Epic Gave Away $2.2 Billion in Games to Build a Store Where Nobody Buys

In 2024 alone, the Epic Games Store gave away 89 PC games with a combined retail value of roughly $2.2 billion, and players claimed them 595 million times.⁷ Picture that line stretching out: hundreds of millions of downloads, a free game every week, a giveaway machine running for years without pause. It is one of the most generous customer-acquisition campaigns in the history of software. It also produced a store where, five years and 187 million new users later, third-party game revenue had barely twitched.

The official story is that Epic is winning a war of attrition against Steam — building scale now, monetizing later, doing to PC gaming what every patient platform does. The real story is closer to the opposite: Epic trained 295 million people to expect games for free, and they learned the lesson. The giveaway didn't seed a buyer base. It built a coupon-clipping audience that shows up for the free thing and leaves before the cash register.

The cut that started the war

Epic launched the store in December 2018 with a genuinely aggressive weapon: a 12% revenue cut for developers, against Steam's 30%, plus a waiver on Unreal Engine licensing fees for games sold through the store.² That number was the thesis. Tim Sweeney framed the 12% as a way to offer 'super-competitive deals for partners while building an enduring and profitable store business for Epic' — and noted that after direct costs Epic nets only about 5% of the 12%, possibly 6–7% at scale.⁸ The arithmetic is the tell. A store living on a 5% net take must run on enormous volume. The whole model assumes that the people you lure in with cheap economics and free games eventually become a river of paid transactions wide enough to make a sliver pay.

So Epic spent. To pull players over, it didn't just undercut Steam on the take rate — it bought exclusives with cash advances and gave away premium games for free, week after week. The losses arrived on schedule. Court filings put EGS losses at $181M in 2019, with a projected $273M loss in 2020 and a further $139M loss in 2021.¹ This was not a rounding error in a side project. It was a deliberate nine-figure annual subsidy, defended as the cost of building the road.

The users came. The buyers didn't.

Here is the number that breaks the growth story. Between 2019 and 2024, EGS grew from 108 million to 295 million registered PC users — a 173% jump.⁶ By any user-acquisition scorecard, that is a triumph. But over the same five years, third-party revenue — the money players spend on games Epic doesn't own — grew about 1.6%, from roughly $251M to $255M.⁶ Read those two facts side by side and the whole strategy inverts. Epic nearly tripled its audience and the audience bought essentially the same amount of stuff. The new 187 million arrived, claimed their free games, and declined to open their wallets.

It got worse in the most recent data. In 2024, third-party revenue actually fell 18% year-over-year — to $255M — even as total platform spending rose to $1.09 billion.⁶⁷ Total spending up, third-party spending down: that is the signature of cannibalization, not growth. The platform's money increasingly comes from Epic's own titles. Of that $1.09B, roughly 74% derived from Fortnite and Rocket League.⁷ Strip out Epic's own games and the 'store' is a thin and shrinking marketplace bolted to the side of a hit shooter. The free-game program didn't build a buying habit. It built the opposite — a population conditioned to treat the store as a place where games are free, and to feel slightly cheated when one isn't.

Meanwhile, Steam just got bigger

The cruelest part is what the incumbent did in response. Steam didn't sit still on its 30% cut — and that flat 30% was always a half-truth. In 2018, the same year Epic announced its 12%, Steam introduced a tiered structure: 30% on the first $10M, 25% from $10M to $50M, and 20% above that, which sharply narrows the gap for exactly the large publishers Epic most wanted to poach. Then it simply kept selling games. In 2024, Steam generated $10.8 billion in revenue, up 24% year-over-year.⁷ Set that beside EGS's $255M in third-party revenue and the gap isn't closing — it's roughly 40-to-1 on the third-party number, and widening. Epic spent seven years and billions to grow a user base, and the gap to the thing it was attacking grew faster than its losses could close it.

That testimony matters because it quietly buried Epic's own promise. The company had projected annual profitability would begin in 2023. In November 2023, its own store GM admitted on the stand that the store still wasn't profitable on any basis, and that the goal remained growth.³ When the original timeline arrives and the answer is still 'we're growing,' growth has stopped being a phase and become a permanent excuse.

Isn't this just the long game working as designed?

The fair objection is that this is how platforms are built — bleed for years, achieve scale, then turn the dial. Amazon lost money for a decade; nobody calls that a defeat. And 295 million users is a real asset; a store with that reach has optionality a smaller one doesn't. The exclusivity advances, often conflated with the free-game spend, were a separate bet — Epic committed about $444M to minimum-guarantee deals in 2020 alone — and some of those will recoup.⁵ The honest version of Epic's case is that it bought distribution at a discount and the revenue follows distribution, eventually.

But the counter has a hole the data drove a truck through. Amazon's losses funded customers who bought more every year; the loss leader led. Epic's losses funded customers who bought roughly the same every year, and in 2024 less.⁶ Scale only becomes leverage if it converts. A platform whose third-party revenue is flat to falling at five times the user count has not delayed monetization — it has disproven its monetization thesis. The free games weren't priming a purchase. They were the purchase, paid by Epic, on the customer's behalf, forever. That isn't a flywheel waiting to spin. It's a treadmill that costs $2.2 billion a year to stand still on.

Epic set out to prove that a generous platform could out-compete an entrenched one by giving more away. It proved something else, and it proved it at a cost few companies could survive: that user acquisition without purchase intent is not a moat — it's a liability with a login. The 295 million didn't migrate from Steam so much as add a second free-game inbox to their week. The store got everyone's attention and almost none of their money, and you cannot build an enduring, profitable business on a sliver of revenue from people you spent billions teaching not to pay.