Steam Early Access to 1.0 Launch: Sales and Review Data (2026)

Only 21% of games leaving Steam Early Access in 2026 (through mid-May) earned more in their first 30 days at 1.0 than they did at Early Access launch, per GameDiscoverCo’s graduate tracking. In 2025 the figure was 20% of 225 graduates, and the median 1.0 month brought in roughly 40% of the median Early Access launch month. If you are planning your development around a big 1.0 payday, the data says your Early Access launch was your launch.

That does not make the 1.0 transition worthless. It is a real marketing beat with a second wishlist notification, a fresh shot at New & Trending, and your one clean chance to reprice. Below: what Steam actually grants at full release per the Steamworks docs, how reviews carry through graduation, the raise-or-hold pricing call, and the survivorship bias baked into every Early Access success story you have read.

GameDiscoverCo tracks every Steam game leaving Early Access that sold more than 5,000 copies (or made $10,000+ if free) in its first 30 days of EA. Two years of that tracking now exist, and the numbers are consistent:

Four out of five graduates make less money in their 1.0 month than they did in their Early Access launch month. The median gap is large: in the 2025 cohort, median 30-day revenue at 1.0 was about 40% of median 30-day EA revenue. And these games took their time getting there — the median Early Access run was 437 days in the 2025 cohort and about 15 months for 2026 graduates.

The mechanism is not mysterious. Your Early Access launch spent your accumulated wishlists and your press moment. By 1.0, most of the people who wanted the game already own it. GameDiscoverCo’s own framing: the 1.0 debut is “really another notification to wishlisters and marketing beat,” not a second launch.

The 20% who beat their EA launch share a pattern worth noticing: they either grew substantially during Early Access (News Tower, up 138% at 1.0; Mars First Logistics, up 181%; Escape the Backrooms, up 189%) or they entered EA quietly and built an audience over time. Games that exploded on day one of Early Access almost never repeat it: Supermarket Simulator’s 1.0 month was 95% below its EA launch month, Slime Rancher 2 was down 85%, and Backpack Battles sold only 13% of its lifetime units after 1.0.

I checked the Steamworks Early Access documentation on July 2, and the policy language is worth reading precisely. Early Access launches get real but reduced treatment: wishlist notifications go out, the game appears in the Early Access new releases tab, and it can hit New & Trending and top sellers if it performs. But the docs state plainly that “Early Access titles do not receive the same launch visibility as a v1.0 release,” and that once you fully release, “additional visibility areas apply.”

In practice, the 1.0 transition gets you:

• —A second wishlist notification. Everyone who wishlisted during Early Access and never bought gets an email and activity-feed entry when you release. This is the single biggest lever, and it is why your EA-period wishlist growth matters more than your EA-period sales. Our wishlist conversion data covers what conversion rates to expect from that pool.
• —Full launch placement eligibility. New & Trending and Popular New Releases treat your 1.0 as a new release. Your slot there depends on revenue velocity in the first days, so the launch-week playbook from a normal release applies — see the Popular Upcoming and New & Trending breakdown.
• —The Early Access banner comes off your page. A meaningful slice of shoppers filter out or bounce off Early Access titles. Chris Zukowski’s data (below) shows EA games also review worse, so removing the banner improves both conversion and the sentiment of incoming reviews.

What 1.0 does not get you: a reset of your Boxleiter math, a fresh algorithmic honeymoon at EA-launch scale, or featuring. Between EA launch and 1.0, your recurring visibility tool is the update round — the update visibility rounds guide covers how major updates reach owners and wishlisters during the EA period, and that same mechanism is how you warm the audience in the weeks before graduation.

Your review score does not reset at 1.0. Every review written during Early Access carries over, counts toward your overall percentage, and stays on the page with an “Early Access Review” tag. This is observable behavior on every graduated store page rather than a documented policy, but it has held for years without exception.

That carry-over cuts both ways, and the Recent/Overall split is what makes it survivable. Steam shows a Recent score (last 30 days) next to the All-time score, so a game that launched EA rough and improved shows shoppers “Recent: Very Positive” above “All: Mixed” — the recovery story, told automatically. The full label mechanics, including the 70% and 80% boundaries where shopper behavior visibly changes, are in our review score thresholds table.

Zukowski’s Early Access analysis found that games still in Early Access average a 69% positive score versus 72% for normal launches and 75% for 1.0 releases; restrict the sample to games with 100+ reviews and Early Access still trails, 78% versus 82%. Unfinished games get reviewed as unfinished games. The practical consequence: if your EA score sits in the low 70s, graduation is your best chance to push the lifetime score over 80%, because 1.0 week brings your largest wave of new reviews since EA launch and those reviewers are judging the finished product. Run your current numbers through the review score calculator to see how many positive reviews you need to cross your next label boundary before the wave arrives.

One warning from the same math: shipping a 1.0 that players consider incomplete is worse than staying in Early Access. The Steamworks docs put the expectation in writing — once the EA label is off, “the expectation from customers will be that your product is stable and delivers a complete experience.” A wave of “this is not a 1.0” reviews lands on your biggest traffic week of the year.

