Steam Next Fest: what actually happens before, during and after

Last updated: 18 September 2026

You get one shot at Steam Next Fest. Not one per year, not one per game per stage of development — one Next Fest per game, ever. That's Valve's rule, and it's the thing that turns "should I do Next Fest" into a real decision.

This is a developer preparation guide, not a list of free demos. The October example below shows how to work backwards from edition-specific requirements. For a future entry, use that edition's actual deadlines; do not reuse October's dates.

The Steam Next Fest October 2026 calendar

Steam Next Fest runs three times a year — February, June and October. The October 2026 edition runs 19 October to 26 October, both at 10:00 AM PDT (UTC−7). That's 19:00 CEST for the start, if you're in Europe.

After that, the next edition Valve has actually named is February 2027. The Fest runs three times a year, so a June 2027 edition is the obvious expectation — but Valve hasn't announced one, and I'd rather tell you what's confirmed than what's likely.

If you're reading this after these dates have passed, check Valve's Next Fest documentation for the current edition. The mechanics below don't change. Only the calendar does.

Valve's countdown table and Valve's calendar disagree

This one is worth a section of its own, because it's the kind of thing that eats a Fest.

The evergreen Next Fest documentation carries a general workback timeline, and it's the thing most people plan from:

Do that arithmetic against 19 October and the two submission deadlines land on 21 September and 5 October.

The October 2026 page says 14 September and 28 September.

That's five weeks and three weeks, not four and two. Both real deadlines land a full week earlier than the countdown table implies. The only line where the two agree is Press Preview: 11 days before, which is 8 October either way.

So the rule is simple and I'd write it on a sticky note: the edition page wins. The countdown table is a shape, useful for planning six months out. The dated page is the contract. If you plan from the generic table, you miss the deadline that matters by a week and find out when your build isn't in review.

I'll be honest about how I know: the earlier version of this post carried 21 September and 5 October. Those are exactly the countdown numbers. Somewhere between the table and the page, the wrong pair got written down, and it sat there for three weeks telling readers they had until October. That's the correction this update exists for.

What has to be finished before 28 September

Working backwards from that date, and using Valve's own advice to submit 3–5 business days before any deadline in case you fail review and have to resubmit:

If you can add a working buffer, submit by 23 September. That gives you one failed review and a second attempt without losing the Fest.

The one-Fest rule, and why it's a real decision

Valve's eligibility list includes a line that's easy to skim past: a title may only participate in one Next Fest. Not one per year. One, full stop.

So the question isn't "should I do Next Fest." It's which Next Fest, and Valve refuses to answer it for you. From their documentation: some developers join early to learn what players enjoy, others pick the edition closest to launch to build interest before release. Their verdict? "There's no right answer."

That's unusually honest for platform documentation, and I think it's correct. The two options do different jobs.

Early in development gets you feedback while you can still act on it. You'll find out that your tutorial is confusing, that people quit at the third level, that the mechanic you're proudest of goes unnoticed. What you won't get is a wishlist spike that's still warm at launch.

Close to launch converts attention into wishlists at the moment those wishlists matter most. But you're locked in. Whatever the demo shows is what people judge you on, and there's no second Fest to fix it in.

If you missed October and you're now looking at February 2027, that's not a loss. It's five months to make the demo better and to build something for the Fest to amplify. Which brings me to the numbers.

What the numbers actually say

This is the part most Next Fest guides skip, and it's the part that should shape your expectations.

Simon Carless at GameDiscoverCo has analysed every Next Fest since February 2023. The June 2026 edition was the biggest yet: 4,382 demos, which he notes is "66% more than the 2,645 demos in June 2025's Next Fest."

Here's what a year of that growth did to the per-game return:

• A game in the top 10% added +121 followers in June 2026, down from +163 in June 2025. About 25% less.
• A game in the top 1% added +1,330 followers, down from +1,759. Also about 25% less.

Same event, same analyst, same method, one year apart. The field grew 66% and the reward shrank a quarter.

For the median game, the most recent figure I have is from February 2026, where 3,500+ demos participated and the median game gained 11 Steam followers. Carless estimates that translates to roughly 200 wishlists. He didn't publish a median for June 2026, so I'm not going to invent one — the February number is a February number.

Where these numbers actually come from

And I want to flag how all of these are built, because it matters more than the digits. Carless measures followers, which are publicly visible, and converts to wishlists with a rough multiplier. He hedges it himself: "perhaps 200 wishlists", "probably ~3k wishlists?" — question mark his. Treat every wishlist figure in this space as an order of magnitude, not a measurement. Anyone quoting you a precise Next Fest wishlist median is quoting an estimate of an estimate.

