”

And when a Google lawyer asked why it might be bad to compete on price outside of “strategic partners” more generally, VP Paul Gennai responded: “If that were to become known, other developers would come to us to do the same, and … negotiate it down, and that would play out badly for us.”

Oh, and that’s before we even get to the fact that Google didn’t force its own YouTube app to use Google Play Billing at times it was encouraging (and later, forcing) other app developers to do so.

The best defenses Google has been able to muster is that Spotify was so important that people wouldn’t buy Android phones without it, and that companies like Spotify qualify as “partners” because they help invest in new hardware platforms.

“A small number of developers that invest more directly in Android and Play may have different service fees as part of a broader partnership that includes substantial financial investments and product integrations across different form factors,” Google spokesperson Dan Jackson told The Verge.

What Google doesn’t say, and I weirdly didn’t hear Epic suggest either: if some companies get sweetheart deals, everyone else effectively has to pay more to compete. Can you imagine trying to compete with Spotify with your new music app while paying Google 11 percent, at the same time Spotify is paying zero?

Apple is just as self-serving, of course — if not more.

10. Google’s “User Choice Billing” is all but a fake choice.

This one leaves a bad taste in my mouth, too. Spotify is effectively getting a 15 percentage point discount to use its own billing system (paying Google nothing for any customer that picks Spotify payments). Ordinary developers only get 4 percentage points off — even though Google knows full well that 4 percentage points isn’t enough of a discount to make that a real choice, because you still have to pay your alternative payment processor.

Epic found the receipts and even an internal graph showing that Google knows this: “A key element of this optionality proposal is we don’t want to give any artificial reasons to incent devs to switch off Play Billing,” a Googler wrote.

It’s not entirely a false choice. Dating app Bumble seems to have tried User Choice Billing because it found users prefer one-day subscriptions, something Google Play didn’t offer on its own.

11. Epic lost the sideloading and security arguments IMO.

Epic has repeatedly tried to suggest it’s not true that apps can simply bypass the Google Play Store, because of all the friction involved in installing an app downloaded directly from the web. It also won some points when it revealed internal documents showing that Google believed the friction might “drastically limit” the reach of Fortnite.

But that was 2018. In court, Google showed us it literally takes 25 seconds to install the Epic Games Store on a modern Android phone, and Epic was never able to show that Google intentionally designed earlier friction to keep rival app stores down. In fact, Amazon’s failure with the Amazon Appstore increasingly looks like a self-own. But I guess it does turn a tiny (for Amazon) profit.

While Epic won some points IMO for pointing out that Google’s “Unknown Sources” scare screens could and should be less scary and less discriminatory, Google made Epic look like a bit of a hypocrite, too. Google’s head of Android security repeatedly made it seem like practical, common sense that Google would have a far easier time protecting against threats within its app store rather than without, and drove home that the threats do exist.

I did like this Epic argument that Google’s broader ecosystem stacks the deck against sideloading, but Epic didn’t spend a lot of time on that.

12. Google had a hell of a time justifying its up to 30 percent fee.

It’s not easy to justify something publicly when your own employees are questioning it internally. Epic showed that time and again — starting in 2008 when Google promised it wouldn’t take a percentage at all while quietly taking 5 percent, then in 2009 when it decided to pocket the extra 25 percent it had publicly earmarked for carriers whenever it could.

  • “Things could get ugly once people find out.”
  • “We are now lying… we are now lying to our developers by not making this change public (for 6 months now!)”

A decade later, Google internally estimated many app devs overpaid as much as $1.43 billion per year compared to the value they received. Though Google argued that was just one way of doing the math, these figures seem to have influenced business decisions.

After concluding that “Tinder is now deriving only 10 percent of the revenue share value versus the 30 percent they pay,” one exec suggested Google should offer a special 15 percent rev share deal to Tinder’s parent company.

And while Google calculated that the value of Google Play to developers was in discovery, payment processing, and mass distribution, that discovery bit is in question: “Our value prop is limited to that of a data pipe,” wrote one Play bizdev lead, adding, “Play Store offers little value for app discovery.” Google never wound up meaningfully challenging that in court.

13. Google’s profit margins sound unreal.

Another reason it’s tough for Google to justify 30 percent: it appears that Google Play has become exactly the “profit center” it promised not to be. In the first half of 2020, the company internally bragged that it was “one of the most profitable businesses” in the world, and “very close to being a Fortune 100 company all by itself.”

Epic’s expert calculated the Google Play Store grew from 26 percent operating profit margins to 71 percent in 2021, generating north of $12 billion in operating profit that year.

While Google suggests that doesn’t account for the $40 billion R&D costs of developing the Android operating system, among other things, we saw that Android’s costs are seemingly accounted for in its own profit and loss statement (P&L).

14. Google did successfully argue it deserves something for Play.

If not 30 percent or 15 percent, how much? That’s not exactly a question before the court — but Google did repeatedly and successfully point out that its Google Play service fee is the primary way it gets paid for downloads.

While Google does sell ads on Google Play, and ads in Google Search (the default search engine on most phones), one of Epic’s economists conceded Google would get paid nothing to distribute apps across the world if a developer chose an alternative payment processor. Sweeney himself agreed that “a payment processor doesn’t make it possible to distribute a game to 60 million users,” to use Google’s words, and that “there’s nothing wrong with making your revenue off a limited number of customers.”

