Valve has rewired one of the busiest real-estate zones on Steam’s homepage. On September 29, 2026, Valve announced that the Discounts & Events area, which used to rely on manual, calendar-based placements, is moving toward dynamic, personalized recommendations, and that it had spent the previous few months testing the change in place of some curated placements. Valve reported that the test exposed customers to roughly 10 times more games per day in that section. Valve has not set a firm release date, and secondary sources report a broader rollout in early 2027, so for now this is a test and a planned change rather than a finished rewrite of how developers get discovered. The test also increased store-page visits, wishlists, and cart additions. Steam discovery is described as several independent systems, including the Discovery Queue, New Releases Queue, New & Trending, tag pages, search and browse views, not one universal algorithm. Valve’s documented system emphasizes customer response, especially purchases and playtime.
Don't miss new tech stories on Google
Add Tech Insider once in the Google app and our stories appear in your news suggestions.
What Changed on Steam’s Storefront in September 2026
The shift was first documented by 80.lv on September 17, 2026, and the report was updated five days later as more developers chimed in with their own traffic numbers. The Discounts & Events grid sits on the Steam homepage, a spot that historically leaned hard on whatever was selling the most copies that week. According to the reporting, Valve has replaced or significantly diluted that bestseller-first logic with a feed that looks more like a personalized recommendation engine, similar in spirit to what streaming and music services show each individual account.
Smaller and older titles, games that would previously never crack a merchandising slot dominated by new AAA releases, started showing up in the grid for players whose history matched the game’s audience, and the shift reportedly hit featured grid positions specifically, leaving Valve’s manually curated Daily Deals slots untouched for the time being, according to 80.lv. The practical result: one indie title tracked by 80.lv logged 7,355 grid visits over six months, with 7,341 of them arriving in a single recent month once the personalized placements kicked in, a surge from a section of the homepage that had been close to irrelevant for anyone without a top-ten seller. Steam did not publish a developer blog post, a Steamworks documentation update, or a changelog entry describing the new logic when it rolled out in September. That silence is itself part of the story, and it is consistent with how Valve has rolled out algorithmic changes in the past: test quietly, watch the data, and only confirm after outside observers have already pieced it together.
Personalization Signals: Playtime, Wishlists, and Purchase History
Coverage from GamesRadar, published between September 25 and 27, 2026, described the redesigned feed as spending less space on identical best-seller tiles and devoting more slots to recommendations tailored to an individual account’s playtime, wishlist contents, and purchase history, and a separate September 2026 report from TechTimes confirmed the underlying recommendation model leans on purchase history, wishlist data, and signals drawn from the behavior of similar players, not playtime alone. That lines up with what Steam has been quietly building for years: a recommendation stack that already powers the Discovery Queue and the “More Like This” shelves on individual game pages. What is new in September 2026 is applying that same personalization logic to a homepage section that used to function as a flat, one-size-fits-all merchandising board.
Simon Carless, founder of the developer analytics firm GameDiscoverCo, was among the first industry voices to flag the behavior change as meaningful rather than cosmetic, according to the 80.lv report, and GameDiscoverCo went on to report that Valve itself acknowledged the personalized placements were driving measurable increases in store visits, wishlist additions, and cart additions compared with the old curated grid. The distinction matters because Valve runs constant behind-the-scenes experiments on its storefront, and most of them never surface publicly. This one did, because the traffic swing was large enough for developers to notice on their own Steamworks dashboards.
Why This Follows Steam’s June 2026 Store Home Overhaul
This is not Valve’s first swing at the homepage in 2026. On June 4, 2026, Steam pushed its biggest Store Home redesign in years to every user, closing out a beta that had run since April. That overhaul added hover-triggered micro-trailers, a personalized release calendar, and a wider layout built around data pulled from Steam’s own hardware telemetry, including signals from Steam Deck usage. The headline consequence for developers at the time was not the visual refresh, it was that the effective wishlist count needed to appear in Steam’s “Popular Upcoming” section jumped from roughly 7,000 to around 80,000, according to reporting cited by Inkl. That single change reset expectations for what a “successful” pre-launch wishlist campaign looked like almost overnight.
The September change targets a different shelf, but the pattern is the same: Valve adjusts a discovery mechanism that indie studios have learned to optimize around, and does so without warning. Where June 2026 made the front page harder to reach for upcoming games with modest wishlist counts, September 2026 appears to be doing the opposite for already-released games sitting in the Discounts & Events area, surfacing more of them to more relevant audiences instead of funneling everyone toward the same bestseller list. Developers who spent the summer recalibrating marketing plans around the June wishlist threshold are now recalibrating again around a discovery surface that, by design, behaves differently for every player.
