Steam Marketing Snapshot
Methodology
How the comparable games are chosen
The candidate pool comes from Steam's own More Like This recommendations (labelled close fits) plus a pre-computed dataset of the top games sharing your most distinctive community tags (labelled genre leaders).
Each candidate is scored on weighted tag and genre similarity, price model, release state and audience scale. Generic tags like Indie or Singleplayer are downweighted so they cannot dominate the match.
A relevance floor runs first and every game that clears it is kept; when fewer than five clear it, the most relevant of the rest top the list up. The final rank is relevance scaled by at most 15 percent for audience size, so popularity can break a near-tie and nothing more.
Where the marketing evidence comes from
Live YouTube and Reddit searches run at report time and are cached for 24 hours. These two sources are the only ones that feed the channel scores.
A second pass over TikTok, X and Instagram starts once the report has landed. It adds examples to the report and never changes a score. Discord invites are read from the public Steam pages and checked for public member counts.
Every result must pass a title-collision filter before it is attached to a game. A Steam AppID link is enough on its own. A distinctive title is enough by itself; a title that is also an everyday phrase (Death's Door, Lost Ark) needs a second signal such as the developer's name, a matching gameplay tag, or a channel or subreddit named after the game. On TikTok, X and Instagram every title needs that second signal.
Piracy and deal subreddits are dropped, and channels that disclosed a paid sponsorship are left off the creator outreach list.
The wishlist estimate
Only a game's developer can see its wishlist count, in Steamworks. For unreleased games the tool shows an estimate: the public Steam follower count multiplied by 9.6, the documented pre-launch median. Released games get no estimate, because the ratio stops holding after launch.