The docs note that “most developers have chosen to start out with lower price than their target launch price,” which makes a raise at 1.0 the default pattern Valve designed for. Documented cases go both ways:

There is no single right answer, but there is a right process. If your EA price was set low to compensate buyers for an unfinished game, a raise at 1.0 is fair and expected — announce it weeks ahead, because “price goes up at full release” is one of the few honest urgency levers you have, and it demonstrably pulls wishlist conversions forward. If you entered EA at your target price (the Baldur’s Gate 3 model, viable when your studio has trust), hold it and spend the 1.0 beat on a discount instead. The broader framework is in our Steam pricing strategy guide.

The trap is sequencing, and it is a hard Steamworks rule: you cannot run a discount within 30 days of a price increase, and that includes your 1.0 launch discount. The docs spell out your three legal orderings:

1. 01Hold your EA price through release, run the 1.0 launch discount, raise the base price after the discount ends.
2. 02Raise the price at 1.0 and skip the launch discount entirely.
3. 03Raise the price 30+ days before 1.0, then run a normal launch discount at release.

Option 1 is the popular one because it stacks the “last chance at the old price” message with a visible discount. Option 3 works if your 1.0 date is locked far in advance. Picking option 2 by accident — raising the price a week before launch and discovering your launch discount is blocked — is a known self-inflicted wound. The full rule set, including the 30-day discount cooldown, is in Steam’s discount rules for 2026.

Hades is the graduation story everyone cites, so here are its actual numbers. Supergiant sold 700,000 copies across roughly 21 months of Early Access (December 2018 on Epic, December 2019 on Steam), then sold 300,000 more in the first three days after its September 17, 2020 full release to cross one million. Nearly a third of its first million came in one 1.0 weekend — the model everyone hopes to copy.

Now the context that gets left out: Hades was Supergiant’s fourth game, from a studio with three beloved titles and an existing audience, it won Game of the Year awards within months of 1.0, and it held Overwhelmingly Positive reviews through most of Early Access. It is the far tail of the 20%.

The 2025 cohort gives a more usable picture of what “good” looks like for everyone else. News Tower turned a modest EA launch into ~170,000 units in its 1.0 debut month. Mars First Logistics launched into EA in 2023 with roughly $300,000 in its first month, then grew its 1.0 month 181% past that and reached 150,000 units total. In both cases the EA period grew the audience through updates and steady wishlist accumulation rather than spending it. Meanwhile Supermarket Simulator, one of 2024’s breakout EA launches at 3.2 million lifetime units, saw its 1.0 arrive to a 95% smaller month than its EA debut. It did not fail; its launch had simply already happened.

The dividing line in the data is not quality. It is whether the game’s audience at 1.0 was meaningfully larger than its audience at EA launch.

Every Early Access revenue comparison you have seen — including the ones in this post — is filtered through survivorship. Zukowski’s dataset makes the filter explicit: 90% of the games that entered Early Access in 2016 never shipped a 1.0. Later cohorts finish more often, but graduates are still the minority. They are, by definition, the games that earned enough during EA to justify finishing.

That filter inflates the 1.0 numbers:

Read naively, that chart says 1.0 releases out-earn normal launches by 77%. Read honestly, it says: among games good enough to collect 100+ reviews, the ones that additionally survived a year-plus of Early Access and chose to graduate did best. The three bars are three different survival filters, not three strategies you can pick between. Zukowski’s own conclusion is that Early Access is “high risk, and slightly higher reward,” and his prerequisite stands: roughly 7,000 wishlists before EA launch, because “EA Launch is the culmination of your visibility, not a catalyst.”

Apply the same skepticism to this post. The 20-21% graduation success rate is measured only among games that sold 5,000+ copies in EA month one — already a minority outcome, as our Early Access strategy guide covers. If your EA launch missed that bar, your realistic 1.0 outcome is the left side of every distribution here.

Treat 1.0 as your biggest marketing beat since EA launch and plan it like a mid-sized release, not a rebirth. Sequence it like this:

1. 01Lock the date 2-3 months out and check it against seasonal sales and Next Fest windows with the launch date planner — a 1.0 that lands during a major sale drowns.
2. 02Decide the price move now so the 30-day discount rule works for you instead of against you (see the three orderings above). If raising, announce the raise at least two weeks ahead and say the exact new price.
3. 03Ship a roadmap-closing update 4-6 weeks before 1.0 and use its visibility round to tell owners and wishlisters the date.
4. 04Prime reviews before the traffic spike. If you are within a few points of 80%, an in-game review prompt and a fixes-focused patch can move you over the Very Positive line before launch week traffic sees the label.
5. 05Run launch week like a launch. Announcement post at release, press and creator keys out a week early, and the standard day-one sequence from our launch day checklist.
6. 06Model the realistic outcome. Take your EA launch month revenue, multiply by 0.4 for the median case and 1.5-2x for the optimistic case, and run both through the revenue calculator so nobody on the team is surprised.

And before any of it, put your store page through the free Steam Page Analyzer. Your 1.0 page has one job the EA page did not: convincing the 80%+ of your wishlist that did not buy during Early Access that the finished game is worth it now. Every capsule, screenshot, and description decision gets its highest-traffic test of the game’s life during that week — it is worth the ten minutes to check.