Steam Next Fest Suite OverviewPart I · Three-Year ArcPart II · June 2026Part III · AI Disclosure
Mode Council — Deep Intelligence Report

Steam Next Fest
June 2026

The largest demo showcase in Steam history. Top 30 performance, roguelite category deep-dive, five-edition historical comparison, and strategic implications for October 2026.

Published June 24, 2026
Edition June 15–22, 2026
Coverage Recognizing Patterns
4,931
Total demos
~200,000
Peak WL gain (IRON NEST)
551
AI-flagged entries
~788
Games w/ 2k+ wishlists
~200
True median wishlists (all games)
Jun 25
Steam Summer Sale opened
Section 01

Key Dates & Press Preview Pipeline

Plan ahead, not on the day.

The Press Preview gave journalists and content creators an 11-day head start on participating demos before public opening. Games that received coverage during this window entered Day 1 of the public fest with algorithmic momentum already building.

April 27
Registration closed — 11:59pm PDT
May 4
Trailers pulled for official Steam marketing
May 18
Demo build deadline for Press Preview
This is a floor, not a target. Top performers had demos live months earlier.
June 3
All required items submitted for Valve review
June 4
Press Preview begins — 10am PDT
Valve Q&A confirmed embargo is optional. Developers can allow coverage to drop immediately, building D1 momentum.
June 4
Trailer opt-out deadline
June 15
Public fest opens — 10am PDT
June 22
Public fest closes — 10am PDT
June 25
Steam Summer Sale opens
Games with active demos entering the Summer Sale window had a compounding visibility advantage.
Demo timing finding: Of the top 10 wishlist earners at February 2026 Next Fest, 5 had launched demos months in advance, 4 launched one month prior, and only 1 surprise-dropped during the fest. The submission deadline is a floor, not a strategy.
Section 02

Top 30 — Performance Heat List

Official and inferred.

Ranks 1–10 are Valve's official most-played demos (June 15–18 aggregate, unique players). Ranks 11–30 are an inferred heat list built from repeated cross-source coverage: Chainplay Top 10, ClemmyGames June 18 and 19 roundups, Polygon, and Game Rant. These are not confirmed Steam rankings — they represent visibility and press heat as of June 22.

Data note: Percentiles for February 2025 onward are drawn from the How To Market A Game survey series. Earlier columns are estimates based on smaller sample sizes and should be read as directional, not definitive.

#GameGenreStatusSignal / Notes

Valve published this as a 3-day aggregate (June 15–18), not a daily ranking. A publicly available full daily top-30 breakdown does not exist for this edition.

#GameGenreNotes
IRON NEST anomaly: IRON NEST was ranked #4 in the official Valve most-played list (June 15–18) but independently confirmed ~200,000 wishlists over the full festival (the ~200,000 (124,622 was a 5-day mid-event snapshot) figure was a five-day mid-event snapshot) — the highest confirmed developer-verified wishlist gain of any game in the festival. Most-played and most-wishlisted do not always align.

Known and estimated wishlist performance

Data caveat: Valve does not publish per-game wishlist data. IRON NEST's ~200,000 total is developer-verified; the ~200,000 (124,622 was a 5-day mid-event snapshot) figure cited in early reporting reflects a five-day mid-event snapshot from a Reddit post by the developers. All other figures are community estimates or media reports. Final post-fest survey data from GameDiscoverCo and HowToMarketAGame is expected several weeks after the event closes.

Breakdown of the top 30 by genre cluster.

Genre ClusterCount in Top 30GamesPattern
Action / Soulslike7Embers of the Uncrowned, Valor Mortis, Onimusha, Swordcery, SPRAWL zero, EMPULSE, CasualtiesSoulslike and fast-action dominated press picks
Strategy / RTS6Dust Front RTS, Wall World Strategy, Anomaly President, Don't Kill Them All, Entropy, Dust FrontRTS/strategy cluster overperformed vs expectations
RPG / Adventure5Echoes of Aincrad, Mistfall Hunter, FantasyJourney, Edge of Memories, Divine FrequencyAnime/RPG very strong; China signal in FantasyJourney
Co-op / Multiplayer4BOMBANANA, Blood Dungeon, TOEM 2, AscendersCo-op presence consistent with Feb 2026 trend
Horror / Thriller3The Mound: Omen of Cthulhu, The Sinking City 2, Entropy (psychological)Lovecraftian wave continues
Simulation / Cozy3IRON NEST, Penguin Colony, Cozy Game RestorationSimulation proving resilient across editions
FPS / Shooter2SPRAWL zero, EMPULSEWall-running / retro FPS niche performing
Section 03

Roguelites — Category Deep Dive

Saturated but not dead.

