Only 45 of 764 Steam store pages show fewer than 15 community tags. That is 5.9%, measured by pulling the live tag cloud off every page in our consented public set on September 14, 2026. The advice you have read a hundred times, fill your tag slots, is already followed here: 599 of the 764 pages (78.4%) sit at the 20-tag display cap, and the median page shows 20.
Which makes the count the boring half of the story. The interesting half is the order. Valve’s Steamworks tag documentation says “Your title’s top 5 tags should paint a fairly clear picture of your game as those tags will also be used to describe your game.” On 61.9% of these pages, at least one of those five slots holds a word already printed on the page as a Steam genre, which tells a shopper nothing the genre line did not.
Five numbers from the sample, every one of them from the same September 14 pull:
- 764 pages, 392 distinct tags, median 20 tags per page, mean 19.03.
- 45 pages (5.9%) under 15 tags. Pre-launch pages are worse: 7.7% of the 391 coming-soon pages against 4.0% of the 373 released ones.
- 651 of 3,815 first-five slots (17.1%) hold a genre word, across the 763 pages carrying at least five tags.
- 242 pages (31.7%) open with a tag that is one of their own Steam genres. The single most-weighted slot on the page, spent restating the category header.
- Singleplayer is the most common tag in the corpus, on 550 pages (72.0%), and its median position is slot 14.
How I counted the tags
The sample is the consented public set behind our leaderboard: apps someone ran through the free analyzer, still flagged public by their owner, with a resolved Steam name. One row per Steam App ID, Steam type game only. That is 785 apps. I requested each store page once on September 14, 2026 at one request per 1.5 seconds, and 21 pages did not come back (age gate, region block or a delisting), leaving 764.
Tags came off the live page, not out of our database, and that distinction is the whole reason this post exists. Until we fixed it on July 26, our stored tag column merged three different things into one array: community tags, Steam genres, and Steam feature declarations like Single-player. The rubric bug writeup has the damage report. Any tag count read from a capture before that date is padded by a median of 11 entries, so the only honest way to publish tag numbers was to re-pull every page today. The script that does it and the per-page rows it wrote are both committed, so anyone can re-run the count.
Two limits come from Steam itself. The page shows at most 20 tags, so 20 is a ceiling rather than a count of what a developer applied. And tags move: Valve notes that “tags and their relative weights may change over time as members of the community apply them to your game”, so a released page carries the developer’s sorting plus the voting.
Almost every page is already at the cap
Valve requires at least 5 tags before launch and recommends adding up to 20. In this sample exactly one page sits under five, and 94.1% carry 15 or more. If you are hunting for a page-quality problem worth a morning, empty tag slots is probably not it.
The one place the gap is still real is before release. Coming-soon pages run 7.7% under 15 against 4.0% for released pages, and 72.6% at the cap against 84.5%. That difference is what you would expect from community voting doing the work after launch, which is also a warning: the pre-launch page is the one where your sorting is the only signal, and it is the one most likely to be short.
If you have not looked at your own tag list since you wrote the page, this is the cheap check.
The first five slots are where pages differ
Valve gives two reasons the front of the list matters. “The tags given the most weight govern your visibility more than those given less”, and “Some store filters prioritize the first 15 tags, so be sure that these are sorted in order of relevance to your title.” Chris Zukowski’s Steam 101 tagging guide, written in 2020 and still the clearest version of the advice, calls the broad ones “empty calorie tags” and tells you to uncheck Indie because shoppers will add it anyway.
The corpus agrees with him, and you can see it in where the generic tags sit:
| Tag | Pages carrying it | Pages with it in the first five |
|---|---|---|
| Singleplayer | 550 (72.0%) | 48 (6.3%) |
| Indie | 349 (45.7%) | 42 (5.5%) |
| 3D | 267 (34.9%) | 27 (3.5%) |
| 2D | 241 (31.5%) | 26 (3.4%) |
| Multiplayer | 212 (27.7%) | 94 (12.3%) |
| Early Access | 111 (14.5%) | 52 (6.8%) |
Singleplayer is on nearly three quarters of these pages and almost never near the front, which is the right instinct. Early Access is the odd row: 50 pages open with it, spending the most-weighted slot on a business model that Steam already prints as a banner across the page.
A sixth of the first five repeats the genre line
Here is the pattern the sorting advice misses. A tag can be specific and still be redundant, because Steam already shows your genres on the same page, a few pixels away. I counted a first-five tag as a genre word when its text matches one of that page’s own Steam genre strings.
