Start from your narrowest tag's median, not the genre total
The usual advice is to size a genre by its total revenue. That answers a question you are not asking: you are not competing for the whole tag, only for the slice a game at your scale can reach.
The starting number is therefore the median revenue of the narrowest tag that honestly describes your game. Two words there are doing the work.
Narrowest means the most specific tag you would genuinely defend, and that choice moves the answer more than any modelling decision after it.
Across the 108 tags surveyed for this page, 107 of which returned data, the smallest carries 224 games and the largest 53,193, a spread of over 230 times.
Call your game an Action game and you are sizing a 53,193-game shelf; call it a Social Deduction game and you are sizing a 224-game shelf.
Median means the middle game in that tag, which is what a new entrant should plan for. Per-tag figures are on what the median game in your genre earns on Steam.
Why a genre total describes almost nobody: 28,200 times the median
A Steam tag is one of the labels on a store page, applied by the developer, by players with non-limited accounts, and by Steam moderators, and it is the unit that matters here because Steam's discovery surfaces work on tags rather than broad store genres.
Revenue inside a tag is concentrated at the very top, and more extremely than most people expect. Comparing the average revenue of the paid titles among a tag's top 20 against the median game in the same tag, across 107 sub-genres:
| Concentration ratio | Value |
|---|---|
| Lower quartile | 8,740 times |
| Median | 28,200 times |
| Upper quartile | 72,916 times |
Read the middle row literally: the average of the paid titles among a tag's twenty biggest is roughly twenty-eight thousand times what the median game in that tag earns. Horror runs to 147,758 times, and even at the lower quartile the gap is three to four orders of magnitude.
A genre total is the sum of a distribution shaped like that, which makes it very nearly a description of its twenty biggest members.
How to build the estimate from data you can actually get
Most of this list is estimated, not measured.
| Input | Where it comes from | How good it is |
|---|---|---|
| Games carrying your tag | Steam store tag browse | Exact count |
| Median revenue for the tag | Published tag-level revenue estimates | Modelled, real error bars |
| Units sold for a specific comp | Review count times a genre multiplier, from 9x for Souls-likes to about 39x for cozy audiences and 40x for strategy | Estimate, roughly 30 to 50 percent error |
| Price | Steam's own store API | Exact |
| Pre-launch demand for a comp | The game's Steam community hub follower count | Exact and public |
| Wishlist count for a comp | Nowhere | Not public, ever |
That last row is not a tooling gap. Steam publishes wishlist counts to the game's owner and nobody else, so any figure presented as another game's wishlist count is a follower count multiplied by a guess.
The arithmetic is deliberately crude, because the inputs do not support precision:
- Take the median revenue for your narrowest tag. That is your centre of gravity.
- Sanity-check it against your own comp set, built with how to pick comparable titles for your Steam game. Take the middle of that set, not the best member.
- Convert gross to net. Every published figure is gross; how much Steam takes models the developer's share at roughly 35 to 57 percent of list price once Steam's cut, VAT, refunds and discounting come out. Treat it as a planning band, not a measured rate: where you land depends mostly on how much of your sales mix sits outside the US and EU.
Worked through: a tag with a median gross estimate of about $17,000 leaves roughly $6,000 to $9,700 in hand. That is the number to hold your plan against.
What to do when the estimate disagrees with your plan
The useful outcome is a disagreement you can act on: the gap between the estimate and your plan tells you what to question first.
If the estimate is far below what you need, question the tag before the estimate. A median that looks impossibly low often means you named a broad, cheap-to-enter tag that fills with unfinished projects. Narrow it and re-read.
If the estimate is far above what you need, check whether the tag is full of large productions. A high median in a tag dominated by publisher-funded games tells you about their budgets, not your opportunity.
If your plan only works at the top of the tag, that is the finding. Beating the median by two orders of magnitude means being one of the top twenty titles. That can be a goal, but state it rather than bury it in a spreadsheet.
If your comp set genuinely sits above the tag median, name the reason (a narrower tag, a higher price, a bigger production) and plan against that instead. Where this estimate sits in the wider sequence is how to market a Steam game, whose first stage is genre-market fit.
Sources: tag-level revenue and copies estimates for 107 Steam sub-genres, pulled September 2026 and presented as modelled estimates rather than measured sales. Steam publishes per-game sales and wishlist counts only to the game's owner.
