If you have ever wondered why one game with a few thousand wishlists outsells another with ten times the social buzz, the answer increasingly lives in a quiet corner of every Steam page: the tags. Valve has just run its largest tag cleanup in years, adding 17 new tags, removing 28, and merging or renaming several more. It is tempting to file this under minor storefront housekeeping and move on. That would be a mistake. Tags are the raw material Steam’s recommendation engine uses to decide which players ever see your game.
Tags are the foundation upon which each of the store hubs are built.
What Valve Actually Changed
The shape of the update is worth understanding. In its official announcement, Valve added 17 tags, retired 28, and merged a handful more.
New tags lean toward how players actually describe games now: “Bullet Heaven,” “Desktop Companion,” cozy mechanics like “Organizing,” “Cleaning,” and “Decorating,” plus culture-specific genres such as “Wuxia” and “Xianxia.”
Removed tags fell into three groups. Some overlapped with clearer options (“NSFW” and “Mature” duplicated “Gore,” “Violent,” and “Sexual Content“). Some were too subjective to apply consistently (“Well-Written,” “Masterpiece“). And some pointed at specific intellectual property (“LEGO,” “Games Workshop”) that franchise pages now handle better.
Even the merges carry a lesson: “Pool” became “Billiards,” because players kept applying it to anything featuring a swimming pool.
One detail many developers miss: you are not the only author of your tags. Valve notes that tags can be applied by the developer, by players with non-limited accounts, and by Steam moderators, which means your tag set shifts over time as more people play and weigh in. That makes your store page a living document whether you tend it or not.
Why This Is Really About Discovery
Valve was explicit about the goal: the changes help players “identify the games that best fit their interests” and help “Steam generate appropriate recommendations.” Tags are not decoration. They build the “More Like This” carousel, the genre store hubs, and every player’s personalized Discovery Queue. When a tag is vague, overlapping, or simply wrong, you are handing the algorithm bad instructions about who should see your game.
The old instinct was to pile on every popular tag you could half-justify and ride a trend. Under a recommendation engine that rewards relevance, that instinct works against you. A precise, honest tag set is a targeting tool. It points Steam at the players most likely to buy your game, finish it, and tell a friend.
The Proof: tinyBuild’s 500k Lesson
Consider The King is Watching, a roguelite kingdom builder from Serbian studio Hypnohead, published by tinyBuild. Since its July 2025 launch it has sold more than 500,000 copies, one of the publisher’s biggest hits to date. Its origins were humble: a Ludum Dare 54 game jam entry that placed third and earned a publishing contract.
Now you can really find a concrete niche for your game.
The road from prototype to half a million sales was not smooth. After a year of iteration, playtest scores were solid but the time players spent in the game was not where the team wanted it. So Hypnohead and tinyBuild made a nervous call and pivoted to a roguelite structure, purely to deepen replayability. CEO Alex Nichiporchik is candid that the choice was not about chasing a trendy genre.
That replayability mattered because of how Steam now measures momentum. Concurrent players, Nichiporchik points out, track closely with how long people stay in a game, and longer sessions keep a title visible. The other half of the story is targeting. A few years ago, success ran on a familiar formula of wishlists, social buzz, and review scores. Today it runs on reaching the right audience and letting tag-driven recommendations carry the game to them.
tinyBuild leaned hard into that. Free demos on Itch.io and the Opera GX browser pulled in more than a million downloads, and a deliberate China strategy (three to four times the studio’s usual localization spend, plus localized news and a social presence) drove over 30% of total sales. None of that is a single growth hack. It is patient, deliberate store-page and audience management, and it paid off.
Do something that is really fun to play, and then pick and choose elements of genres you believe fit what you are trying to develop.
Your Discoverability Checklist
You do not need a publisher’s budget to act on this. You need an afternoon and some honesty about what your game actually is.
- Audit your Steam page now. For each tag ask: does this genuinely describe my game, or did I add it because it was trending?
- Re-align with Valve’s clearer options. Replace removed or vague tags with accurate ones, and check which tags players and moderators have added on your behalf.
- Study your real competitors. Find the games in your exact sub-genre and learn how the ones winning your niche describe themselves.
The studios that treat their store page as a living document, not a launch-day checkbox, are the ones the algorithm rewards.
OGDA runs regular jams, workshops, and meetups where Okanagan developers trade exactly this kind of hard-won craft. Find the next one at ogda.ca, and come compare notes in the OGDA Discord.