Steam’s Scale: Why a Homepage Tweak Moves the Whole Market
Any change Valve makes to Steam’s front page carries outsized weight because of the platform’s sheer size relative to every other PC storefront. Steam hit an all-time concurrent-user record of 42,042,778 players online at once on January 11, 2026, according to SteamDB tracking, a number reached without a single blockbuster launch driving it. Valve’s own hardware and software survey data, cross-referenced with analyst estimates from Alinea Analytics, puts Steam’s monthly active user base at roughly 132 million to 147 million depending on the measurement window. First-half 2026 gross revenue on the platform is estimated at $11.1 billion, up 14.5% year over year.
The catalog Valve is trying to sort through has also kept growing. Steam’s storefront listed 135,003 games as of July 18, 2026, and roughly 20,000 new titles launched on the platform in 2025 alone, up from 18,477 in 2024, an average of more than 50 new releases hitting the store every single day. Every one of those games is competing for the same limited set of homepage slots, which is exactly why a tweak to a single grid can decide whether a small studio’s launch month is a success story or a rounding error.
| Metric | Figure | Period / Source |
|---|---|---|
| Peak concurrent users | 42,042,778 | January 11, 2026, SteamDB all-time record |
| Estimated monthly active users | 132M-147M | 2026, Valve/Alinea Analytics estimates |
| H1 2026 gross revenue | $11.1 billion (+14.5% YoY) | Alinea Analytics estimate |
| Total games in catalog | 135,003 | As of July 18, 2026 |
| New games launched in 2025 | ~20,000 (~55/day) | Full-year 2025 |
| Share of PC digital distribution revenue | 72%-75% | 2026 analyst estimates |
| Wishlist threshold for “Popular Upcoming” shelf | ~7,000 to ~80,000 | Changed June 4, 2026 redesign |
How Steam’s Discovery Model Compares to Epic, GOG, and itch.io
Steam’s move toward a personalized grid puts more daylight between it and its closest PC storefront competitors, none of which run anything close to the same scale of algorithmic merchandising. Epic Games Store leans on curated weekly free games and manual editorial picks rather than an always-on recommendation feed, and its discovery model has stayed largely unchanged even as Epic disclosed a record 78 million monthly active users in December 2025 at its State of Unreal event on June 17, 2026. Epic’s 2025 full-year revenue came in at roughly $1.16 billion, up 6% year over year, a fraction of what Valve pulls in over a single half.
GOG, which holds an estimated 2% to 3% of the PC digital distribution market, has built its identity around human curation and a DRM-free promise rather than algorithmic personalization, and it has shown no public sign of following Steam’s approach. itch.io sits even further from Steam’s model: it is an open, low-friction upload platform where discovery leans on creator pages, curated collections, and community tagging rather than a recommendation engine reacting to purchase history. None of these platforms face the same discovery problem Steam has, because none of them carry anywhere near 135,000 competing store pages.
| Platform | Discovery Model | Scale Signal |
|---|---|---|
| Steam | Algorithmic personalization (Discovery Queue, homepage grids, now Discounts & Events) | 135,003 games; 72-75% market share |
| Epic Games Store | Curated weekly free games, manual editorial features | 78M peak MAU (Dec. 2025) |
| GOG | Human curation, DRM-free positioning, no algorithmic feed | ~2-3% market share |
| itch.io | Creator pages, community tags, curated collections | Open-upload long-tail catalog |
The Developer Reaction: Opportunity Mixed With Unease
The reaction from the developer side has leaned positive so far, tempered by the same complaint that follows every unannounced Valve change: nobody outside Valve knows exactly how the system weighs its inputs. The GamesRadar report described the CEO of publisher Hooded Horse characterizing Steam as being on an extended run of platform improvements that have measurably helped developers, a framing that lines up with what smaller studios have been reporting on social media since mid-September, and with 80.lv’s own reporting that the personalized system gives smaller and older games substantially more exposure than the bestseller-first grid it replaced. For a developer whose game has been sitting quietly in the back catalog for a year or two, a personalized surface that resurfaces the title to players with a matching playtime history is close to free marketing, something that previously required a paid Discovery Queue placement or a discount deep enough to catch a curator’s eye.
The unease comes from the same place it always does with Valve: opacity. Valve’s Steamworks documentation does say that customer response, especially purchases and playtime, drives visibility, and that tags decide relevance and eligibility for particular surfaces. How those signals are weighed against each other, and what a developer can do to influence their odds of appearing in the grid, is still hard to see from the outside. That leaves marketing decisions, and in many cases a studio’s entire revenue forecast, dependent on a system nobody outside Valve can fully audit. Valve holds the levers, and the rest of the industry often finds out how they work by watching store-page analytics after the fact.