The roguelite category arrived at June 2026 under a known tension: structurally saturated but not dead. The February 2026 BI analysis put roguelite median wishlists at approximately 900 — below co-op (3,200) and RPG (4,800) but substantially above narrative and F2P. The operative phrase from that data set: "It's a roguelite" is table stakes now, not a selling point. "It's a roguelite where X" is the minimum bar.

Roguelite median WL (Feb 26)
~900
vs 806 overall median; modest premium for the tag
Tears of Metal (Oct 25)
50k+ WL
Co-op roguelite breakout; Oct 2025 second-tier standout
Ball X Pit (Jun 25)
67.5k WL
Action roguelike; pure mechanic differentiation story
Saturation index
High
Genre is viable only with strong mechanical hook in first 10 seconds

Top 10 Roguelite / Roguelike Demos — June 2026 (by heat signal)

No official Valve roguelite sub-ranking exists. This list is assembled from available data: the official most-played top 10, ClemmyGames roundups, Chainplay Top 10, Althera Games post-mortem analysis, and creator coverage. Deckbuilder / roguelite-twist titles are included where the mechanical identity qualifies.

#GameSubgenreSignal tierKey differentiatorvs. category avg

Roguelite performance vs. other top genres — June 2026

The "familiar core + surprise layer" formula: Althera Games' June 2026 post-mortem analysis specifically named deckbuilders and roguelite-twist demos as the games that "produced concentrated wishlist gains" in the category. The formula that worked: a mechanically familiar loop (run, die, upgrade) combined with one layer of surprise that is not legible from the genre tag alone — a physics twist, a timing mechanic, a mid-run narrative choice. Games that leaned on the genre tag alone as a selling point did not break out.

Historical roguelite trajectory across editions

EditionNotable roguelite(s)WL gainedCategory context
June 2026Blood Dungeon, Embers of the Uncrowned, Rizz DungeonEst. 5k–20k eachDeckbuilder-twist formula dominant; pure roguelites in mid-tier
Feb 2026Lootbound (roguelite adjacent)34,000 activations34k activations (not WL); category median ~900 WL surveyed
Oct 2025Tears of Metal (co-op roguelike)50,000+ WLCo-op modifier drove breakout; pure roguelites mid-tier
Jun 2025Ball X Pit (#9 WL), Ratatan (#10), Moonlighter 2 (#13), Morbid Metal (#12)Ball X Pit: 67.5k WLFour roguelite titles in top 15 by WL; healthy but not dominant
Feb 2025Haste: Broken Worlds (top 5 CCU)Estimated top-10%Physics-differentiated roguelite; CCU over pure wishlists
Strategic read for October 2026: A roguelite entering October 2026 Next Fest needs to pass the "10-second test" on the store page capsule: the differentiating mechanic must be visible immediately, not buried in a description. The co-op modifier remains the single most reliable way to elevate a roguelite above category average — Tears of Metal proved it in October 2025, and the February 2026 data confirmed co-op lifts median performance by 3x or more across genres.
Section 04

Wishlist Benchmarks — June 2026

Know your tier.

Performance tiers are built from accumulated survey data across multiple Next Fest editions. The June 2026 final survey will be published weeks post-event; the February 2026 data remains the best current proxy.

Diamond tier
10,000+
Enters Popular Upcoming at launch
Gold tier
7,000–9,999
Strong validated market interest
Silver tier
1,000–6,999
Actionable; not explosive
Bronze tier
0–999
Positioning issues to address before launch

February 2026 percentile reference — best proxy for June 2026

PercentileWishlists gainedWhat this means
30th382Below median; organic visibility only
Median (50th)806Surveyed median; true median ~200 when all games included
70th1,839Solid showing; Silver tier confirmed
90th (est.)~7,400Gold tier entry; publisher-visible
95th13,461Near-Diamond; launch day visibility secured
Highest earner57,074February 2026 maximum; June 2026 ceiling estimated 60–120k+
True median caveat: The surveyed median (806) is inflated by self-selection — only marketing-active developers submit to surveys. The estimated true median across all 4,900+ demos in June 2026 is closer to 200 wishlists. Only 788 of 4,175 analyzed games entered with more than 2,000 wishlists. Those 788 games — 18% of the field — are effectively the only real competition for discovery attention.

