Almost every piece of advice a small studio hears is about making the game. Almost none of it is about the harder problem that starts the moment the game is finished: getting anyone to see it. In 2025 that gap stopped being a nuisance and became the defining constraint of the business. Roughly nineteen thousand games shipped on Steam over the year — on the order of fifty-four or fifty-five a day — and about half of them were seen by almost no one at all. Development is a solvable engineering problem; you can grind your way to a finished, competent game. Discovery is a market, and that market is saturated. This post reads the real numbers on that saturation, with every caveat the data demands, and asks the only question that matters afterward: what actually moves visibility for a game nobody is waiting for?
The wall is discovery, not development
The instinct of most first-time developers is to treat shipping as the finish line. It is closer to a starting gun fired into a stampede. The scarce resource in the modern indie market is not talent, or tools, or even funding — it is attention. There are more competent, finished, genuinely enjoyable games released every week than any player, curator, or algorithm can meaningfully sort through. That reframes the whole problem. A studio that spends its last month polishing a mechanic almost nobody will encounter has optimized the wrong variable; the same month spent building an audience before launch would have moved the needle far more. None of the numbers below are a verdict on the quality of the games in the long tail. They are a measurement of a distribution problem — supply has scaled far faster than the shelf space that gets a game in front of a buyer.
The flood, in one chart
Start with the raw supply curve, because it explains everything downstream. Steam’s annual release count has climbed from roughly 2,800 games in 2015 to about 9,700 in 2020, past 14,100 in 2023, to around 18,500 in 2024 and then north of 19,000 in 2025 (SteamDB’s year summary lists closer to 20,003, a figure it is still adjusting). At the high end that is roughly fifty-four or fifty-five new titles every single day competing for the same finite front-page real estate, the same curators, and the same recommendation slots.
| Year | Games released |
|---|---|
| 2015 | ~2,800 |
| 2020 | ~9,700 |
| 2023 | ~14,100 |
| 2024 | ~18,500 |
| 2025 | ~19,000+ (SteamDB summary ~20,003) |
One honest note before the doom sets in. The 2025 increase over 2024 was, in percentage terms, one of the smaller annual jumps since Steam Direct removed the gatekeeping in 2017 — roughly 11%, described by HowToMarketAGame as about the third-lowest percentage rise on record. The absolute numbers are enormous, but the rate of acceleration has cooled. The flood is real; it is not a runaway exponential. That distinction matters, because it means the problem is a stable, structural feature of the platform to plan around, not a moving target that makes planning pointless.
Half of last year’s releases went unseen
Supply is only half the story; the other half is how brutally attention concentrates. Of the roughly 19,000 games that launched on Steam in 2025, about 9,327 — nearly half — finished the year with fewer than ten user reviews. Around 2,229 of those got zero reviews, and roughly 7,100 landed somewhere between one and nine. At the other end, only about 6.2% of all 2025 releases cleared 500 reviews — a rough, commonly-used proxy for “broke through and found a real audience.” This is not a one-year anomaly: in 2024 the same shape held, with about 43.56% of releases under ten reviews and 7.2% over five hundred. The curve is the platform’s resting state.
| Reviews | Games | Share of ~19,000 |
|---|---|---|
| 0 reviews | ~2,229 | ~12% |
| 1–9 reviews | ~7,100 | ~37% |
| 10–499 reviews | ~8,500 (derived remainder) | ~45% |
| 500+ reviews | ~1,180 | 6.2% |
The revenue picture rhymes with the visibility one, and it is worth stating carefully rather than reaching for a scary headline number. Steam took in on the order of $16.2 billion in gross game revenue over the first eleven months of 2025, and indie titles are a huge and growing slice of that — VG Insights put indie gross at roughly $4 billion through the first three quarters of 2024, about 48% of full-game revenue, spread across the 98.9% of releases that are indie by their definition. But the money inside that slice is intensely concentrated: VG Insights reports that the largest “triple-I” studios captured about 53% of all indie revenue in 2024, and games shipped with a third-party publisher earned roughly five times the median revenue of self-published ones. A tiny share of titles takes most of the money. We are deliberately not going to quote a single “median lifetime revenue per indie game” figure — the only defensible per-game medians sit behind VG Insights’ paywalled platform, and the freely-reported numbers describe the skew, not a clean midpoint. Anyone who hands you a confident “the median indie game earns $X” without citing a purchased, dated report is guessing. The honest read is the shape: a power-law, not a bell curve, and getting seen is what determines which side of it you land on. We took the winners’ side of that same distribution apart in what 2026’s top games have in common.
