Which benchmark should I compare my launch with?

The one that matches your price, your wishlist count and the time window you measure. Unless a table says otherwise, every number on this page is first-week (or later) copies divided by the wishlist balance on launch day. Buyers who never wishlisted are in the copies, so the ratio is not the share of wishlisters who bought. How the ratio works, and how to measure your own, is in the glossary entry linked below.

Valve gives no benchmark of its own. Its documentation says that because everyone wishlists for different reasons, there is no formula to accurately predict sales from wishlists. So read each row below as the middle of a wide spread, not as a forecast.

Which published median to use
Your launchMedian to compare withSource and sample
Above $10, more than 25,000 launch wishlists0.10 in week oneGameDiscoverCo estimates, launches in the year to August 2025
$10 or less, more than 25,000 launch wishlists0.15 in week oneThe same GameDiscoverCo estimates, all prices
Fewer than 25,000 launch wishlists (from about 5,000)0.105 in week oneGameDiscoverCo estimates, launches from August to October 2024
You measure the first month, not the first week0.22 in month oneHow To Market A Game survey, launches from 2023 to 2025

Sources: GameDiscoverCo (3 December 2024, 17 October 2025); How To Market A Game (13 October 2026); Steamworks Wishlists documentation.

What are the published benchmarks by year and game size?

The median fell from about 0.20 in developer surveys of 2020 to 2024 to 0.10 to 0.15 in the estimates of 2024 and 2025. The table lists each published week-one median with its sample. All rows except one come from GameDiscoverCo. The Gamalytic row is a second estimate of public data, with a filter that removes small launches.

Read the sample column before you compare rows. The survey rows came from developers who sent their own numbers, and GameDiscoverCo says games that convert well are probably over-represented. The estimate rows cover every Steam launch above a wishlist limit, so they include games that did badly. A part of the fall from 0.20 to 0.15 is this change of method.

For a game above $10 that launches in 2026 or 2027, 0.10 is the most realistic published median. For a cheaper game, use 0.15. GameDiscoverCo says very inexpensive games convert better, but gives no separate number for them.

Published medians of first-week sales per launch wishlist on Steam
PublishedSampleGame size or priceMedian
June 2020GameDiscoverCo survey, 50 responses, games of different agesAll games0.20 (average 0.36)
June 2020The same surveyMore than 1,000 wishlists0.16
June 2020The same surveyMore than 10,000 wishlists0.20
June 2020The same surveyEarly Access launches0.16
October 2021GameDiscoverCo survey, more than 75 responses, games from 2019 to 2021All games0.20
April 2024GameDiscoverCo survey, more than 100 games launched since 2020All games0.20
April 2024The same surveyMore than 10,000 wishlists0.17
April 2024Gamalytic estimates, about 700 games released after September 2023, at least 500 copies in month oneAt least 5,000 wishlists0.22
December 2024GameDiscoverCo estimates, every launch from August to October 2024 with a Hype score above 250At least 5,000 wishlists, with a few smaller games0.105
October 2025GameDiscoverCo estimates, all launches in the year to August 2025More than 25,000 wishlists0.15
October 2025The same, without adult gamesMore than 25,000 wishlists0.14
October 2025The same, priced above $10 at launchMore than 25,000 wishlists0.10

Sources: GameDiscoverCo (15 June 2020, 6 October 2021, 15 April 2024, 3 December 2024, 17 October 2025); Gamalytic (29 April 2024).

How wide is the range around the median?

Very wide. In its estimate of every launch from August to October 2024, GameDiscoverCo found 0.019 at the 10th percentile and 0.599 at the 90th. The 25th percentile game, which ranked about 90th of 120 launches in August, converted only 0.048. Only 15 to 30 games a month passed 0.5, and some of those had only about 10,000 wishlists.

The surveys show the same shape. In the 2024 survey, single games came in as low as a tenth of the 0.17 median and as high as 10 times it. In 2020 the results ran from 0.05 to more than 1.0. So a ratio of about 0.26 or more puts a launch in the top quarter, and a ratio below about 0.05 puts it in the bottom quarter.

Week-one sales per launch wishlist by percentile, Steam launches from August to October 2024 (GameDiscoverCo estimate, chart values)
PercentileWeek-one sales per launch wishlistWhat it means
10th0.0199 in 10 launches did better
25th0.048The bottom quarter starts here
50th (median)0.105The middle launch
75th0.26The top quarter starts here
90th0.599Only 1 in 10 launches did better

Sources: GameDiscoverCo (15 June 2020, 3 December 2024, 17 October 2025).