Carless's own summary of the dynamic is the sentence I'd read before registering:

"it's tough to do as well as games used to, if you're in the top 10% of games in Next Fest - because duh, there's lots more games"

Existing interest may help, and players may also discover a game during the event. The cases and estimates here do not establish a guaranteed result or prove that starting without an audience makes participation pointless.

Which means the work that decides your Next Fest outcome mostly happens before Next Fest. Your store page, your capsule art, your tags — those convert the traffic Valve sends you. The Fest is a traffic event, not a conversion event.

Two smaller findings from the same June 2026 analysis, both interesting and neither actionable: Godot's share of Next Fest demos rose from 9.2% to 12.6% while Unity and Unreal both dipped, and the share of demos with Valve's AI content disclosure enabled went from 750 (21.2%) in February to 1,163 (26.5%) in June. Carless adds the caveat himself: that tickbox "could cover a multitude of usages."

Are you even eligible?

The full criteria, from Valve — Next Fest is open to upcoming games that meet all of the following:

What Valve does not require

There are no requirements regarding exclusivity, newness, or length of demo. Valve says that in as many words. An old demo you've already had live is fine. A fifteen-minute demo is fine.

There's no requirement about your release date, as long as it's after the Fest. You can be two years out.

Early Access and free-to-play games are eligible, as long as the game hasn't released yet. Entering Early Access counts as releasing.

One wording trap: Steam distinguishes the demo app from the full game app, and registration happens from the base game, not the demo. For most titles it appears automatically on your App Landing Page, and you should see a green checkmark once it's saved.

Steam Playtest and Next Fest: pick one

Worth its own heading, because the answer is not the one people expect.

A Playtest doesn't affect your eligibility at all. Valve is explicit about that. But they follow it immediately with a recommendation against running both:

"we do not recommend having a Playtest live during Steam Next Fest, as this splits your players between the two builds instead of concentrating them in one spot. As a demo is required for Next Fest, we would recommend pushing all your users towards your demo during the event."

The logic holds up. Next Fest's entire mechanic is concentration: a week where Steam pushes a lot of people at a lot of demos, and where the back half of your visibility depends on whether the people who tried yours kept playing. Splitting that group across two builds halves the signal you're trying to generate, and gives you two sets of feedback about two different things.

The two features do different jobs anyway. A Playtest is for finding out whether the game works, with a group you can control the size of. A demo at Next Fest is for showing a finished slice to strangers. The Playtest decision is worth making deliberately, and it's mostly a question of what you still need to learn — but if you're in the October Fest, the sequencing part is answered: run the Playtest before, not during.

How visibility inside Steam Next Fest actually works

This is the most useful mechanical detail in the whole documentation, and it's buried.

The opening days are largely randomised. Valve's wording: the first few days are "completely randomized in all locations except for explicitly named lists and personalized segments, which are based on the user's play behavior." After a few days, logged-in users get carousels personalised by an algorithm running on behaviour during the Fest.

That exception clause is doing more work than it looks. Carless, watching the June 2026 edition, noted that an "exploration mode" toggle made the randomisation explicit until Wednesday — and added that even before Wednesday, "a chunk of the page is non-randomized, btw, inc. recommended tags & discovery queue."

So: the front of the Fest is a lottery for the first few days, and it's by design. Don't read your day-one numbers as a verdict. But the surfaces that run on your tags keep working the whole time, which is an argument for auditing them, not against.

Then, once personalisation kicks in, it runs on behaviour during the Fest. Which means the thing that determines your back-half performance is whether people who tried your demo kept playing it. Not your marketing. The demo.

Your tags decide where you appear

When you register you pick one or two primary categories. Valve combines those with your base game's Steam tags to sort you into sections. With 4,382 participating games in the June 2026 edition, that sorting is most of what decides whether anyone sees you at all.

And the failure mode, from Valve's tips page: if you don't select any categories, your game will not appear in Next Fest at all. Registered, demo built, review passed, invisible.

If your game isn't showing up where you expect once the event starts, Valve's own checklist is: confirm the game is registered and has one or two categories selected, confirm your demo is actually released (only games with released demos appear), and compare your store tags against the sections you see in the Developer Preview.

That Developer Preview Page goes live about five weeks before the event and shows exactly where you've landed. Check it. If you're in a section that doesn't fit, you still have time to adjust your tags — and tags are adjustable in a way that almost nothing else about your Fest placement is.

