Launch success on Steam: are there correlating metrics?
Or are we at 'launch and fingers crossed' again? Also: the latest most-streamed games & lots of news
[The GameDiscoverCo game discovery newsletter is written by ‘how people find your game’ expert & company founder Simon Carless, and is a regular look at how people discover and buy video games in the 2020s.]
Well, it’s time to explore PC/console game discovery again, whether you like it or not - and today, we’re poking around somewhere angels fear to tread: is ‘get lots of wishlists and hope’ actually the best method of creating a successful game launch in 2026?
Before we start, we hadn’t spotted that the NYU Game Center has been putting a bunch of its guest lectures online, and we really enjoyed Getting Over It & Baby Steps’ Bennett Foddy talking about his provocative, masocore game design style: “I particularly love it feeling like life or death when you’re trying to do something normal.”
[THE DEEPEST PC/CONSOLE DATA? You can get a free demo of our GameDiscoverCo Pro company-wide ‘Steam deep dive’ & console data by reaching out today - >90 orgs have it. Or, signing up to GDCo Plus gets the rest of this newsletter and Discord access, plus more.]
Game discovery news: Stranger than Subnautica?
Let’s kick things off by vigorously probing the game platform & discovery news of the week so far, like so:
According to GDCo Pro's 'trending' unreleased Steam games by 'new wishlists in the last 7 days', May 4th to 11th, underwater survival sequel Subnautica 2 (out this week!) is safely at #1, and Forza Horizon 6 - out next week - is also booming again at #3 (and getting a lot of pre-orders - we have it at 950k on Steam alone.)
New entry-wise, we see Stranger Than Heaven (#2), the latest epic ARPG title from the devs of Yakuza, WWI trench warfare simulator Dig In (#6), and awaited shooter sequel Aliens: Fireteam Elite 2 (#8). Also: there’s a surge for Soviet kiosk sim with open playtest, I Have No Change (#5) - quite a diverse set of genres.
Xbox and Discord officially announced their Game Pass collab, including “a starter edition of Xbox Game Pass” with 50+ games & 10 hours of cloud play, part of the new Nitro Rewards bonuses for paid Discord subs. (Game Pass subscribers also get extra Discord orbs monthly and when completing quests - it’s symbiotic.)
Much confusion on Forza Horizon 6 getting leaked & pirated on PC early. Not the facts (it has!), but claims of Playground Games making pre-load files playable early were denied by the dev. It’s the third recent AAA game to get a pre-release hack, with Death Stranding 2 and Pragmata also getting pirated before release.
Our 2c on the leak: we think a Steam reviewer key (or token/file manifest?) got to pirates for FH6. (So does SteamDB, which rather awkwardly got donated a token for the full game around the time this blew up.) Outside chance of a Steam or encryption hack, but human-level failure (influencer/reviewer hack?) seems likely.
Roblox’s top titles on the platform for April 2026 included anime-ish ARPG Sailor Piece - which hangs out around 400,000 CCU - making it all the way to #1 for the on minutes played (and #5 on monetization.) Related: PC Gamer interviewed the Roblox-native dev of current Top 10 survival title 99 Nights In The Forest.
Nintendo results aftermath: the company is forecasting Switch 2 hardware sales down from 19.9 million (this year) to 16.5 million units (next year), and the new hardware price “does not fully account for all cost increases”. (Unrelated: we dug it that their Switch price increase PR also included $ changes for hanafuda cards.)
After we talked about the
PopularUpcoming section of the new Steam Beta UI, it appears that it’s been further shifted to show “a select set of the top games coming out in the next few days”, vs. the Popular Upcoming-style view. (If it ships, some devs will miss a pre-launch wishlist bump, tho we’re not sure they were HQ ones.)PlayStation results aftermath: the game division saw profits up 11% to $2.9b for the fiscal year, even with a $760m writedown on the Bungie acquisition. But it’s forecasting revenue down in FY2026, since “we plan to base our PS5 hardware sales in FY26 on the volume of memory we can procure at reasonable prices.” (More squeeze.)
