Last week I updated the benchmarks with the latest data from the February Next Fest. In today’s post I wanted to look across the raw data and make some wild ass assumptions (somewhat supported by the data) and see what it means about where SNF is headed.
The upper bounds of Next Fest is 15,000 wishlists
The brass ring of Steam Next Fest is to appear in top of the charts: Most Wishlisted, Most Played Demo, Trending.
Many of the developers of games that appeared in those charts submitted their data and gave me some background of what it is like to be there. My gut feel: if your game makes it into the bottom to middle of those top 100 charts, you are going to get around15,000 wishlists from SNF.
There are a few games that go SUPER Viral and are at the very very top of those charts and they get in the 45,000-60,000 wishlist range. There are usually only 3-5 games that reach that prize every next fest.
So my gut feel, back of the napkin, don’t quote me on it, rules for Steam Next Fest estimating is:
- 1,000 wishlists or less if your marketing is struggling.
- 2,000-3,000 wishlists for middle class games that have earned decent but not great Streamer Coverage.
- 15,000 wishlists for the big boys who are at the top of their game
- 30,000-45,000 wishlists for the breakout hits that “win” next fest. (don’t bet on doing this)
The new stasis point
I have been running a SNF survey every year since its inception so I can track how they have evolved. The following charts track the last 2 years which represent 2 big changes:
- A machine learning visibility algorithm where everyone gets baseline visibility for SNF day 1 and 2 (Started October 2024)
- The steady increase in the number of games participating in SNF
The following table tracks the changes at all levels through the years.
| Percentile | June 2024 | October 2024 | February 2025 | February 2026 |
| 30th Percentile | 384 | 420 | 584 | 382 |
| Median | 884 | 916 | 1079 | 806 |
| 70th Percentile | 2365 | 1741 | 2298 | 1839 |
| 95th Percentile | 30802 | 7800 | 10879 | 13461 |
| Highest earning game | 69040 | 18349 | 45655 | 57074 |
Here are a couple charts that show it in graph form (the second graph zooms in just on the 70th percentile and below for readability. The data in both of them comes from surveys I ran in the past of data provided by readers.
The key thing here is there is a big dip in October 2024. What happened?
To sum it up, here is my THEORY of what the October 2024 dip is:
- In June 2024, the big games (95th percentile and Max) got a WHOLE LOT of extra visibility. Valve may have thought that was too lopsided and too winner-take-all so they changed it to be more egalitarian.
- In the next Steam Next Fest (October 2024) they implemented the “first two days are open to everyone” algorithm to kind of spread things around.
- But the algorithm wasn’t tuned right and nobody got visibility. It hurt the 95p+ games too. This also coincided with the year that 3000 games flooded the October 2024 next Fest. There was a huge dip for everyone. I even wrote about this dip in my summary back then: “Yes, games underperformed compared to previous Next Fests. And, it seems the biggest reduction in visibility was towards top-performing games. Basically, the new algorithm effectively lowered the maximum potential visibility while maintaining the minimum.”
- After October 2024 Valve recalibrated the algorithm and in February 2025 and February 2026 the visibility was back up for the upper tier games. The 70th percentile and lower were pretty much the same.
I don’t know if this is the case for sure. Just a wild ass theory I have with a bit of data to back it up.
Evidence of this
This June 2024 boost and subsequent fall was brought to my attention by the guys at Butterscotch Shenanigans.
Their game Crashlands 2 (CL2) was in the June 2024 Steam Next Fest and it did great:
- Impressions from NF: 6,247,962
- Visits from NF: 38,406
- WL additions during NF for CL2: 15,987
Then their next game How Many Dudes? (HMD) was in February 2026 and did well but not quite the same as Crashlands 2.
- Impressions from NF: 645,745
- Visits from NF: 15,342
- WL Additions during NF for HMD: 14,740
Basically both games were in the top 95 percentile of Steam Next Fest (~15K wishlists) but Crashlands 2 got 10x more impressions and more than 2x as many visits.
But their wishlists basically ended up the same.
Maybe Valve saw similar behavior across all top tier games and said “that is wasted visibility that isn’t targeted enough”. Now almost 2 years later Valve has tuned the algorithm to better match players to their preferences. So fewer impressions and wishlists but BASICALLY the same number of wishlists.
Just a wild ass theory here.
