You get one Next Fest, and what it pays is set before you enter

There is not much time left before the October Next Fest, registration closes at the start of September. The one thing you need to know before you decide is this, a game can enter Steam Next Fest only once in its life. There are three a year, February, June and October, but you get one shot. Registering does not spend that shot, what spends it is being in the fest. Until the fest starts you can go back to the registration page and withdraw, and that is the only flexibility you have.

I put Metal Thunder into the October 2024 SNF with 300 wishlists. The registration deadline was 3 September and I opened the page a day before that. Between building the game in 1 month, 300 WL was all I could collect. The game came out of the fest with 3300 wishlists and made the top 10% of games by wishlists gained. For a 1 month old game that counted as a success back then.

It was also the first run of the new format. Before that there was a format with livestreams, and whoever packed the most marketing into the fewest days won. The ones who piled their marketing into the 1-2 days before the fest would take something like 50-100k WL depending on the size of their budget and fly off. In the new format every game gets equal visibility and the game is tested against the algorithm. If you did relatively well at SNF you usually behave similarly at launch, so it turned into a kind of soft launch.

This is why the first 3 days are the ones that matter. On day one everyone gets the same share of visibility, on day 2 that share is adjusted according to how day one went, and on day 3 it becomes clear whether you are going to win or lose at SNF. Slop games are eliminated by day 2 anyway, and if you are one of those games you more or less have to make the game again.

In those three days your game is not even being played. The ratio I have seen in the games we put into SNF is this, if 3 people wishlist, only one of those three plays the demo. They probably put off playing the demo to the following week, but there is no good way to measure that, it is pure assumption. What I can measure is this, the fest is not testing your game, it is testing your page and your appeal.

This format has a problem of its own. There is a maximum WL you can win, they do not say it openly but it runs entirely on ratios. You usually land between 1.5-2x, the logic being that if you go in with 10k you come out with 15-20k. Zukowski’s survey of 182 developers who entered the February 2026 fest came out similar, the correlation between the wishlists you have going into the fest and the wishlists you gain from it is 0.825. The same survey also found that games going in with under 2,000 wishlists got almost no support from the algorithm.

The real issue is that the multiplier is shrinking too. Turning that 300 into 3300 was an 11x for me, and if I went in with the same 300 today I probably would not even see 1000 wishlists. The format did not change, the number and the quality of the games went up with every fest.

According to GameDiscoverCo, 2,645 demos entered the fest in June 2025 and 4,382 entered in June 2026, a 66% rise in one year. In the same year the followers gained by both the top 1% and the top 10% fell by 25%. According to Zukowski, Valve meanwhile tuned the algorithm to match the player more accurately instead of growing raw impressions, so total impressions fell while the wishlist result of the games at the top stayed similar. The part crushed by the crowd is the bottom and the middle segment.

Zukowski’s survey gives you today’s picture. The median game takes 806 wishlists out of the fest, the bottom third stays under even 400, the middle segment with good streamer coverage sits in the 2,000-3,000 band, the top 5% passes 13,000 and the games that win the fest start at 15,000. So in a fest 4,382 games entered, what the game in the middle gets is 806 wishlists. And the 3,000 I got in 2024 with a top 10% finish is the upper end of the middle segment in today’s picture. Set your own expectation against this picture before you enter, and remember you get to run this experiment once.

Next Fest is not a promotion tool, it is a multiplier. Multiply zero and it stays zero, and you can only use that multiplier once. It may not matter as much as it used to for the big ones with big budgets, but for small games this is still bread and butter. That leaves one decision, when you are going to spend that one shot.

One of the biggest problems is entering the nearest fest the moment the demo is ready. That is close to throwing Next Fest in the bin. My demo is ready, why would I wait for the next fest, you will say. Because the right question is not am I ready, it is which fest is closest to my release. If you go in ten months before release the wishlists you win go cold before launch, and the same game going in two or three months out carries the same list much warmer. I did exactly this in 2024, opened the page on the last day of registration and entered the nearest fest, and back then it worked. Someone doing the same today disappears inside that 806 median.

The three do not have the same character either. June is the busiest, in 2026 it broke a six year record with 4,382 demos, but against that the Summer Sale starts three days after the fest ends, so a buying window opens right in front of the wishlists you collected. February is in the middle, in 2026 it passed 3,500 demos and was 19% busier than the October before it. October is the quietest of the three, it sits right before the holiday season and it is the fest for people planning a Q4 release. Every fest from here will get busier, GenAI has a hand in that, but those games are eliminated by day 2 anyway and players start being shown games they will like, so do not get too hung up on the raw demo count.

Still, since visibility is handed out proportionally, the crowd is the direct enemy of the small game. The first rule is the fest closest to your release, that does not change. If you are stuck between two fests and you do not have a big list, pick October.

Do not leave your marketing to the last day of the fest either. The moment the fest starts your announcement drops into the same box as the announcements of every other demo, and you disappear in there. Publish the demo 1.5-2 months before the fest and start the promotion in that gap, you should already be talked about by the time you enter. I got the demo out two days before the fest, back then it saved me, today it would not. Do not release right behind the fest either, a lot of games come out the moment SNF ends and if you did not make the top 50 at the fest you disappear into that crowd, aim for at least 3 weeks later. And do not take the demo down when the fest ends, leave it up until release.

Count backwards from your release. Set a rough release date, mark the nearest fest before it, and write down today how many wishlists you will enter that fest with. Register today as well, because registering spends nothing, and if that number is still under 2,000 in September you withdraw and save it for the next one, but do not forget to actually withdraw. You have to send the demo build for review at least two weeks before the fest, four weeks if you want to be in the press preview, and registration closes seven or eight weeks before the fest. Last time, having sent the page for review was enough to register. Even if you miss the deadline you can write to Valve up to a week later and talk them into it. And do go through Zukowski’s SNF benchmarks, he is pretty consistent on this. The October fest is on 19-26 October and registration closes at the start of September, so you have two weeks left.

If you have not opened your page yet, or you have never put the game in front of players, SNF should be the last of your worries. First you need to do your general marketing beats and see where the buyer for this game is, because that is what the fest is going to measure anyway.

Next Fest does not get anyone discovered, that was a while ago, right now it only grows what has already been seen. The marketing you have to do before you enter matters more than the fest itself, and if you have not done it, do not spend your one shot.