The Verge has gone in-depth on Netflix’s new ‘TV games first’ strategy, suggesting the “tab at the top of the app alongside movies and television” without quitting out makes them feel way more seamless, and concluding: “Games are no longer a separate service Netflix operates, they’re just part of Netflix.”
Microlinks: Twitch is punishing streamers who use viewbots by capping their concurrent viewership; Steam recently added more detailed docs on the Roadmap image feature on Steam; “It took Apple 98 days to remove an infringing game after receiving my DMCA notice”, vs. 1-7 days for Roblox.
OUR SPONSOR: FirstLook: Build better gaming communities
Your Discord rolled chaotic neutral. 8,000 members, three active ones, a #general that’s mostly cat pics and patch complaints. Iconic. But who are your power players? What do they actually think of your last patch? Silence.
FirstLook rolls your community a +5 to insight: onboarding funnels, playtest tracking, sentiment capture, feedback dashboards. Stop guessing what players want. Start shipping what they’ll actually play, share, and stay for.
Apply for a pro Discord audit → GameDiscoverCo readers skip the queue.
Launch success on Steam: correlating metrics?
Look, it’s simple: more people are interested in your PC game, then you get more Steam wishlists, and then when you launch, the game’s a big hit. Just go get more wishlists and you’ll be fiiiine. It’s all about ‘number go up’. Riiight?
Except: as we’ve proven with analysis: “The idea of wishlist ‘conversions’ on Steam as some kind of reliable metric is near-fatally flawed. Why? Because the performance range of ‘wishlists at launch’ compared to ‘sales at the end of Week 1*’ varies by 10-20x, not 10-20%.”
How do we project slightly more certainty into that outcome? At GameDiscoverCo, we do delve into why individual games were a hit. But we don’t as a rule, try to predict outcomes. However, in chatting to Lumo Publishing’s Steve Stopps recently, we realized there are absolutely things you can metricize that might correlate - in part.
And the first one of these we wanted to highlight was Steam follower to wishlist multiplier. Chris Z has a brief explainer of the metrics, but the follower button on Steam allows players to see more info about a game in their activity feed & news hub.
You can see public follower counts, and extrapolate wishlist numbers (or see them directly, if you’re the dev of a title.) So you can get to a multiplier, which we’ve documented before and was a median of 1 follower for every 9.6 wishlists in 2021, and 1 follower for every 12 wishlists in 2023.
The claim Steve made - which we’ve nodded along to in the past, but hadn’t explored, was that the higher your ‘follower to wishlist’ multiplier, the lower the conversion for your game. Why? Because wishlists without a follow are likely to be more ‘casual’ in nature. (Also perhaps: most inorganic calls-to-action are wishlist-first, nowadays.)
So we grabbed games released in the last 12 months with >50k wishlists, and - removing the biggest outliers - mapped that multiplier (X axis) with success in terms of outperformance of expected conversion metrics (Y axis). And indeed, you can see a downward tilt to the ‘line of best fit’ as the ‘follower to wishlist’ multiplier goes up:
To give a couple of recent examples, Subnautica 2 is out tomorrow, and is at 5 million wishlists and 352,000 followers, a 14.2x multiplier - and below the 16.1x median. It’s also a meaty, in-depth title. So - no reason to hint at underperformance from that.
On the other hand, RV crafter Outbound came out yesterday and debuted with 1.6 million wishlists and 64k followers, a 25.6x multiplier. Based on current CCU metrics (max 6k), it may struggle to get to a median conversion of 200k units in Month 1 - though still do decently.
Steve Stopps also analyzed GameDiscoverCo Pro data on around 5,000 games released in 2025 - a much broader range. And here’s a subset of his data on ‘follower to wishlist multiplier’ vs. lifetime sales, showing that games with higher multiplier do worse:
Needless to say, there’s all kind of issues with generalizing on these stats. For example, according to our data, RV There Yet? had a sky-high 42x follower to wishlist multiplier and still sold 3.7m copies in Month 1, haha. But here’s how we think about it:
Does your game have a very casual audience? If you’re friendslop-adjacent, or Escape Simulator 2, it’s expected to have an above-median ‘follower to wishlist’ multiplier. These games also have sky-high ‘reviews to sales’ multipliers - players are parachuting in to play the game, not mess with Steam features.
