Do Steam Demos Increase Sales? What the Data Shows

Yes, for most indie games a Steam demo increases sales, but indirectly: demos generate wishlists, and wishlists convert to launch sales at a median of roughly 12% on day one. The strongest single number in the public data comes from Chris Zukowski’s Next Fest surveys, where games that put their demo up months before the fest earned about 2.5x the median wishlists of games that shipped the demo during the fest itself.

The old fear that “a demo satisfies curiosity and kills the purchase” comes from 2013 console data and does not describe how Steam works in 2026. But demos are not free wins. Launch one too early or cut it wrong for your genre and it can cost you real sales. The rest of this post shows the survey numbers behind that 2.5x, plus the four situations where holding the demo back is the smarter play.

Almost nobody plays a demo, closes it, and buys the full game on the spot. The full game usually is not even out yet. The actual chain is:

1. 01The demo makes your game playable, which unlocks streamers, festivals, and Steam’s free-game discovery surfaces.
2. 02Those channels put your store page in front of people who were never going to find it otherwise.
3. 03Some of them wishlist the game.
4. 04Wishlists convert to sales at launch — our data on wishlist conversion rates puts the median day-one conversion around 12%.

The wishlist step is where the leverage is. In Zukowski’s 2023 analysis of demo visibility, Mob Factory’s wishlist chart was mostly flat for months of demo-less marketing, then spiked within about 3 days of uploading a demo, when the streamer Wanderbots picked it up. Dome Keeper collected roughly 40,000 wishlists in a month on the strength of a demo that streamers kept discovering. In both cases the audience mostly did not play the demo. They watched someone else play it and wishlisted.

So the honest answer to the title question: a demo will rarely convert an individual fence-sitter on the spot, and it does not need to. It turns strangers into wishlisters at scale.

The stereotype has a real source. At the 2013 DICE conference, Jesse Schell presented Xbox 360 data showing that games with a demo plus a trailer sold notably worse than games with a trailer alone. That finding got repeated for a decade as “demos cut sales.”

None of it maps to a Steam indie in 2026:

• —It measured boxed and XBLA console games in an era with no streamer economy. Nobody was wishlisting anything, because wishlists barely mattered.
• —The games studied were largely known quantities from established publishers. If shoppers already know what your game is, a demo can only subtract mystery. Indie games have the opposite problem: nobody knows they exist.
• —Steam has since rebuilt its discovery systems around demos. Valve does not build event infrastructure around a feature that suppresses sales on its platform.

If your game has a marketing budget in the millions and a known franchise name, the 2013 logic might still apply to you. For everyone else, the visibility a demo unlocks is worth far more than the mystery it spends.

In July 2024 Valve shipped a major rework of how demos behave on the store, now baked into the Steamworks demos documentation, and the design intent is hard to misread. Demos gained:

• —Their own optional store pages, so a demo can rank and be found on its own.
• —Separate user reviews, scored independently of the full game.
• —Placement in discovery surfaces like New & Trending, New on Steam, and tag pages, treated like a free game release.
• —A wishlist notification: you can email everyone who wishlisted your game the moment the demo goes live.
• —An “add to library” button, so players can grab the demo now and install later.

Two of these matter more than the rest. Demo reviews mean a rough demo now leaves a permanent, visible mark (more on that below). And the discovery placement means a demo launch is a visibility event in its own right, separate from Next Fest — shipping a demo on a quiet week can generate a traffic spike that used to require a festival slot. The mechanics of setting this up are covered in our demo best practices guide.

The best public dataset on demo outcomes is Zukowski’s post-fest developer surveys on howtomarketagame.com. His running benchmark survey covers 182 self-reported games from the February 2026 fest, and his February 2026 write-up digs into the same data. These are survey numbers from developers, not Valve data, so treat them as directional rather than exact.

The headline finding: what you earn during the fest is mostly determined by what you brought into it. Pre-fest wishlist count correlated with fest earnings at r = 0.825, the strongest relationship in the whole dataset.

In his fest-by-fest tallies the medians have stayed in a narrow band: 884 wishlists (June 2024), 916 (October 2024), 1,079 (February 2025), 806 (February 2026). Call it roughly 800 to 1,100 wishlists for the median participating game. The 95th-percentile games in February 2026 earned around 13,500, and the single biggest breakout in that survey hit 57,000.

