According to Polygon, Ichiro Lambe, a developer who worked with Valve on improving how users find games, has declared Steam’s discoverability “broken.” In a LinkedIn post, Lambe argues that despite features like Top Sellers, New and Trending tabs, and themed sales events, most players don’t discover new titles by browsing Steam itself. He points to the “Rule of Seven,” suggesting consumers need multiple light exposures to a game before buying, something the dense Steam store page can’t provide. Lambe states the platform’s primary function is to be a checkout aisle, not a discovery engine, and that real discovery happens elsewhere. He cites data from the GameDiscoverCo newsletter showing that in 2025, only about 20% of games performed better after launching from Early Access than during it.
The Tallest Hobbit Problem
Lambe’s analogy is brutal but effective. Steam is “miles ahead of every other media platform” for discovery, but that’s “like saying they’re the tallest hobbit.” And he’s right. Look at the alternatives—the Epic Games Store, GOG, even console marketplaces. They’re arguably worse. But being the best of a bad bunch doesn’t mean the job is getting done. The core issue he identifies is behavioral: people don’t hang out on Steam to browse. They go there with intent, usually to buy something they already know about. All those algorithmic tabs and fest banners? They’re mostly preaching to the choir, or to an empty room. The real discovery, as he says, happens on YouTube, Twitch, Discord, and TikTok. Steam is just the warehouse where you go to pick up the item you saw advertised somewhere else.
Why This Is Devastating For Indies
Here’s the thing: this isn’t a problem for the *Escape From Tarkov*s of the world. Those games have massive communities and hype trains that start off-platform. The crisis is for the unknown indie dev with a “deserving game.” They’re competing in a store that adds thousands of new titles a year, relying on systems that, by Lambe’s admission, are built for checkout, not curation. The Early Access data he references is a perfect, depressing example. Everyone thinks launching to 1.0 is a second chance at a first impression, a new news cycle. But the data says nope—for 80% of games, it doesn’t move the needle. That’s a gut punch if you’ve spent years in early access building a game, hoping the full launch will be your breakout moment. It turns out the breakout moment needed to have happened long before you ever hit that “release” button.
Valve’s Incentive Is The Real Story
Lambe’s most crucial point challenges the entire premise. We assume a storefront *wants* to help us discover things. But does Valve’s incentive structure truly support that? They take a 30% cut. A sale is a sale, whether it’s *Call of Duty* or a tiny indie gem. Investing heavily in a perfect, human-centric discovery system that surfaces niche titles is astronomically hard and expensive. Why overhaul a broken system when the current one still prints money from the big hits and the players who arrive already knowing what they want? Solving discovery might even be counterproductive for them—it could divert sales from sure-thing AAA titles to riskier, cheaper indies. I’m not saying they’re actively suppressing discovery, but the business case for truly fixing it is murky. It’s a wicked problem, as detailed by outlets like GameDiscoverCo. Every “solution” creates new problems.
So What’s The Answer?
Honestly? There might not be one, at least not coming from Steam. Lambe doesn’t offer a fix, and that’s telling. The platform is too big, too crowded, and too transactional. The “answer” has already emerged, and it’s the ecosystem Lambe describes: off-platform community building. Successful games are now launched from Discord servers, built by YouTubers, and validated on streaming platforms. Steam is the endpoint. For developers, this means marketing can no longer be an afterthought you hope Steam will handle. It has to be the primary focus from day one. You’re not just making a game; you’re making content for the real discovery engines. That’s exhausting, but it’s the reality. Steam’s discovery is broken because, in a way, it was never really built to work in the first place. It was built to sell. And at that, it’s terrifyingly efficient.