If not casual, but a high multiplier, beware & examine forecasts: Not picking on them, but less casual games we can see with high follower-to-wishlist multipliers like Death By Scrolling, Keeper or Skate Story failed to outperform median expectations. (They still did fiiine, but not as well as the devs might have hoped.)
Low multipliers are good, because people are strapped in for the ride: you don’t get 10% of wishlisters signing up for news unless legit interested! The 100 highest-multiplier games in GDCo’s 600+ game sample with >50k launch wishlists have median ‘launch wishlist balance to month 1 sales’ conversion of 0.22x. But the lowest 100 have a median of 0.37x - that’s 68% better.
Of course, the problem here is there’s WAY too many variables. Outbound, for example, had crazy-good shortform video reach. But maybe many of those people were maybe a) wishlisting the general idea of ‘RV chill times’, not the exact game b) more onboard with paying RV There Yet?-ish prices of $8, than the $23 it launched at.
Or it could be a price/execution combo. Though many Outbound players turned up to try the demo, with a max CCU of over 3,500, the game just slipped to Mixed reviews, another thing that will dampen sales. Steve Stopps also provided us with select data on this idea, showing that poor reviews drop sales by 2-5x over time:
Anyhow, the general point here is: maybe it’s not OK to just say ‘we’re going for max wishlists, and let’s plug in median conversion numbers!’ Some of the things you might want to take into account, besides ‘follower to wishlist’ multiplier and review estimates, are:
Active/repeated demo players: in addition to CCU, are people engaging and re-engaging with a demo version? What’s the NPS score for players?
Discord & community activity & sentiment: we often look at Steam forum traffic pre-release, particularly volume. And Discord server data is very helpful.
Organic influencer activity: are people excited - when not incented - to talk about your game in visual formats? This is a biggie & should have high weighting.
What’s wishlist momentum coming into launch? Are we seeing the game in the Top 20 for new wishlist adds for multiple weeks, or is it stumbling into launch? You might also consider ‘wishlists added per month’ instead of ‘total wishlists’.
Not sure if anyone is doing this yet, but we could see a matrix of all of these factors changing first week/month conversion estimates across a portfolio. We haven’t been keen to do it en masse, because there are so many outliers that we will regularly look stupid. (And you may end up maths-ing yourself to death.) But there’s hints here…
Top streamed, April: Overwatch not underwatched
To end, here’s the newest data via livestream analytics platform Stream Hatchet, bringing the Top 100 most-streamed games for April 2026 across the big (non-China) game video streaming platforms like Twitch & other medium/small platforms.
As we do, here’s the full list of April’s Top 100 (Google Drive link) with GDCo notes. And yep, Stream Hatchet’s Mark Rowland joins us on commentary to check trends:
Overwatch continues to move forward, as others step back: five of the top seven games had marginal decreases, but Blizzard’s Overwatch continues good results after its Feb. rebrand, with 15.1% growth to 46 million hours watched at #6, sneaking past DOTA 2 and Fortnite. (Did the Underwatch April Fools gag help?)
Rust blasted into the Top 10 for the first time in >18 months: the brutal survival game had 39 million hours watched, a 265% increase from April. That’s partly due to one particular creator: TheBurntPeanut, who set up a new server, so was watched playing Rust on Twitch for 8.4 million hours (!) during the month.
Nintendo-related games - and Windrose - topped debuts: specifically, quirky life sim Tomodachi Life hit 21.6 million hours watched (#19), followed by a Twitch Rivals-enhanced mobile/Switch F2P debut Pokémon Champions (#24, 16.9m). Also big? Steam piratical hit Windrose (#26, 14.7m) and Pragmata (#33, 11.7m.)
Further down the charts, a couple of games that hit 1.0 on Steam did make it in - PvE tactical FPS Gray Zone Warfare (#56) and survival game Soulmask (#88), showing that Early Access graduates can still summon up new mojo under positive circumstances!
[We’re GameDiscoverCo, an analysis firm based around one simple issue: how do players find, buy and enjoy your PC or console game? We run the newsletter you’re reading, and provide real-time data services for publishers, funds, and other smart game industry folks.]