Steam doesn’t run one algorithm, and Valve’s visibility documentation is blunter about the inputs than most guides are. Purchases and playtime are the signal Valve names as strong; tags decide which shelves and “similar games” lists you can appear on; supported languages decide who gets shown your capsule. The same page states that store page traffic, store page conversion rate and wishlists are not factors in algorithmic visibility, and that review score stops counting once you are above 40% positive. I re-read all of it on September 15, 2026.
That leaves a large gap between what Valve publishes and what developers watch happen in their own dashboards. This post takes each signal in turn, marks which column it belongs in, and says what to change first.
What Valve states and what is inferred
Most writing about the Steam algorithm, including earlier versions of this post, blends two different things: rules Valve publishes and patterns developers infer from Steamworks. Here is the split, checked against Valve’s visibility documentation, tag documentation and update visibility rounds on September 15, 2026.
| Signal | What Valve states | What is inferred |
|---|---|---|
| Purchases and playtime | “Are people buying and playing your game? This is a strong signal to Steam that there is interest in your game” | Nothing. This is the documented core |
| Tags | The top 20 “determine the tag pages and similar browse views that your game can appear on” and “impact which games Steam thinks are similar to yours”. “The tags given the most weight govern your visibility more than those given less” | Which specific tags pay best, and by how much |
| Languages | “Steam will be more likely to show your game to players that speak a language supported by your game” | The size of the effect |
| Store page traffic | “The traffic landing on your store page is not a factor in your Steam store visibility” | Nothing to infer. This is a flat no |
| Store page conversion | “The conversion rate of users landing on your store page and making a purchase is not a factor in visibility on Steam” | Nothing to infer. This is a flat no |
| Wishlists | “With a few exceptions like the Popular Upcoming tab, wishlists are not a factor in your game’s algorithmic visibility on Steam” | That velocity rather than balance drives the Popular Upcoming cut |
| Review score | “As long as your game’s reviews are Mixed or above (40%+), review score is not a factor in algorithmic visibility” | Every threshold above 40%, including the 95% and 80% label breakpoints |
| Update rounds | Each product “will start with five rounds”; a round runs “up to 30 days from the date it started or cap out at 1M impressions on the homepage, whichever comes first”; “Additional rounds may be granted to products that are selling well” | How Valve decides what counts as selling well |
| Capsule click-through rate | Nothing. Valve publishes no CTR threshold anywhere in the partner docs | Every band in the next section |
Read the right column as a working hypothesis you can act on. The left column is the part you can cite.
How does Steam discoverability work?
Discoverability on Steam is not one queue you either make or miss. It is several systems running in parallel, each picking games on different inputs, and understanding which one you’re optimizing for changes what actions you take.
The Discovery Queue shows players a personalized stream based on their play history and browsing behavior, and for many indie games it is the largest single source of organic traffic. Valve does not publish the selection rule, so treat any list of its inputs, including this post’s old one, as inference. Our breakdown of discovery queue mechanics covers what is observable from outside.
Search results work differently. When a player types “roguelike deckbuilder,” the tags on your page are what make you eligible to be that result at all, which is the part Valve does document. How the eligible set is then ordered is not published. Good tagging strategy matters here more than anywhere else.
Then there’s the “More Like This” section on every store page, and this one Valve states outright: your top 20 tags “impact which games Steam thinks are similar to yours”. Tag yourself accurately and you land next to the games your buyers already own. Our Steam page optimization guide ranks the fixes by how often 405 analyzed pages get flagged for each.
The front page and featured sections are a mix. Some are hand-curated by Valve. Others are algorithmically selected based on trending data and personalization. You can’t really optimize for curation, but you can optimize for the algorithmic parts.
Steam click-through rate (CTR): why it’s everything
One correction before the numbers, because this section used to overstate its case. Valve publishes no click-through threshold and never names CTR as an input. What Valve does name is purchases and playtime. So the honest mechanism is indirect: a capsule that earns clicks earns store page visits, visits are where purchases come from, and purchases are the documented signal. Clicks are the tap that fills the funnel rather than a ranking lever of their own.