How many copies will my wishlists sell?

Multiply your launch wishlists by the median for your price, then check a low and a high case. The example below is arithmetic, not data, and it uses the same math as our wishlist to sales calculator. It is a $15 game with a balance of 50,000 wishlists on launch day.

The calculator's low and high cases are our own band: 0.5 times and 2 times the median. The percentile rows show how far real launches spread past that band. A budget that only works in the middle case is a risk, so check the 25th percentile too.

The gross amounts are copies times the full price. They are before a launch discount, refunds, taxes and the share that Steam keeps.

  1. Take the wishlist balance on launch day: 50,000.
  2. Choose the median for the price. The game costs more than $10, so use 0.10. At $10 or less, use 0.15.
  3. Middle case: 50,000 x 0.10 = 5,000 copies in week one.
  4. Low and high case: 5,000 x 0.5 = 2,500 copies, and 5,000 x 2 = 10,000 copies.
  5. Gross sales at $15 in the middle case: 5,000 x $15 = $75,000.
Worked example: 50,000 launch wishlists at $15 (arithmetic)
CaseRatioFirst-week copiesGross at $15
Calculator low: median x 0.50.052,500$37,500
Calculator middle: median above $100.105,000$75,000
Calculator high: median x 20.2010,000$150,000
10th percentile, August to October 20240.019950$14,250
25th percentile, August to October 20240.0482,400$36,000
75th percentile, August to October 20240.2613,000$195,000
90th percentile, August to October 20240.59929,950$449,250

Sources: GameDiscoverCo (3 December 2024, 17 October 2025); our own arithmetic.

What do wishlists sell after week one?

About two to three times the first week by the end of year two. In July 2026, GameDiscoverCo estimated sales for about 5,000 Steam games launched from July 2023 to July 2025 with more than 1,000 copies in month one. At the median, gross revenue reached 1.42 times week one by the end of month one, 2.46 times by year one and 2.99 times by year two. Copies reached 1.39, 2.59 and 3.32 times.

In our example, $75,000 in week one becomes $184,500 gross by the end of year one and $224,250 by year two (our arithmetic). GameDiscoverCo found a longer tail for games at $10 or more (2.7 times week-one copies in year one), for games that sold 10,000 to 100,000 copies in week one (3.16 times) and for co-op games (3.53 times). Games under $10 reached 2.32 times.

How To Market A Game measured the same thing per launch wishlist, in a survey of 160 games. For launches from 2023 to 2025, the median game had sold 0.22 copies per launch wishlist after one month, 0.43 after six months and 0.70 after one year. For launches from 2020 to 2022 the values were 0.45, 0.82 and 0.99. Survey games are self-selected, and a month is longer than a week, so these values sit above the GameDiscoverCo week-one estimates.

Median copies sold per launch wishlist after launch, How To Market A Game survey of 160 games (chart values)
Time since launchLaunches from 2020 to 2022 (34 games)Launches from 2023 to 2025 (126 games)
1 month0.450.22
6 months0.820.43
1 year0.990.70

Sources: GameDiscoverCo (21 July 2026); How To Market A Game (13 October 2026); our own arithmetic.

What share of wishlisters buy the game?

Fewer than the launch ratio suggests, because the launch ratio also counts buyers who never wishlisted. The share of wishlisters who bought is a second number, and the published values run from about 5% to 20%.

In the 2020 GameDiscoverCo survey, the average game converted about 9% of its wishlists into purchases in 7 days, and GameDiscoverCo wrote that Valve cited 13% at that time. Among games with more than 10,000 wishlists, the top half converted 12.3% of wishlists in the first week and the bottom half 4.5%.

The Steamworks cohort report gives the lifetime share for each month of wishlists. How To Market A Game published two of them. These values keep growing while the game sells.

Published shares of wishlisters who bought the game
SourceWhat it measuresValue
GameDiscoverCo, June 2020Average share of wishlists that bought in 7 days, survey of 50About 9%
GameDiscoverCo, June 2020The same, more than 10,000 wishlists: top half and bottom half12.3% and 4.5%
How To Market A Game, January 2025Song of Iron (2021): lifetime average of the monthly cohorts, data from January 20258.02%
How To Market A Game, January 2025Song of Iron: the Steam Next Fest month, the lowest cohort5.51%
How To Market A Game, January 2025Cosmoteer: lifetime average of the monthly cohorts19.31%

Sources: GameDiscoverCo (15 June 2020); How To Market A Game (27 January 2025); Steamworks Wishlist Reporting documentation.