Wishlists: a real signal with a very wide band
If visibility is the constraint, wishlists are the currency most studios track toward launch — they seed Steam’s launch-day visibility, trigger notification emails, and feed the algorithm early demand. The temptation is to convert them into a tidy planning constant: “X wishlists means Y sales.” The data does not support that, and the most authoritative source on it says so plainly. GameDiscoverCo’s cohort analysis puts the median first-week wishlist-to-sale conversion for games launching with more than 25,000 wishlists at roughly 0.15× — about fifteen sales in week one per hundred wishlists — and closer to 0.10× for games priced above $10. But the crucial part is the spread, not the midpoint: GameDiscoverCo states the ratio “can vary by an enormous margin — anywhere from ten times lower to twenty times higher than the median.” That is an order-of-magnitude band in each direction. Treated as a point estimate, 0.15× is a planning trap; treated as a median with a wide band around it, it is an honest expectation. We present it only that way, and never as a single number you can multiply against a forecast.
There is a subtler, more important lesson buried in the same research, and it reframes where a studio should spend its worry. GameDiscoverCo’s thesis is that conversion itself is not the problem — conversion rates have not meaningfully weakened over time. “The real obstacle,” they conclude, “is building wishlists at all.” In other words, the bottleneck is upstream: it is generating the demand in the first place, on a store page nobody has found yet. And here we hit a hard limit in the data that we will respect rather than paper over. There is no authoritative, dated figure for the “typical” or median number of wishlists a small indie game launches with. Community benchmark pages float rules of thumb, but they are undated and unsourced, so we are not going to invent a “most indies launch with N wishlists” number. If you see one quoted confidently, ask where it came from; the honest answer is usually nowhere citable.
Levers that measurably help
So what actually moves the needle? The clearest evidence points to Steam Next Fest, the recurring multi-day demo event — but the mechanism is not the one most people assume. Surveying participating developers, HowToMarketAGame found that somewhere between 68% and 88% of the wishlists a game gains during the fest come from users who never played the demo at all. That is a profound finding: the store page itself — the capsule art, the trailer, the short-description copy, the tags — does the overwhelming majority of the conversion work. The demo unlocks the fest’s visibility; it is not what converts most of the browsers who arrive. If your capsule and first two sentences do not land in a scroll-by, no amount of demo polish rescues the page.
Next Fest also has a floor and a shape worth knowing. The data suggests a rough ~2,000-wishlist threshold below which the fest gives a game little to no algorithmic lift — the event is an amplifier of existing momentum, not a launchpad from zero. There is a strong correlation between pre-fest wishlists and fest gains, and the event runs in two phases: an early roughly-48-hour equal-opportunity window, after which Steam’s recommendation engine personalizes and amplifies the early performers while quietly suppressing the weak ones. First-impression performance compounds. One lever is unusually concrete: games that were livestreamed during the fest saw on the order of two-to-four times more page visits, per developer reports — directional, but a rare case of a tactic with a measurable multiple attached.
The balancing view keeps this from becoming a Next-Fest infomercial, and it comes from Alinea Analytics: the fest’s impact is “modest even for the biggest games and minimal for anything outside the top 25 (even top 10).” The visibility, like everything else on the platform, concentrates violently at the top of the participant list. Wishlists gained can become a vanity metric if they are not the right wishlists; Alinea argues that demo playtime and demo-to-wishlist engagement predict eventual launch conversion better than raw wishlist counts do. Read the two sources together and the practical synthesis is clear: Next Fest is a genuine, measurable lever, but it rewards a page that was already working and a game that was already generating engagement. It multiplies momentum; it does not manufacture it.