That still makes CTR the first thing to fix, because everything downstream is capped by it. A page nobody clicks cannot sell to anybody.
What the numbers actually mean
These bands are community observation from developer-shared Steamworks screenshots, not a Valve-published rule. Treat them as orientation:
- Below 2% is where most indie capsules sit.
- 2-4% is ordinary.
- 4-7% is where developers report the traffic curve visibly bending upward.
- Above 7% is rare outside of anticipated releases and strong-hook genres.
Nobody has published a controlled test behind those bands, ours included. They are the shape of what developers report, and the reason they still matter is arithmetic rather than algorithmic: doubling the share of impressions that become visits doubles the traffic reaching the page that does the selling. For CTR data by genre and where to find your own number in Steamworks, see our Steam CTR benchmarks guide.
These bands grade wishlists on a Coming Soon page. If your game is released, read the purchase ranges in the table above rather than entering purchases here.
Enter your impressions, visits, and wishlists to see a verdict.
Aggregated Steamworks traffic screenshots shared by developers, plus our own store page reviews. No published sample size, as of March 2026. See the table above.
How to actually improve it
CTR comes down to one thing: your capsule image. Players make sub-second decisions based on that tiny rectangle. Everything else on your page is irrelevant until someone clicks. For a deep dive on what makes capsules convert, see our capsule design guide. You can also run your images through the Capsule Validator to check dimensions and formatting before uploading.
What tends to work: cleaner and bolder compositions, removing text from the capsule (or keeping it minimal), higher contrast, and showing what makes your game distinct from the hundred other games in your genre. I’ve seen devs boost their CTR by 2-3 percentage points just by simplifying an over-busy capsule.
Test different capsule versions during Steam events. The traffic spikes make A/B testing actually meaningful.
Tags and discoverability
Tags are the one discovery input Valve documents in detail, which makes this the section with the least guesswork in it. Valve requires at least 5 tags before launch and recommends up to 20. Only the top 20 are visible and used, they “determine the tag pages and similar browse views that your game can appear on”, and ordering is weighted: “The tags given the most weight govern your visibility more than those given less.” Valve also says your “top 5 tags should paint a fairly clear picture of your game as those tags will also be used to describe your game.”
Players apply tags too, and those count. If players consistently tag your game with something you didn’t include, add it. If they dispute tags you set, reconsider those choices.
Two numbers from our own corpus say where the work actually is. We pulled the live community tag cloud from 764 consented public Steam store pages on September 14, 2026. 599 of them (78.4%) already sit at the 20-tag display ceiling and only 45 (5.9%) show fewer than 15. Filling the slots is close to a solved problem.
Ordering is not. Across the 763 pages carrying at least five tags, 651 of 3,815 first-five slots (17.1%) hold a word the page already prints as a Steam genre, and 61.9% of pages spend at least one of those five that way. Valve says the top five describe your game. A page whose top five are Action, Adventure, Indie, Casual and Singleplayer has described the store, not the game. The per-genre tables are in our tag study, and the Tag Optimizer works on your own five. For the selection strategy behind them, see our Steam tagging guide.
One rule survives from the old version of this section: do not spam popular tags that do not fit. Tagging a turn-based RPG as “Action” puts your capsule in front of people who want something else, and Valve’s similar-games lists will then pair you with games your buyers are not looking for.
Wishlist velocity
This is the claim that needed the biggest correction. Valve’s visibility doc states that “with a few exceptions like the Popular Upcoming tab, wishlists are not a factor in your game’s algorithmic visibility on Steam.” So wishlist velocity is not a general ranking input. It buys you one specific, documented placement, and that placement is worth chasing because it sits on the front page. Our Popular Upcoming breakdown covers how the cut works.
What wishlists do unambiguously is fire a notification. Valve’s wishlist documentation says every user holding your game on release gets “an e-mail and/or mobile push notification about the game being available”, and a discount of 20% or more does the same later. That is a mailing list you build in advance, not a ranking score.
The typical pattern is spikes during events and campaigns, then gradual decline between them. That’s normal. Don’t panic when you see the graph crater after a Steam Next Fest or marketing push. The key is planning regular activities that create new spikes rather than betting everything on a single moment. Our Steam Next Fest checklist covers how to maximize one of the biggest spike opportunities available.