The second finding is the one you can act on: demo timing. Games that launched their demo months before the fest earned roughly 2.5x the median wishlists of games that shadow-dropped the demo during the fest. Among the top ten earners in February 2026, five had demos live for months beforehand, four for about a month, and only one shipped days before. An early demo has time to get picked up by streamers and build the pre-fest wishlist base that the r = 0.825 correlation rewards. If you are aiming at the October fest, the October 2026 Next Fest dates and deadlines give you the working-backward schedule; the short version is that the fest runs October 19-26 and your demo build is due September 21.

Once your demo is live, Steamworks reports a demo conversion rate: players who wishlisted after playing, divided by total demo players. Zukowski’s June 2026 analysis of this metric gives the distribution from his survey data:

So a 20% conversion rate is genuinely good, and the median game sees about 1 in 5 demo players wishlist. That sounds low until you remember the mechanism from earlier: the demo’s job includes reaching all the people who watch it played and never install it.

His sharper point is that conversion rate barely predicts total outcome. His survey included games with a 30% conversion rate that earned only 1,000 wishlists, and games below 15% that earned over 15,000. Reach beats efficiency. Do not redesign your game because your play-to-wishlist ratio is 16% instead of 22%; spend that energy getting more people in front of the demo, and read our guide on how to get Steam wishlists for where that reach comes from.

Demos help on average. Your game is not an average, so here are the specific cases where the data and the postmortems say to hold off.

1. The demo is not ready. Since the 2024 update, demos carry their own review score. A janky demo now produces a visible “Mixed” label attached to your franchise before launch, and first impressions from streamers do not get a second take. If you need feedback but the build is rough, that is what a playtest is for — the playtest vs demo comparison covers exactly this decision. Playtests are gated, unreviewed, and invisible to the store. Rough build, use a playtest; polished slice, ship the demo.

2. Short narrative games with one core twist. A 2-hour story game where the demo gives away the hook can cannibalize itself, especially when Steam’s refund window means a player can finish a short game inside 2 hours anyway. Systems-driven genres (roguelikes, factory and automation games, colony sims, deckbuilders) are the opposite case: their demos work like free samples of a loop that gets better with depth, which is part of why those genres dominate Next Fest charts.

3. The demo is the wrong length. Too short and players bounce before the hook lands; too long and they play the demo instead of buying the game. The working consensus sits around 30-90 minutes of content depending on genre, and we break down the numbers in how long a Steam demo should be.

4. Launching the demo on top of your release. A demo that goes live weeks before launch competes with your own launch messaging and gives refund-window shoppers a reason to hesitate. The demo’s wishlist-building work needs months, not weeks. If you are inside 6-8 weeks of release and have no demo, the higher-value move is usually spending that time on the launch itself. (Post-launch is a separate question: leaving a demo up after release is generally fine and keeps the streamer channel open.)

To be strict about the title question, wishlists have to turn into money. The canonical numbers: median day-one wishlist conversion is about 12%, and GameDiscoverCo’s 2024-2025 conversion study found games launching with 25,000+ wishlists converted at a median of about 0.15x over the first week — roughly 15,000 first-week sales per 100,000 wishlists, with outliers far above that when word of mouth kicks in.

Run the median Next Fest outcome through that math. A fest earning you 1,000 wishlists is worth on the order of 120 extra day-one sales, before the long tail. A demo that performs at the 15,000-wishlist tier is worth around 1,800 day-one sales. You can model your own numbers, including price and Steam’s cut, with the wishlist calculator.

That is also the ceiling on the claim. A demo is a multiplier on attention, not a substitute for it. A game with 300 pre-fest wishlists and no outside traffic will most likely land near that 322-wishlist median for its tier, demo or not.

Work through these in order:

• —Is your game systems-driven (roguelike, sim, strategy, deckbuilder, factory)? Ship a demo. This is the strongest yes in the data.
• —Is it a short narrative game built on one reveal? Consider a vertical slice that stops before the hook, or skip the demo and lean on trailer plus press.
• —Is the build polished enough to review? If not, run a playtest first. Demo reviews are permanent.
• —Are you more than 3 months from launch? Good demo window. Inside 2 months, think hard before adding a demo to your launch workload.
• —Is a Next Fest on your calendar? Get the demo live at least a month before the fest starts; the 2.5x timing effect is the cheapest multiplier available. Deadlines for the October event are in our Next Fest October 2026 guide, and the Next Fest prep tool turns them into a task list.
• —Is the demo 30-90 minutes with the hook in the first 10? Trim until it is. Details in the demo length guide and the broader demo best practices post.

And before any of it goes live, make sure the page the demo points at can convert the traffic. A demo that sends 10,000 curious players to a weak store page wastes its one spike — run yours through the free Steam Page Analyzer first.