Epic Won the War It Wasn't Fighting and Lost the One It Started

In December 2018, Epic Games went to war with the most entrenched standard in PC gaming. Steam took 30 cents of every dollar spent on a game; Epic launched a rival store and offered developers 88 cents of that dollar instead of 70.⁸ It was a clean, aggressive bet: undercut the toll, buy exclusivity, and pry open a market Valve had owned for fifteen years. Six years later that store reached 295 million PC users — and was still losing money, with more than three-quarters of its revenue coming from Epic's own games.⁸ Meanwhile, almost as a footnote, Epic was quietly winning a different standards war it barely talked about. That second war is the one that paid the bills.

The official story is that Epic is the scrappy challenger taking on the gatekeepers — Steam, Apple, Google — and winning. The real story is stranger and more useful. Epic won the standards war it wasn't fighting, and is steadily losing the one it loudly started. The engine became a standard while the store stayed a subsidy.

The war Epic actually won is the one it rarely fights about

Unreal Engine doesn't run loud exclusivity campaigns or sue Apple. It just sits underneath other people's games, taking a small royalty when they succeed. And by 2024, that quiet position had become a standard: Unreal powered 28% of all games released on Steam and captured 31% of total Steam game-sales revenue.² Here's the part that matters. Unity still leads by raw game count — roughly 51% of Steam releases to Unreal's 28%.² So Unreal is not the most-used engine. It's the most-earning engine, because it concentrates in the high-budget AAA titles where the money is. That is what winning a standards war actually looks like: not the biggest install base, but ownership of the segment that pays. Unreal Engine revenue climbed from $150M in 2021 to $275M in 2023 — small against Epic's billions, but pure, defensible royalty that compounds with every blockbuster shipped on the engine.¹

The store was never a mistake — it was a land-grab that didn't land

The easy read on the Epic Games Store is that Epic stumbled into a money pit. That's wrong, and the correction matters. The losses were deliberate, funded, and disclosed in internal projections shown at the Apple trial — Epic expected the store to stay unprofitable until at least 2025 and to accumulate a combined loss of $965M by 2027.⁵ This was a chosen war, paid for out of Fortnite's profits, to pry developers off Steam's 70/30 split with a far more generous 88/12.⁸ The bet wasn't crazy. The problem is what the spending bought. By 2024 the store reached 295 million users and $1.09B in total spend — but only $255M came from third-party games, meaning over 74% of revenue still came from Epic's own titles like Fortnite. Worse, third-party revenue actually fell 18% year over year.⁸ A store that spent years buying exclusives to attract outside developers was, six years in, mostly a way for Epic to sell Fortnite to itself. In 2023 the store's general manager admitted under oath that it still wasn't profitable.⁵

That sentence is the whole strategic problem in one breath. The engine war Epic won throws off thin, durable royalties. The store war Epic started was financed by Fortnite — and Fortnite is the one asset cooling off. When the funding source softens, the subsidized war can't be subsidized anymore. In September 2023 Epic cut roughly 830 employees, about 16% of its workforce; in March 2026 it cut over 1,000 more.⁶¹⁰ Two rounds, not one — a detail the single-restructuring narrative quietly drops, and the one that tells you the distress is structural, not a blip.

Didn't Epic just beat Apple and Google in court?

This is where the popular scoreboard is most wrong, and it's worth being precise. Against Apple, Epic lost. The judge ruled for Apple on 9 of 10 counts in 2021 — every federal antitrust claim — and Epic prevailed only on a single California anti-steering claim; the Ninth Circuit affirmed the split in 2023, and the Supreme Court declined to hear further appeals in 2024.³ Against Google, Epic won: a jury found Google's conduct anticompetitive in 2023, the Ninth Circuit affirmed, and Google settled in March 2026 agreeing to charge 20% or less on Epic's transactions.⁴ Opposite outcomes, different mechanisms — one a bench trial decided by a judge, one a jury verdict. Conflating them into 'Epic beat the gatekeepers' flatters a record that is genuinely split. And note the deeper point: even the win against Google lowers a toll. It doesn't make Epic's own store profitable. Litigation can pry open a market; it cannot manufacture demand for the platform you built to fill the opening.

The fair objection is that this is too neat — that a store with 295 million users and a court win over Google is hardly losing.⁴⁸ True, and the optionality is real: Disney put $1.5B into Epic in 2024.⁷ But notice the price. That round valued Epic at $22.5B, down from $31.5B in 2022 — a third of the company's paper worth gone while it was supposedly winning.⁷ The market is reading the same split the financials show: the engine is an asset, the store is a bet still being funded by a hit game that is no longer growing. Optionality you keep paying for is not the same as victory.

Epic Games is two companies wearing one logo. One quietly became a standard by taking a sliver of everyone else's blockbuster; the other spent years and a billion dollars trying to become a standard by giving money away, and is still waiting. The engine war was won by being unavoidable. The store war is being lost by being optional — a generous alternative to Steam that most developers simply don't need, financed by a game whose players are drifting off. Sweeney was right that it was a real plan. The trouble is the cheapest standard to own is the one you collect on. Epic built one of those. It just spent the proceeds fighting for the other kind.