Competing with 4,900 games — the real number

Entry wishlist bracketGames in bracket (est.)% of fieldDiscovery outcome
Under 100 wishlists~1,24130%Effectively invisible after Day 2
100–1,999 wishlists~2,14651%Unlikely to break Silver without viral moment
2,000–9,999 wishlists~60014%Algorithm-eligible from Day 3
10,000+ wishlists~1884.5%Diamond candidate; full algorithmic boost
Section 05

Historical Comparison — Five Editions

The pattern is the signal.

A cross-edition analysis covering February 2025, June 2025, October 2025, February 2026, and June 2026. Data sourced from HowToMarketAGame survey series, GameDiscoverCo, Alinea Analytics, WN Hub, and games.gg.

Scale and participation growth

Wishlist percentile shifts across five editions

Percentile Feb 2025 Jun 2025 Oct 2025 Feb 2026 Jun 2026 (est.)
30th pct 584 ~400 ~320 382 ~200–350
Median (50th) 1,079 336 ~480 806 ~400–750 est.
70th pct 2,298 ~1,400 ~1,500 1,839 ~1,500–2,000
95th pct 10,879 ~15,000 ~7,000 13,461 ~12,000–20,000
Top earner 45,655 227,235 (Stellar Blade) ~150,000 (Reanimal) 57,074 ~200,000 (IRON NEST, full festival; ~200,000 was a five-day snapshot)

June 2026 column = estimates pending final survey. All other columns: HowToMarketAGame survey data + GameDiscoverCo analysis.

The October 2024 algorithm dip and recovery: The sharp drop in October 2025 95th-percentile figures reflects an algorithmic recalibration Valve made after the June 2024 edition. The new "Days 1–2 randomised, Days 3+ engagement-ranked" system compressed top-earner maximums initially. By February 2025 and onward, the upper tier recovered. The median and lower percentiles have remained relatively flat, meaning the algorithm benefited top performers more than it helped the average developer.

Most-played top 3 by edition

Edition #1 Most Played #2 Most Played #3 Most Played Dominant genre(s)
Feb 2025 Mecha BREAK (317k CCU) Among Us 3D RoadCraft Mech shooter, party, construction sim
Jun 2025 Vindictus: Defying Fate Wildgate Jump Ship Action RPG, co-op space shooter
Oct 2025 Half Sword Heroes of Might & Magic: Olden Era YapYap Physics combat, strategy, co-op magic
Feb 2026 Windrose (351k WL) Marathon (Multiple co-op titles) Co-op survival, live service shooter
Jun 2026 Echoes of Aincrad Mistfall Hunter over the hill Anime RPG, action RPG, cozy narrative

AI content disclosure — edition over edition

Data note: The 26.5% AI disclosure rate applies to GameDiscoverCo's tracked pool of 4,382 eligible demos. For the full event of 8,682 entries, the disclosure rate is nearer to 20 percent.

EditionAI-disclosed entries% of total demosNotes
Oct 2025507~17%First edition with systematic disclosure tracking
Feb 2026~600 est.~17%Community complaints about "AI slop" began surfacing
Jun 2026551 (confirmed)~11%Kotaku: 10 of 16 random hub games triggered AI warning; disclosure rate may undercount actual AI content

Co-op performance trend

Co-op titles in top 20 wishlisted Co-op % of all entries (est.)
Windrose as the thesis statement for co-op: Windrose entered February 2026 Next Fest with 40,000 Discord members already in place and earned 351,000 wishlists during the event — the highest verified figure in any recent edition outside of major console port titles like Stellar Blade. It is the most direct proof available that community-first development, where the audience is gathered before the fest rather than harvested during it, produces outcomes that are structurally unrepeatable by late-entry competitors.
Section 06

Genre Performance — June 2026 Context

Know your ceiling.

Genre data for June 2026 is projected from the February 2026 BI analysis (the most recent comprehensive breakdown), with adjustments from June 2026 press coverage and the Althera Games post-mortem. Final survey data is pending.