Market Impact: Marketing Budgets and Platform Concentration Risk
For an industry that spends heavily on pre-launch wishlist campaigns, paid Discovery Queue slots, and influencer keys, a discovery mechanism that rewards organic playtime and purchase-history matching could shift where marketing dollars go. If personalized placement genuinely drives thousands of extra visits without a media buy, smaller studios have an incentive to redirect some of that budget toward retention and post-launch content instead of pure pre-launch hype, since an engaged existing player base now appears to feed future visibility on the storefront itself.
There is a broader concentration-risk angle here too. With Steam controlling an estimated 72% to 75% of PC digital distribution revenue, any single algorithmic decision Valve makes has more influence over which games succeed commercially than a comparable decision at any other storefront could ever have. Release-week concurrent-user data from late September illustrates the gap in visibility blockbusters already enjoy versus everything else: EA Sports FC 27 peaked at 95,985 concurrent players on September 27, 2026, while a mid-tier release like Graveyard Keeper 2 peaked at 60,538 concurrent players the same day. A personalization layer that nudges even a fraction of that traffic toward smaller titles matching a player’s history is one of the only mechanisms available to counterbalance the gravitational pull of the biggest releases, which is exactly why developers are paying close attention to a change Valve has so far described only as an experiment ahead of a planned rollout in early 2027.
Historical Context: Steam’s Long Run of Silent Algorithm Changes
Valve has a well-established pattern of adjusting Steam’s discovery mechanics without formal announcements, then confirming or clarifying only after the community and press have already reverse-engineered the behavior. The Discovery Queue itself, introduced years ago, was built on an interest-based recommendation model that Valve refined repeatedly without much public documentation. The tagging system, curator program, and “More Like This” widgets all evolved the same way: quiet rollout, community observation, occasional after-the-fact confirmation. The Steam Direct submission fee, a flat $100 per product that is recoupable once a game earns enough revenue, was one of the few discovery-adjacent policies Valve did announce clearly, largely because it replaced the far more controversial Greenlight voting system.
The September 2026 Discounts & Events personalization fits that pattern. Valve announced it in a Steamworks post on September 29, 2026, as a test with a broader rollout expected in early 2027, and it rewards different behavior from developers than the old curated layout did. That consistency is arguably the most predictable thing about Steam’s discovery system: the mechanics change often, but the communication style around those changes almost never does.
What This Means for Indie Developers Right Now
For a studio currently planning a Steam launch or trying to extend the life of an older release, the practical takeaway is that pure wishlist-chasing is no longer the only lever worth pulling. A game with strong retention and a loyal, if small, player base may now get more homepage exposure than a game with a large wishlist count but weak post-launch engagement, since the personalization signals reportedly include playtime and purchase history rather than wishlist volume alone. That is a meaningful pivot from the logic Valve pushed just three months earlier with the “Popular Upcoming” threshold hike, which was entirely about pre-launch wishlist accumulation.
Developers building marketing plans for Q4 2026 and into 2027 should treat both changes as evidence that Valve is actively experimenting with how much weight pre-launch hype versus post-launch engagement should carry in the platform’s discovery math, and the timeline is no longer purely speculative: TechTimes and GameReactor both reported in September 2026 that Valve had paused scheduling new curated Daily, Midweek, and Weekend Deals, with plans to replace those manual placements with the algorithmic system in early 2027, though no firm rollout date has been confirmed. Studios that have leaned entirely on wishlist campaigns may need to budget more attention toward content updates, community engagement, and retention mechanics that keep players actively logging playtime, since that appears to be one of the signals now feeding a homepage grid that used to be reserved almost exclusively for the platform’s biggest sellers.
Predictions: Where Steam’s Discovery System Goes From Here
- Valve will face developer pressure for documentation. As more studios notice traffic swings they cannot explain, expect renewed calls for Steamworks documentation detailing how the Discounts & Events personalization actually weighs its signals.
- Rivals will experiment with their own personalized surfaces. Epic Games Store and GOG have relied on curation for years, but Steam’s traffic results could push at least one competitor to test algorithmic merchandising of its own during 2027.
- A cottage industry of algorithm-watching will grow. Expect analytics firms like GameDiscoverCo to build tracking tools aimed at inferring which signals, such as playtime, wishlist age, or purchase recency, carry the most weight, mirroring how SEO agencies reverse-engineer search rankings.
- Regulatory scrutiny of Steam’s market power could intensify. With Steam holding an estimated 72% to 75% of PC digital distribution revenue and now exercising more granular algorithmic control over commercial outcomes, antitrust and platform-fairness questions are likely to resurface in industry and regulatory discussions through 2027.
- Back-catalog titles could see a modest revenue bump. Older games that match a player’s established genre preferences and playtime history stand to benefit from repeated exposure in a personalized grid, a dynamic that favors sustained post-launch support over one-time launch hype.