And while Epic tried to show that Google’s actual cost of payment processing is 6 percent and developers could get the same by ditching the middleman, Google successfully pointed out that you’d pay more than 6 percent for the most common microtransactions at rivals because they often charge a per-transaction fee in addition to their cut.

Google concedes it could have figured out a different way to charge for its many services but felt this was the simplest.

15. Epic’s attempt to show Google and Apple getting cozy looks like a nothingburger.

Epic talked a big game about how Apple and Google’s vision was “that we work as if we are one company,” based on a direct meeting in 2018 between Apple CEO Tim Cook and Google’s Pichai.

But when we got down to brass tacks, it turned out that Epic only had secondhand notes from that meeting — and that the “quote” was simply a paraphrase. Google explained away the CEO meetup as a one-time thing and pointed out there’s no record of the meeting having anything to do with Google Play or app stores.

Pichai did admit that Google pays Apple a 36 percent revenue share for search results, in exchange for Google being the default search engine on the iPhone, however.

16. Both Google and Epic planted unfair stories in the press to advance their agendas.

We saw the receipts proving that Google decided to single out Epic by embarrassing the company with news stories about a (potentially nasty) bug in the Fortnite Launcher even though Epic fixed that bug in a single day — and even though Google never publicized the same kind of vulnerability in apps of its own. Here’s how Google seemingly justified it internally.

Epic accused Google of planting a second story, too.

But Epic’s hands aren’t clean: the company’s Coalition for App Fairness was actually solely created by Epic to lobby for its desired outcomes, and we learned in court that Spotify, Basecamp, Match, Tile, Blix, Deezer, and others weren’t actually founding members despite what it told press.

17. Google tried and failed to establish the conditions for an easy victory.

Google’s core argument didn’t land. “You cannot separate the quality of a phone from the quality of the apps in its app store, and that means Google and Apple compete against each other,” said Google lead attorney Glenn Pomerantz in his opening argument.

But Epic showed that Google had an “existential question” that had nothing to do with Apple — and that again and again and again, Google wasn’t thinking about Apple when it was thinking about securing the future of Google Play. The court docs we saw almost always showed it thinking about the potential contagion of rival app stores on Android instead.

Google did find a couple oblique mentions of Apple in the end and spent quite some time explaining all the ways it does compete with the iPhone maker on pricing, features, and security. Execs insisted that many of Google’s actions were pro-competitive because they helped Android phones compete with the iPhone at a time when Google was concerned that Android may be weak — “losing users it really doesn’t want to be losing” like teens and wealthier users, Google’s economist Catherine Tucker pointed out.

But Judge Donato wasn’t buying her arguments that Apple directly participates in the markets we’re concerned with — and has said he wasn’t willing to apply aftermarket theories that Apple might be a participant because people buy their phones for apps. He said we hadn’t seen evidence of that.

That’s tough for Google because Tucker’s proposed market definition — “the facilitation of digital content transactions” — is close to the one that helped Apple win several years ago. Instead, the jury will likely decide whether Google has monopoly power within the Android ecosystem.

18. Epic will likely pay Google $398,000 even if it wins.

Epic didn’t even try to suggest it didn’t breach its contract with Google, which means Epic probably owes Google money for the short period of time that Epic managed to sneak its own billing system into Fortnite before it got booted off the Play Store. Epic owes $398,931, according to Google’s calculations.

Epic didn’t substantively challenge that amount, either — simply asking if it factors in payment processing fees.

But in court documents filed Monday, we learned that Epic does want the court to decide whether the contract was even valid before Epic pays up.

19. Epic didn’t sue just out of the goodness of Sweeney’s heart.

“Epic decided to stand up because that’s what you do to a bully,” claimed Epic lead attorney Gary Bornstein in his opening arguments — and it’s true that Epic isn’t suing for cash. The company’s asking for an injunction that stops Google from blocking alternative app stores and billing systems, not damages.

But Sweeney deflated some of the hero narrative when he took the stand, agreeing that his company stands to make billions of dollars by paying Google nothing. He did turn it around a bit, though, when he pointed out that’s how PC applications have (largely) worked for ages.

20. Everything could hinge on missing chats.

On Friday, Judge Donato vowed to investigate Google for intentionally and systematically suppressing evidence, calling the company’s behavior “a frontal assault on the fair administration of justice.”

He decided to stop short of telling the jury in this case that they should assume Google deleted damning evidence. But he did so partly because Epic has already suggested that countless times during this case. Epic grilled executive after executive, lawyer after lawyer, about why they didn’t change the default setting that would auto-delete their Google Chats after 24 hours — or worse, intentionally got colleagues to turn off chat history to keep things from being preserved — despite their legal obligations to preserve any relevant evidence.

The jury will be told that they may assume Google deleted important evidence, and that may make all the difference.

This was perhaps the most stunning exchange in the entire trial:

“You cannot guarantee that the documents that were destroyed will contradict the testimony we’re going to hear?” Lopez couldn’t.

“We’ll never know that,” Moskowitz replied.


Leave a Reply

Your email address will not be published. Required fields are marked *