Median wishlists by genre — February 2026 (best available proxy)

Genre / TagMedian WL (Feb 26)10k+ hit rateJune 2026 signal
Open World Survival Craft22,204High64× festival median; Everwinds (Oct 25) confirmed pattern
RPG~4,80020.2%Strongest category; anime RPG performing in Jun 26
Co-op~3,200High8 of top 20 most-wishlisted in Feb 26; co-op modifier lifts any genre
Action / Action-Adventure~1,400MediumSoulslike subgenre outperforming broader action in Jun 26
Roguelite / Roguelike~900Medium-lowSaturated; co-op modifier or mechanical twist required
Simulation~700MediumIRON NEST breakout proves mechanical uniqueness overrides genre ceiling
Narrative / Visual Novel<500LowFormat structurally poorly suited to demo-driven discovery
Free-to-Play~3205.2%vs 16.7% for paid titles; F2P underperforms despite player counts
Deckbuilder / Roguelite-twist~1,200 est.MediumJun 26 post-mortem: "concentrated WL gains" for this formula
Cozy + psychological horror divergence: Althera Games named both "cozy sim" and "psychological horror" as the two genre clusters with the strongest demo-to-wishlist conversion rates in June 2026 specifically. Cozy has heavy competition but high conversion when hit. Psychological horror produced "short, atmospheric, mechanically signature demos" that concentrated wishlists efficiently. Both genres were underrepresented in the February 2026 top tier — suggesting the June 2026 edition may have shifted the genre mix at the top of the wishlist curve.
Section 07

East Asia & China — The Silent Determinant

The multiplier most studios ignore.

The Chinese and Taiwanese publishing presence in Steam Next Fest has moved from a notable data point to a structural feature of the event. Eight of the top 10 most-wishlisted games in the June 2026 early chart were reported as coming from Taiwanese publishers. Simplified Chinese was the second most common localization language across the February 2026 edition — 46% of all 3,449 entries included it.

Simplified Chinese — Feb 2026
46%
Of all entries had SC localization — #2 most common language
Taiwanese publishers — Jun 26
8 / 10
Of top-wishlisted games in early charts
Stellar Blade CCU — Jun 25
68%
Of Stellar Blade's 192k CCU at Jun 25 Next Fest were Chinese players
Missing SC localization
Hard cap
On East Asian discovery; Bilibili/QQ ecosystem bypassed entirely

The China multiplier in Next Fest context

The Stellar Blade data point from June 2025 is the clearest single illustration of the China multiplier at work within Next Fest itself: 192,000 concurrent players, 68% of whom were Chinese, for a PlayStation-published game that was not even marketed as a China-first title. The implication for indie developers is that reaching Chinese-language players on Steam is not an edge case — it is one of the primary drivers of CCU and wishlist momentum during the festival window.

East Asian publishers, particularly Taiwanese studios, have systematically optimised for Next Fest over the past two years: strong tag selection, polished trailers, fully localised store copy in multiple Asian languages. The June 2026 data point of 8 of 10 top-wishlisted games coming from Taiwanese publishers is the output of that systematic preparation, not a one-time anomaly.

Practical implication: A game entering October 2026 Next Fest without Simplified Chinese localization is effectively opting out of engagement from one of Steam's largest player segments. Chinese-language creator coverage on Bilibili mirrors the Western YouTube/Twitch influencer ecosystem in scale. Locking out that ecosystem is not a neutral decision — it is a ceiling on discovery.

FantasyJourney — the June 2026 China signal

FantasyJourney, explicitly identified in ClemmyGames' June 19 roundup as a Chinese-developed title, reached the inferred top-20 of the June 2026 heat list without any significant Western press coverage prior to the event. This pattern — strong East Asian entry achieving Western visibility purely through community-driven discovery — is consistent with the Once Human trajectory from June 2024 and the broader trend of Chinese-developed games outperforming their Western press profile on Steam metrics.

Section 08

Algorithm & Discovery Mechanics

Equal for 48 hours, then ranked.

Valve's ranking system for June 2026 operated identically to the post-October 2024 recalibrated version. The architecture is now well-understood: equal access for 48 hours, then a hard pivot to engagement-driven ranking.

Days 1–2
Equal visibility
Randomised exposure regardless of pre-fest size. Store page conversion dominates.
Day 3 onward
Engagement ranking
Wishlist velocity, demo playtime, and conversion rate drive placement. Diamond games see 3× impression multiplier.
Mid-week shift
Banner displaced
Top banner typically replaced by AAA title pop-under mid-event — expected by developers.