Honest takeaways for a small studio
Strip the numbers down to what a small team can actually act on. First, budget for discovery as a first-class workstream, not a launch-week afterthought — if the game is finished before anyone knows it exists, the finishing was the easy 80% of the job. Second, treat the store page as the product it is: the capsule and the first two sentences do most of the conversion work, so they deserve the same iteration you give a core mechanic. Third, build wishlists early and steadily, because the bottleneck is demand generation, not conversion — and do not anchor plans to a wishlist-to-sales point estimate, because the real distribution spans an order of magnitude in each direction. Fourth, if you enter Next Fest, enter it with momentum already built (the ~2,000-wishlist floor is real) and lean into the tactics with evidence behind them, like livestreaming during the event.
And finally, read all of this without either the doom or the false hope. The long tail is not a graveyard of bad games; it is a distribution problem, and distribution problems have levers even when they do not have guarantees. The odds are honestly long — nearly half of last year’s releases ended the year nearly invisible — but the games that break through are not doing it by luck alone. They are the ones that treated being seen as a design problem from the first day, not the last. If you are also weighing how to monetize once you are seen, we wrote about drawing that line without exploiting players in free-to-play done honestly.
Sources
Every figure above traces to one of the following. Release counts are SteamDB totals read partly through reputable secondary reporting; the review-count breakdown is 80.lv / TechSpot citing Alinea and SteamDB; conversion figures are GameDiscoverCo’s medians presented with their stated order-of-magnitude band; revenue concentration is VG Insights; and discovery-lever figures are HowToMarketAGame and Alinea Analytics. Where a precise number was not reliably sourceable — a typical launch wishlist count, a single median lifetime revenue — we said so in the prose rather than inventing one.
- SteamDB — Steam releases by year — annual release trend (~2,800 in 2015 to ~19,000+ / ~20,003 in 2025); 2024 review-count breakdown (~43.56% under ten reviews, ~7.2% over five hundred); daily rate ~54–55 releases/day at the 2025 total. Primary release-count source; SteamDB retro-adjusts totals.
- 80.lv — Steam earned ~$16B in 2025 but nearly half of 19,000 games got under 10 reviews — 2025 review-count distribution (nearly half under ten reviews; ~2,200 zero; ~7,100 with 1–9; only ~6.2% over five hundred); Steam gross ~$16.2B (Jan–Nov 2025, via Alinea Analytics). Secondary reporting of Alinea / SteamDB.
- TechSpot — Nearly half of the ~19,000 games released on Steam this year went (largely unseen) — corroborating tally (19,112 releases; 9,327 titles under ten reviews; 2,229 with zero). Totals differ slightly by source and measurement date.
- GameDiscoverCo (Simon Carless) — The state of Steam wishlist conversions — median first-week wishlist-to-sale conversion ~0.15× (games >25k wishlists), ~0.10× above $10, with the explicit “ten times lower to twenty times higher than the median” variance band; thesis that building wishlists, not converting them, is the real obstacle. Presented here as a median-with-band, never a point estimate.
- VG Insights (via Game World Observer & gamedevreports) — Indie games revenue on Steam — indie gross ~$4B (Jan–Sep 2024), ~48% of full-game revenue, 98.9% of releases; triple-I studios captured ~53% of all indie revenue; games with a third-party publisher earned ~5× the median revenue of self-published ones. Per-game medians sit behind VG Insights’ paywall, so no single median-revenue figure is quoted here.
- Alinea Analytics — Steam Next Fest’s winners, and why — the balancing view that Next Fest’s impact is modest even for the biggest games and minimal outside the top 25; wishlists risk becoming a vanity metric, with demo playtime and demo-to-wishlist engagement better predictors of launch conversion.
- HowToMarketAGame (Chris Zukowski) — Making sense of the February 2026 Steam Next Fest, with the companion How many games were released in 2025? — 68–88% of fest wishlists from non-demo-players (store page/capsule/tags do the conversion work); ~2,000-wishlist floor below which the fest gives little algorithmic lift; livestreamed games ~2–4× more page visits; and 2025’s ~11% release growth being one of the smallest percentage rises since Steam Direct.
Keep reading: all posts on the WizusLabs Engineering blog.