Has wishlist conversion fallen in recent years?

Per wishlist, yes. In How To Market A Game's October 2026 survey, launches from 2023 to 2025 sold about half as many copies per launch wishlist in the first month as launches from 2020 to 2022. Wishlists still predicted first-month copies well: the R squared on a log-log scale was 0.77 for the early era and 0.72 for the later one, and 0.73 across all games.

GameDiscoverCo saw a smaller change. In May 2024 it used a related measure (first-week reviews against its own interest score) and found a fall from just over 0.15 at the start of 2022 to about 0.125. In October 2025 it wrote that it does not think the ratios are slipping, and that wishlists are harder to get because more games compete.

So treat the medians on this page as a midpoint and plan with margin. Our guide to wishlists before launch has the 2026 wishlist tiers and the Popular Upcoming figures.

Sources: How To Market A Game (13 October 2026); GameDiscoverCo (29 May 2024, 17 October 2025).

Which launches beat the benchmark?

Cheap games, well reviewed games and games that spread through friends and creators. In October 2025, GameDiscoverCo compared top and low converters among about a year of launches, and How To Market A Game compared wishlist cohorts.

Steam helps every launch the same way: it emails or notifies wishlisters when the game releases, and later for a discount of 20% or more that lasts over 8 hours, with a cooldown of about 2 weeks between emails.

  • Price: games above $10 had a median of 0.10, against 0.15 for all prices.
  • Month: December launches converted poorly, during the Steam Winter Sale. No other month was a clear winner.
  • Reviews: the low converters that GameDiscoverCo listed had a median review score of 67% after 7 days. Its selected top converters had 91%.
  • Time on Steam: those low converters had been on Steam for 411 days on average before launch, against 214 days for the selected top converters.
  • Word of mouth: the highest ratios came from online co-op games that spread through creators and groups of friends.
  • Age of the wishlist: How To Market A Game found no sign that old wishlists convert worse. For Song of Iron, the monthly cohorts stayed near the 8.02% average, with a standard deviation of 2%.

Sources: GameDiscoverCo (17 October 2025); How To Market A Game (27 January 2025); Steamworks Wishlists documentation.

How do I measure the result?

Manalens is the analytics and growth engine built for games. Measure your own ratio one week after launch, then compare it with the row of this page that matches your launch. The steps are in the glossary entry linked at the top of this page.

Manalens helps with two parts today, with limits. The Steam UTM report import works now: upload the UTM Analytics CSV from Steamworks and see visits, wishlists and purchases per source and campaign, and import the last week again after a few days, because Steam finalizes conversions later. Steam counts only signed-in players who act within 72 hours of a click, so this is a sample of your wishlists, not all of them. The automatic Steam sync of all wishlists and sales, with a Steamworks Financial API key, is in early access. Manalens does not calculate the ratio for you yet: you divide the two numbers.

Sources: Steamworks UTM Analytics and IPartnerFinancialsService documentation; Manalens Steam documentation.

Frequently asked questions

Is 10% a good Steam wishlist conversion?

It is typical for a game above $10. GameDiscoverCo found a median of 0.10 first-week sales per wishlist for games above $10, among games that launched with more than 25,000 wishlists in the year to August 2025. Across all those launches, the median was 0.15. Results differ 10 to 20 times between games.

What ratio puts a launch in the top quarter?

About 0.26 first-week sales per launch wishlist. That was the 75th percentile in GameDiscoverCo's estimate of every Steam launch from August to October 2024 with a Hype score above 250. The 90th percentile was 0.599, and only 15 to 30 games a month converted more than 0.5.

How many copies does a wishlist sell in the first year?

For launches from 2023 to 2025, How To Market A Game found a median of 0.70 copies per launch wishlist after one year, in a survey of 160 games. GameDiscoverCo estimated that the median game sold 2.59 times its week-one copies by the end of year one, for launches from July 2023 to July 2025.

Why was the median 0.20 in older surveys?

The older numbers came from surveys. GameDiscoverCo found a median of 0.20 in 2020, in 2021 and in April 2024, from developers who sent their own data. It said that games which convert well were probably over-represented. Its later numbers are estimates for all launches above a wishlist limit, and they are lower.

Sources and implementation