Rise in the non-wishlisting players
If I track the median number of players per game you can see there are even more players in February 2026 than at any previous time. It is up and to the right. This is good!
This chart is a zoom in on the 70th percentile and below and the number of players is basically flat or even down. This isn’t great.
Why is this?
I think normie Steam gamers who only play 1-2 games per year and don’t have 400 games on their wishlist have finally discovered Steam Next Fest. This is probably driven by the rise in Co-op “Friendslop” games.
In my previous post I noticed that co-op games have the lowest ratio of plays to wishlists. Meaning a lot of people play them, but not a lot of people wishlist them. Also in February 2026 festival, the top of the charts were dominated by friend-slop games.
My wild-ass theory is that normies are pulled in to play co-op games that are at the top of the charts and leave without wishlisting.
I don’t think this is a bad thing necessarily. These people will probably buy the game when it comes out, they just don’t have a habit of wishlisting. And maybe, eventually, they will learn about the benefits of wishlisting.
It could be good. Maybe they will try more games. Maybe they will wishlist more. Maybe you will sell more games eventually.
Shadow drop a demo?
I have heard whispers that the new meta is to not launch your demo until the days leading into next fest to farm “excitement,” “velocity,” and “word of mouth.” The theory is that these wishlists are “fresher” and the Steam algorithm likes this better.
In this context, velocity means how many wishlists a game got in the 14 days before Steam Next Fest.
I don’t think it is true though. Here is why:
FIRST, if you look at the correlation between wishlists earned in Next Fest, there is still a stronger correlation to having A LOT of wishlists (chart in red) than there is towards velocity (in blue).
Velocity is good, but most of the games that went in with a lot of wishlists also had a high velocity. They had both.
Similarly, if you look across the data, games that launched their demo more than 1 month before Steam Next Fest did much better than the sub 1-month games.
Even if we look at the 10 games in my survey that were 95th percentile “Super Earners” that got 15,000+ SNF wishlists:
- Only 1 game “surprise dropped” the demo a few days before SNF.
- 5 Games had the demo out months and months before SNF
- 4 Games launched the demo 1 month before SNF.
I just think it is too risky to launch an untested product before Steam Next Fest.
Steam Next Fest as Validation
Tomas Sala is a wild man. He both announced and shipped a demo for his new boat-crafting-sandbox game Ship Shaper just days before the February Steam Next Fest. At the start of the fest it shot up and peaked at the #27 most popular demos before settling in at 60-70 rank.
He thinks that quick increase is due to the 5000 wishlists he earned from some viral TikToks right at the announcement of the game.
His strategy was to use Next Fest to test whether his game had “The Magic.” If it did, he would continue working on it. If it didn’t, I don’t know what he would have done.
But lucky for him it proved that it had the magic and he earned 11,566 wishlists during Steam Next Fest.
I am happy for Tomas and I think it is cool what he did, but I don’t think the typical developer can pull this off. Here is why
- He is a very successful dev who has built up an audience over years. His previous games are Bulwark: Falconeer Chronicles and The Falconeer: Revolution Remaster
- He has a big following of fans who like his previous games.
- He knows what it takes to ship a quality game.
- He understands Steam and the audience very well.
- He was just working on a “rapid release” idea and was ok with the downside risk
- He was lucky with some very viral Tiktoks.
- Even with all those stars aligning, he didn’t technically get into the 95th percentile 15,000 wishlist plus club.
If this is your first game and your first Steam Next Fest, do not try to shadow drop your demo. You will not have the luck that Tomas did. Your demo will not be as refined as Tomas’s either. He has a lot of experience making demos and learned a lot by refining his demo for Bullwark publicly over a number of months.
Summary
So, in conclusion I know I should say something like
OMG 3000 games were in this next fest. AI Slop. TOO many games. Steam is always changing. Everything you know is obsolete now.
- Chris in scare tactic mode
That is just fear mongering though for the sake of engagment.
But, I just don’t see that much change this Next Fest. There were more games, but overall the performance matched the years post October 2024. This is still the meta same as before
- I still think the best meta is to release your demo months and months before Steam Next Fest.
- I still think you should try some extra marketing in the weeks before SNF to get a boost in velocity.
- You should still try to get as many wishlists as you can.
- You should do the last Next Fest before you launch.
Keep your head up, you have got this.