Events, marketing pushes, content announcements, and press/streamer coverage all drive wishlists. Spacing these out gives you multiple velocity bumps instead of one big one that fades. And your Coming Soon page guide setup matters here too, because every impression you get during a spike needs to convert into a wishlist.
For context on what wishlist numbers actually translate to in terms of launch sales, check our data on wishlist conversion rates.
Review scores and how they affect visibility
Valve documents exactly one threshold here, and it is much lower than the ladder this section used to print: “As long as your game’s reviews are Mixed or above (40%+), review score is not a factor in algorithmic visibility.” Below 40% you are suppressed. Above it, going from 72% to 94% buys you no extra impressions from the algorithm.
That does not make the label ladder useless, it moves where it acts. Overwhelmingly Positive (95%+), Very Positive (80%+ across 50 or more reviews), Mostly Positive (70%+) and Mixed (40-69%) are what a buyer reads on your page after the capsule earned the click. The thresholds themselves are community-documented rather than published by Valve; the full table with the review-count minimums is in our review score thresholds guide.
So the practical framing is: clear 40% or the algorithm stops helping, then treat everything above that as a conversion lever aimed at humans.
Steam weights recent reviews more heavily than lifetime averages. If your game launched rough but you’ve fixed it, your recent review sentiment can pull your visibility back up. For detailed strategies on getting there, our review management guide breaks down what actually moves the needle.
Update activity signals
This is the one section where Valve gives you a mechanism with numbers in it. Update Visibility Rounds are a deliberate grant, not an ambient signal: each product “will start with five rounds”, a round runs “up to 30 days from the date it started or cap out at 1M impressions on the homepage, whichever comes first”, and “additional rounds may be granted to products that are selling well.” Valve ties the trigger to substance, describing major updates, new content, new DLC and new features as the moment to use one.
So the actionable read is that you hold five homepage placements and spend them, rather than earning drips of visibility by patching often. Our guide on Steam Update Visibility Rounds covers how to time them. Regular patches still matter for the reason the documented signal gives: they keep people playing, and playtime is what Valve names.
Your Steam Community Hub is one of the most effective channels for maintaining these activity signals - announcements, discussion engagement, and patch notes all contribute. And if you want to amplify your reach further, getting Steam curators to feature your game creates additional external signals that the algorithm picks up on.
This ties back into the feedback loop. Active games keep players engaged, which keeps review sentiment fresh, which keeps the algorithm happy.
Conversion rate as a quality signal
It is not one, and this section used to say it was. Valve’s visibility doc closes the question twice over: “The traffic landing on your store page is not a factor in your Steam store visibility” and “the conversion rate of users landing on your store page and making a purchase is not a factor in visibility on Steam.”
Your page still decides your revenue, it just does it one step later. Visits are not the signal, purchases are, and the page is what turns one into the other. A page converting at 3% instead of 12% produces a quarter of the purchases from the same traffic, and purchases are the thing Valve says it watches. The lever is real. The mechanism is not the one this post used to describe. Maybe your capsule implies a different game than what your screenshots show. Maybe your trailer doesn’t represent actual gameplay (our trailer best practices guide can help with that). Maybe your description is confusing people. Use the Screenshot Checker to make sure your visual assets are working as hard as they should.
The fix is always alignment. Make sure every element of your page tells the same story about what your game is. If you want to see where your own page stands, run the free analyzer below and get the specific issues dragging your conversion down.
The Steam algorithm for new games
Valve gives every new release a floor: the doc describes a New Releases Queue that carries titles regardless of how they are selling, and the algorithmic sections after that (New and Trending, top sellers, personalized recommendations) are earned by the purchases and playtime that follow. That is the whole documented story for a new game.
What is inferred, and worth acting on anyway, is that the first week compounds. Purchases in week one produce the signal that earns the placements that produce more purchases, so a flat opening has nothing to feed the loop with. Nothing in Valve’s documentation says a weak launch permanently caps you.