The Bigger Picture: Discovery Is Steam’s Real Product
Steam’s core business has never really been payment processing or game hosting, both of which are commodity functions any storefront can replicate. Its real product is discovery: the ability to take a catalog of more than 135,000 titles and surface the right handful to each of its tens of millions of daily visitors. Every adjustment to that system, whether it is a wishlist threshold on an upcoming-games shelf or a personalization layer on a discounts grid, is Valve tuning the single mechanism that decides which developers get paid and which ones do not. The September 2026 change is not dramatic in isolation, but stacked against June’s wishlist reset, it confirms that Valve is treating 2026 as a year of active experimentation with how Steam decides what players see first, and the rest of the PC gaming industry is once again adjusting its strategy around a system it can observe but not control.
Timeline: How the Discoverability Story Unfolded
The sequence of events matters because it shows how little of this came from Valve directly, at least at the start. On June 4, 2026, Steam shipped its Store Home redesign, which included the wishlist threshold hike for the “Popular Upcoming” shelf. Sometime in the weeks before September 17, 2026, the Discounts & Events grid began behaving differently, though no exact rollout date has been confirmed. On September 17, 80.lv published the first detailed report describing the personalized feed and its effect on smaller developers’ traffic. That report was updated on September 22, 2026 as more data points came in. Between September 25 and 27, 2026, GamesRadar published a follow-up detailing the specific signals involved and gathering developer reaction, including comments from the Hooded Horse CEO. On September 29, 2026, Valve announced it was experimenting with dynamic and personalized recommendations in place of some curated placements, and it updated that announcement on October 20, 2026, with a broader rollout reported for early 2027 but no firm launch date confirmed.
Frequently Asked Questions
What exactly changed on Steam’s storefront in September 2026?
Valve appears to have redesigned the Discounts & Events grid on the Steam homepage so that it shows personalized game recommendations based on an account’s playtime, wishlist, and purchase history, rather than the same bestseller-heavy lineup for every visitor. The change was first reported by 80.lv on September 17, 2026, and Valve confirmed the test in a Steamworks announcement on September 29, 2026, though it has not published the exact mechanics.
Did Valve officially announce this change?
Not initially. When the change first appeared, Valve had not published a blog post, changelog entry, or Steamworks documentation update describing the new logic, and the change was identified independently by developers monitoring their store-page traffic and by outlets including 80.lv and GamesRadar. That changed on September 29, 2026, when Valve announced that it had spent the previous few months testing dynamic, personalized recommendations in Discounts & Events, replacing some manually curated placements. The announcement was updated on October 20, 2026, and it describes an experiment with a planned early 2027 rollout rather than a completed launch.
How is this different from Steam’s June 2026 Store Home redesign?
The June 4, 2026 overhaul changed the wishlist threshold needed to appear in the “Popular Upcoming” shelf, raising it from roughly 7,000 to around 80,000 wishlists, which primarily affected unreleased games building pre-launch hype. The September 2026 change targets the Discounts & Events grid for already-released games, personalizing which discounted or featured titles each account sees based on behavior rather than pre-launch wishlist volume.
How big is Steam compared to other PC game storefronts?
Steam controls an estimated 72% to 75% of PC digital game distribution revenue, hit an all-time concurrent-user record of 42,042,778 players on January 11, 2026, and generated an estimated $11.1 billion in gross revenue in the first half of 2026 alone. Epic Games Store, its largest rival, reported a peak of 78 million monthly active users in December 2025 and roughly $1.16 billion in full-year 2025 revenue.
Does this personalization change help indie developers?
Early reporting suggests yes, at least for some studios. Several independent developers reported thousands of additional monthly store-page visits after their games began appearing in the redesigned grid, according to 80.lv. However, since Valve has not documented the underlying signals, results likely vary significantly by genre, playtime patterns, and existing player base size.
What data does Steam reportedly use to personalize the grid?
Based on reporting from GamesRadar, the redesigned feed factors in a player’s playtime history, wishlisted games, and past purchases. Steam has also previously used hardware telemetry, including data from Steam Deck usage, to inform other parts of its recommendation system, though it is unconfirmed whether that specific signal feeds the Discounts & Events grid.
How much does it cost to list a game on Steam?
Valve charges a flat Steam Direct submission fee of $100 per product, which is recoupable once the game generates sufficient revenue. This fee replaced the earlier Greenlight community-voting system and remains unchanged as of September 2026.
Could competitors like Epic Games Store or GOG copy this approach?
Neither platform has announced plans to adopt algorithmic personalization at Steam’s scale. Epic Games Store currently relies on curated weekly free games and manual editorial features, while GOG’s identity is built around human curation and a DRM-free catalog rather than a recommendation engine. Given Steam’s traffic results, however, pressure to experiment with similar features could grow through 2027.
