What to do after Steam Next Fest

Half Sword topped Valve's list of the most-played demos of the October 2025 Next Fest. GameDiscoverCo's read of that result was that most of its players were people coming back: the demo had run at 5,000 to 15,000 concurrent players for months before the fest, and the players had already wishlisted it. Being first by players did not mean being first by new wishlists.

That is the lesson of the week after. The headline numbers, rank and downloads, are the least useful ones. What matters is how many new people played, how long they stayed, how many wishlisted, and where they came from. For seven days strangers played your game with no one watching, which is the biggest playtest most indie games get before launch, and its data goes stale fast. Read it in the first few days, while you still remember what you changed and when.

Steamworks keeps the reports, but not the context. Before you forget, record for each day of the fest:

• Wishlist additions and deletions, from the wishlist report. Where the curve jumped, note what happened that day: a stream, a creator's video, a patch, a front-page slot.
• Where the traffic came from, from the store page's traffic breakdown. Next Fest's own pages, search, tags, and outside links each tell you something different.
• Demo players and median play time, from the demo app's reports. A short median says most players stopped early.
• Followers, because followers are public for every game and are what you will compare against.
• Reviews and forum threads on the demo, with every bug and confusion players named.

The demo's real job is the wishlist, so the ratio that matters most is new wishlists per demo player. Two demos with the same downloads can differ several times over on it.

Most of what gets written about Next Fest is about the games at the top, and the median game is nowhere near them. Two sources measure the middle:

Reviews are scarce: by the end of the October 2025 fest, only 45.5% of demos had a single review and 21.5% had ten or more, by GameDiscoverCo's count. So the reviews you got are worth more than their number suggests, and silence is information too.

Sort what players said into three piles. Bugs: anything that crashed, softlocked or broke on a machine that is not yours. Confusions: places players got lost, missed a mechanic, or asked a question the game should have answered. Wants: features players asked for. Fix the first pile before anything else, the second before launch, and treat the third as votes, not orders.

Then match the median play time to your demo's structure. If the median player quit at minute six and your hook arrives at minute ten, the problem is not the hook. If they played to the end and did not wishlist, the ending is not asking: the last screen should say what the full game adds and carry a wishlist button.

Valve's guidance is to consider leaving it up, and its reason is blunt: many players are disappointed when demos disappear the moment the fest ends. A demo that stays up keeps earning wishlists from everyone who finds your page later, through tags, creators or your own posts, and it can keep showing in New & Trending Free and the demo sections of tag pages, because demos can appear anywhere a free game can.

The case for taking it down is narrow: a demo that holds most of a short game. If a player can finish your two-hour game's best hour in the demo, deactivate it before launch. Valve's Deactivate Demo tool, under Technical Tools on the demo's landing page, stops new installs; players who have it cannot launch it.

Updates to an approved demo are not reviewed again, so you can ship the fixes from the fest to the demo the same day you ship them to your build.

The fest ends one possible launch window: a game in Next Fest could not release before the edition ended, and now it can. Whether it should depends on its wishlists.

Steam's Popular Upcoming list, which shows unreleased games in the days before launch, took about 5,000 to 7,000 wishlists to enter in January 2026, by Chris Zukowski's daily scrape. A game well under that after the fest gets little from Steam's launch lists, and the months before launch are better spent earning wishlists than shipping. A game well over it can plan its launch around the Steam sale calendar: release at least 30 days before a seasonal sale if you want it as your second wave, and set your launch discount before release, because it cannot be added afterwards.

Do not move a release closer just because the fest went well. The extra wishlists will still be there in two months; the polish you skipped will be in every review.

Each game enters one Next Fest, ever. What is left:

• Themed fests. Steam runs genre and theme fests most months, open to Coming Soon games and games with demos. Our themed festivals guide lists the 2027 dates.
• The demo wishlist email, if unspent. Releasing a demo unlocks one email to every wishlister, but only within two weeks of the demo's first release. If you released the demo for the fest, that window has probably closed; if it has not, use it.
• Creators. A demo is the easiest thing to send a streamer or YouTuber. Send it to the creators who covered games like yours during the fest, with one short clip of the best moment.
• A post about what changed. A store-page event on the fixes you made after the fest gives wishlisters and followers a reason to look again.
• The independent showcases. Showcases run their own submissions months ahead and do not count against Next Fest.

• Days 1 to 3: record wishlists by day and by source, demo players, median play time and followers.
• Days 1 to 7: answer every review and forum thread; sort feedback into bugs, confusions and wants.
• Week 2: ship the bug fixes to the demo and the game; post a store-page event about them.
• Week 2: decide whether the demo stays up, and fix its last screen so it asks for the wishlist.
• Week 3: compare your results with the benchmarks above and set a launch window that fits them.
• Week 4: plan the next beats: a themed fest, creator outreach with the fixed demo, and the sale you want as your second wave.

Flockbay is a free AI game builder on the official Godot engine. Its playtest links send your next build to testers and bring back their errors and crashes, with a clip when the tester allows it, as something your AI can replay and fix. Its Steam numbers read your followers and an estimated wishlist range for free, and your exact wishlists from your own Steamworks key, which never leaves your machine. Video studio cuts the clip for your creator emails from your real gameplay. Hosted playtests and Video studio are free in beta until 1 December 2026.