Across the 763 pages with five or more tags, 651 of 3,815 first-five slots (17.1%) are genre words. 291 pages (38.1%) keep all five clean. 326 (42.7%) give up one slot, 117 (15.3%) give up two, and 29 give up three or more.
One slot out of five is not a catastrophe. It is the easiest one on the page to win back: the fix is a drag in the Tag Wizard, not new art.
What the genres look like from the inside
Same 764 pages, split on Steam’s own genre strings. A game usually carries several genres, so these cohorts overlap and the rows do not sum.
| Steam genre | Pages | Under 15 tags | At the 20 cap | Most common first-five tag |
|---|---|---|---|---|
| Indie | 516 | 3.9% | 80.4% | Casual (14.0%) |
| Action | 338 | 2.1% | 85.2% | Action (33.7%) |
| Adventure | 298 | 2.7% | 86.9% | Adventure (29.2%) |
| Casual | 246 | 7.3% | 75.2% | Casual (35.4%) |
| Simulation | 222 | 4.5% | 80.6% | Simulation (40.5%) |
| Strategy | 207 | 5.8% | 77.3% | Strategy (49.3%) |
| RPG | 147 | 2.0% | 91.2% | RPG (34.0%) |
Six of the seven cohorts have the genre’s own name as their most common first-five tag, and Strategy is the worst of them at 49.3%. Indie is the exception, and only because developers have already been told to drop that one.
The runner-up in each cohort is more useful than the winner. After the genre word, Adventure pages put Exploration in their first five 22.1% of the time and Story Rich 14.1%. Strategy pages put Roguelike there 19.3% of the time, Deckbuilding 15.5%, Card Game 13.0%. Casual pages put Incremental there 14.2% and Idler 13.0%. Those are the tags a shopper filters by when they already know they want the genre, and they are how your page reaches Steam search and the discovery queue instead of sitting in a category of forty thousand.
What to change on your page this week
Four checks, in the order this data says they pay:
- Read your first five tags out loud and delete any that repeat a genre shown on your own page. 61.9% of the 763 pages with five or more tags have at least one, 19.1% have two or more.
- Check what your first tag is. 242 of 764 pages (31.7%) open with their own genre, and 50 open with Early Access. The first slot is the one Valve says carries the most weight.
- Count your tags only if you are pre-launch. 7.7% of the 391 coming-soon pages are under 15 against 4.0% of the 373 released ones, and before launch nobody is voting them up for you.
- Replace each removed slot with the tag your closest comparables share and you are missing. That is exactly what the tag optimizer compares, and the tag strategy guide covers how to choose between a broad tag and a narrow one.
What this data cannot tell you
- It says nothing about wishlists or sales. There are no outcomes in this dataset. A page with clean tags may be selling badly for reasons no tag fixes.
- It is a self-selected sample. These are pages whose developers went looking for a store page audit, so if anything the tag hygiene here is better than Steam-wide.
- It is one snapshot. Tags move with community voting. A page that had a genre word at slot three in September may not in December.
- Genre-word matching is literal. A tag counts only when its text matches a genre on the same page, so “Action Roguelike” on an Action game is not counted.
Frequently asked questions
How many tags can a Steam game have? Valve’s documentation says at least 5 are required before launch and recommends adding up to 20, and that “only the top 20 are visible to users and used by Steam to influence visibility of your title in the store”. Everything past 20 is invisible on the page.
Do the first five Steam tags matter more than the rest? Valve states that the tags given the most weight govern visibility more than those given less, that some store filters prioritize the first 15, and that the top 5 will be used to describe the game. It has never published a weighting formula, so treat the ordering as directional rather than a known multiplier.
Can I control my tag order? You sort them in the Tag Wizard, and that sorting is the starting order. After that, Valve notes the weights change as players apply tags, so a released page drifts. Re-checking every month or two is the practical answer.
Is it bad to have the Indie tag? It is on 45.7% of the pages here and almost never in the first five (5.5%). Zukowski’s tags to watch out for argues shoppers will add it for you, so spending a top slot on it is the waste, not carrying it at all.
Where do these numbers come from? A live pull of the community tag cloud from 764 Steam store pages on September 14, 2026, listed one row per page in the study file committed with this post. Free to cite with the sample size and date attached.