June's Steam Next Fest was the biggest yet: 4,382 playable demos, up around 66% on the same event a year earlier (gamediscover.co). For players, that's a nice problem to have, a whole week of free games to try. For a studio, it's much harder. The crowd you're standing in got bigger, and less players are paying you attention.
That's the real story of game discovery in mid 2026. The platforms have built better ways to put the right game in front of the right player. But getting shown and getting played are two different wins, and the gap between them is where most demos quietly lose people.
PS. big shout out to gamediscover.co for their invaluable insights across the games industry.
More demos fighting for attention
The record levels of demos in Next Fest is the visible half of a much bigger squeeze. More than 19,000 games launched on Steam in 2025, and nearly half of them earned fewer than 10 reviews. Only about 1,200 passed 500 (SteamDB, via PC Gamer).
Next Fest shows the same curve in miniature. This June, a game in the top 10% of the event added around 120 new followers, down from about 160 a year earlier. Even the top 1% slipped, from roughly 1,760 followers to 1,330 (gamediscover.co). The event still works. It's just that each demo now splits a fixed pool of attention with thousands more.
Where discovery actually comes from now
A few surfaces do most of the heavy lifting in 2026, and knowing which ones matters more than chasing every channel at once.
- Wishlist momentum, seeded early. Steam's systems reward games that already have some, and Next Fest is the clearest example. The event gives every demo a baseline of visibility for the first couple of days, then from around day three the algorithm starts rewarding the top performers and quietly demoting everyone else. Chris Zukowski's tracking (howtomarketagame.com) shows demos getting near-identical impressions on days one and two, then sharply diverging on day three based on who's being played and wishlisted. Titles that got their community playing on day one build a head start that's hard to overtake.
- Steam's new homepage and Personal Calendar. The refreshed store home now carries a personalised calendar that suggests upcoming games based on what you and similar players actually play. Because it's on by default, it surfaces smaller titles without them needing huge pre-launch numbers. The gains can be dramatic: gamediscover.co documented one title, Solarpunk, going from a few hundred wishlists a day to 14,000, then 22,000, then 35,000 on consecutive days after the redesign, and a ~1,000-wishlist horror game, Karjala, doubling its total in two days before release. For a game without a marketing budget, it may be the most important discovery shift of the year.
- The demo as its own discovery surface. A demo isn't only a Next Fest entry ticket. It's a second presence on the store and a live stream of player feedback, a way to be found and to learn whether the game holds up at the same time.
Getting seen isn't the same as getting played
Every one of those surfaces shares a catch: they get you seen. None of them tell you what happens after the player clicks.
A wishlist records that someone liked your trailer. The Personal Calendar records that your game fits a player's taste. A Next Fest chart shows people opened your demo. All useful. None of them tell you what kept players engaged once they were actually got playing, and that's the number a successful launch relies on.
That gap plays out in the first few minutes. A confusing menu, a slow intro, a tutorial that skips the one thing a player needed: none of it shows up in a wishlist count. You only see it when you watch someone play your opening cold, telling you what they think as they go, where they hesitate, where they go quiet, where they lean in.
Plan for October, not just June
The next Next Fest runs October 19 to 26, with demos due for review in early October. Wishlists compound, so the studios that do well in the autumn are seeding their communities and testing their opening now, not in the final week before submission.
The work that pays off isn't cramming more polish into the trailer. It's making sure the first five minutes of the demo earn the attention the platform sends you. In a field of several thousand demos, a second chance is rare.
Two jobs, not one
Discovery in mid-2026 is really two jobs. Gaining visibility is the one the platforms keep getting better at; the calendar and the recommendation models are increasingly on your side. Getting the first five minutes of playtime is still entirely on you and will make or break your game, no matter how great your marketing is.
Get only the first right and you've bought a bigger audience for a demo that may not be ready for them. If you're prepping for October, the most useful thing you can do between now and then is see your game played through a first-time player's eyes, before the crowd does.