To give yourself the best shot: build wishlists before launch (10,000+ is the commonly recommended target), coordinate your marketing push for day one, make sure your page is fully optimized, prepare for technical issues, and engage heavily with early players. Planning your launch timing also matters. Launching into a crowded week can dilute your window.
Pricing plays into this too. Your pricing strategy affects both conversion rate and revenue signals during that critical first week.
What to actually do with all this
Knowing how the algorithm works is useful. Doing something about it is what matters. Here’s how I’d prioritize if I were starting from scratch.
First, audit your tags, because this is the only lever on the list Valve documents. Are the top five specific to your game rather than to its genre? 61.9% of the 764 pages we pulled on September 14, 2026 spend at least one of those five on a word their page already prints as a genre. Fix the order before you touch anything else.
Second, fix your capsule. It is not a ranking signal and Valve never says it is, but it decides how much of your impression volume reaches the page that sells. Run it through the Capsule Validator and compare it to top-performing games in your genre on the Steam Page Leaderboard.
Third, keep people playing and keep shipping. Purchases and playtime are the documented signal, and your five update visibility rounds are the documented way to spend a major release on homepage impressions.
Beyond that: build wishlists as a launch-day mailing list rather than as a ranking score, clear the 40% review floor and then treat the label as a conversion lever, and use Steam events as testing opportunities for page experiments. Work the Store Page Checklist after each round of page changes so none of these quietly drift.
None of this is secret knowledge. But the devs who actually do all of it consistently are a small minority, and they’re the ones whose games keep showing up.
Frequently asked questions
How does Steam discoverability work?
Steam runs several systems in parallel rather than one ranking algorithm, and Valve names the inputs it uses: whether people are buying and playing your game, your top 20 tags (which decide the tag pages and similar-games lists you can appear on), and your supported languages. Valve’s visibility documentation explicitly rules out three things developers commonly optimize for: store page traffic, store page conversion rate, and wishlists outside the Popular Upcoming tab. Review score counts only as a floor at 40% positive. Everything else you read about Steam discoverability, including the click-through bands on this page, is inference from developer dashboards.
How does the Steam algorithm treat new games?
New releases get a documented floor rather than a boost you can win: Valve carries them in a New Releases Queue regardless of sales, and the sections after that (New and Trending, top sellers, personalized recommendations) are earned by purchases and playtime. So a launch week matters because purchases compound into the placements that drive more purchases, not because Steam grades your opening and files the result. Nothing Valve publishes says a slow start caps you permanently, and a title that starts selling later starts being recommended later.
How long does it take for Steam’s algorithm to respond to store page changes?
Most changes take effect within a few days. CTR shifts from a new capsule usually show up within 48-72 hours as Steam gathers enough impression data. Tag changes can affect discovery queue placement almost immediately, but search ranking shifts may take longer.
Can I reset my algorithm performance if my game launched poorly?
Not entirely, but you can recover. Steam weights recent data more heavily than historical data, so sustained improvements to your capsule, reviews, and update activity will gradually shift your algorithmic standing. A major update combined with a discount event is one of the most effective reset strategies. Check our review management guide for tactics on turning sentiment around.
Do wishlists still matter after launch?
Yes, but differently. Pre-launch, wishlist velocity affects “Popular Upcoming” placement. Post-launch, your existing wishlist becomes a conversion tool since Steam notifies wishlisters about sales and updates. The wishlist conversion rates post has data on what percentage of wishlisters typically buy and when.
Does the algorithm treat Early Access games differently?
Steam treats Early Access launches as launches for algorithmic purposes. You get the same initial visibility window, and your performance during that window sets your baseline. The main difference is that Early Access games can get a second visibility bump at 1.0 launch, so treat both moments as separate optimization opportunities. Our Steam Early Access strategy guide covers how to maximize both the EA launch and the 1.0 transition.
Want to see where your store page stands across all these signals? Run it through our Steam Page Analyzer for a detailed breakdown of your CTR potential, tag coverage, and conversion signals. Then use the Tag Optimizer and Screenshot Checker to fix what’s dragging your score down.
Browse our genre-specific optimization guides for strategies tailored to your game type, and check the Steam Page Leaderboard to see how top games optimize their store pages.