Build review, and the buffer nobody leaves themselves

Two pieces of Valve's advice I'd underline, because both of them cost people Fests.

Submit 3–5 business days before your actual deadline. That's Valve's own number. If you fail build review the first time, you need room to fix and resubmit. There is no grace period, and "I submitted on the day" is how people end up watching the Fest from outside.

Build review is not QA. Valve is explicit: it validates that the game has the features you promised and that it runs, and "will not catch or even look for most bugs." Passing review tells you nothing about whether your demo is good. Do your own testing.

Two related things worth knowing. You can update your demo build after passing review, and Valve won't re-review it — convenient, and a good way to ship a game-breaking bug on day one if you're careless. And you get to send the demo-release notification email to your wishlisters once, within 14 days of your demo launching. It's separate from Next Fest, so if you used it months ago, there's no second one waiting for you at the Fest.

Press Preview, honestly

Press Preview is a sneak peek of Steam Next Fest that Valve sends to selected press outlets. It starts 11 days before the Fest — 8 October for this edition — and gives them an early look at the full list of registered games.

• All registered games are included, regardless of demo status. If you're cancelling, do it before Press Preview starts.
• Press get no special access. They see what the public sees. If your demo isn't live, they can't play it.
• Releasing your demo early for press releases it for everyone. There's no press-only build.

If you provided contact details on the registration form, Valve shares those with press.

Now the reality check. Journalist Pierre-Hugues Meyer, writing about the February 2026 Fest, reported receiving emails about 157 different Next Fest games — the first on 28 January, the last on 25 February. That's one journalist's inbox, not a statistic, but it tells you what the room looks like. Being in Press Preview is not the same as being covered.

Valve also points out something worth sitting with if you're worried about spending your launch excitement early: demos don't get the launch visibility a full game does. Releasing early for press costs you less than it feels like it does.

During the Fest

The short version: your demo does the work now.

Put a feedback link on your demo's main menu. A button, right there, or in the pause menu, or both. It can point at a Steam sub-forum, your Discord, a bug tracker, wherever. You're paying a week of visibility for feedback — make it trivially easy to give.

Say whether you'll deactivate the demo, and when. Plenty of people download demos meaning to play them later. If yours is going away, tell them so they prioritise it. Valve suggests leaving it up beyond the Fest, and I'd agree.

Make sure the demo button is on your store page. Under Edit Store Page → Special Settings, with "display above purchase options" ticked so it's prominent. If you haven't been in that part of your store page recently, now's the time.

Set the demo live at least 30 minutes before the Fest starts. Valve's minimum recommendation, and it gives you time to notice something's broken while you can still fix it.

Livestreaming is optional, and there's a dedicated section where players watch live streams from participating games. Since the October 2024 edition there's no official Valve livestream any more, and developer streams sit in their own tab. If you don't have the time or the temperament, skip it and put the energy into the demo. If you do stream, test the setup well in advance — broadcasting a demo works differently from broadcasting a regular game, and it takes a few attempts.

If you're reading older guides: the Fest was reorganised in October 2024, away from the genre tabs. Advice written before then describes an event that no longer looks the way it's described.

After the Fest

The Wrap-Up page featuring the most-played demos launches on 26 October, the day the Fest ends. If you land on it, you'll get an email.

Beyond that: you now have data you didn't have before. How many people installed the demo, how long they played, where they stopped, what they said in your feedback channel. In my experience that's the part people underuse. The wishlist number is the headline, but the playtime distribution is the thing that tells you whether your game is working.

And if the numbers were small — which, statistically, they will be for most participants — that's information too, not a verdict on the game. It usually means the audience-building work hadn't happened yet. That work is still available to you. The Fest isn't.

Should you do it at all?

Yes, if you're going to release on Steam anyway. It costs you nothing but preparation time, and even a randomised week in front of Steam's browsing audience is more exposure than most indie games get on any given week.

But go in with the right expectation. Next Fest is not a launch strategy. It's an amplifier, and amplifiers need a signal. Spend the months before it building something to amplify — a store page that converts, tags that put you in the right sections, a demo that people finish — and pick the edition where those are actually ready.

The October submission deadlines below have passed. If you took part, review demo feedback and your own reporting. If you did not, check the next eligible edition before choosing where to spend your entry.

For a future edition, choose a date that fits your demo and release plans. Check the current edition-specific registration and review deadlines; a February event does not mean you can wait until February to register.

FAQ