Ranking signals — priority order

SignalWeightActionable implication
Median demo playtimeCriticalAim for 20–60 min sessions. Sub-10 minute bounces suppress ranking from Day 3.
Wishlist conversion rateCriticalCapsule art and trailer quality. 68–88% of wishlists come from players who never launch the demo.
Pre-fest wishlist baseCriticalEnter with 2,000+ minimum; 5,000+ for meaningful algorithmic boost. Next Fest amplifies momentum.
Live stream on store pageHighContinuous streams during peak US/EU hours Days 1–3 boost session dwell time.
Demo download-to-wishlist conversionHighHealthy range: 25–45%. Below 15% signals capsule or genre-fit issue.
Localisation breadthMediumSimplified Chinese is #2 most common language. Missing it caps your ceiling hard.
Feature completenessMediumAchievements, controller support, cloud saves correlate with higher session medians.
The compression effect: A developer who entered a game at February 2023 Next Fest (~900 games) received 900,000 impressions and 1,600 wishlists. The same game entered at February 2026 (~3,400 games) received 58,000 impressions and 350 wishlists. That is a 15× reduction in impressions at an equivalent starting position. With ~4,900 games in June 2026, the compression is now deeper. Games that are not algorithm-eligible by Day 3 effectively do not exist for the remainder of the event.
Section 09

Council View — Decisions for October 2026

The right question to ask.

October Next Fest runs October 19–26, 2026. The following analysis is structured for allocation decisions: what the June 2026 data demands from a studio entering that window with a new title.

VariableWhat June 2026 showsOctober 2026 targetHard floor
Pre-fest wishlist baseAll Diamond earners entered 3k+; most with 10k+8,000 target / 12,000 stretch2,000 minimum or algorithm-invisible after D2
Demo availabilityTop performers had demos live months beforeDemo live by August 2026Demo must pre-date Oct 19 by 4+ weeks
Demo arc length20–60 min sweet spot; 60+ min ARPG demos lagged on "completed session"30–45 min designed session10 min minimum; below this bounces tank Day 3 ranking
Download-to-wishlist rateHealthy: 25–45%Above 30%Below 15% = capsule/genre-fit intervention needed
Simplified Chinese46% of competitors have it; 8/10 top wishlisted from East Asian publishersLocalisation complete before registrationMissing SC = hard cap on East Asian discovery
Co-op modifier8 of top 20 wishlisted in Feb 26; Tears of Metal (50k+) in Oct 25Add co-op if design permitsN/A — optional but high multiplier
Press preview11-day early access; embargo optional; D1 momentum built in PP windowKeys to 20+ creators before PP opensAt minimum: send to top 5 relevant creators
Roguelite positioningSaturated but viable; "familiar core + surprise layer" formula worksDifferentiator visible in first 10 seconds of trailerGenre tag alone not sufficient
The thesis the data confirms repeatedly: Next Fest is a multiplier, not a launchpad. Every Diamond-tier game in the data set entered with at least 3,000 wishlists. Most entered with 10,000 or more. The wishlist ramp between now and the registration deadline is the campaign — the festival is the amplification window for momentum that already exists. A studio that asks "how do we win Next Fest?" is asking the wrong question. The right question is "what is our wishlist trajectory between now and October 12th?" (the expected registration deadline for the October 19 event).

What the IRON NEST case study means for planning

IRON NEST entered June 2026 with over 250,000 pre-existing wishlists, built by two developers over months of community engagement, developer diaries, and prior demo availability. The ~200,000 wishlists it gained during the festival's first five days — on a near-linear daily growth curve, not a day-one spike — are the direct output of that existing community. The ratio of fest-gain to pre-fest base was roughly 50%, which is consistent with the "roughly 2×" multiplier pattern observed across the broader data set. IRON NEST did not break the pattern; it is the clearest single expression of it at scale.

The lesson for October 2026 planning is not to replicate IRON NEST's genre or mechanics. It is to replicate the precondition: a community that already exists and is already engaged before the festival opens its doors.

AI flood opportunity: The AI content flood in June 2026 — Kotaku flagging 10 of 16 random games triggering AI content warnings — creates an asymmetric advantage for handcrafted titles. In an environment where the average demo is low-effort, a genuinely polished game stands out disproportionately. The discovery problem is finding those players who care about craft quality; the June 2026 data suggests they exist in large numbers and are actively hunting for "relief valve" titles that are not AI-generated. This is the context in which IRON NEST's 99% Overwhelmingly Positive demo rating must be understood.

Abbas Saleem

Principal Consultant · Llama & Griffin

Gaming industry executive, strategic advisor, and independent analyst. 25+ years across gaming, transmedia, and studio strategy — AAA studios, investment committees, government advisory boards, and founder conversations across six continents.

Current work spans executive producing three titles, advisory to VC funds, and co-production frameworks with publishers across Greater China, MENA, and Western markets.

Publishes ongoing analysis on Recognizing Patterns.

Book a Conversation → LinkedIn llamagriffin.com →