Ask a Steam Workshop creator what to do after their first hit, and you'll usually hear one of two answers: double down on the game that worked, or spread out before it dies. Both sound reasonable. Nobody has measured which one actually wins.
Using ModSignal's live dataset of 7.33 million active Workshop items published by 2.06 million distinct creators, we split every creator by how many games their portfolio spans, then measured what each group actually earns, how often they break out, and which games share the same creators. The result is a clear portfolio premium — with two sharp twists at the edges.
The 9% who own 31% of the Workshop
Start with the headline split. Of 2,060,944 creators with at least one active item, 184,459 — 8.95% — publish across two or more games.
That eighth of the creator base holds 30.7% of every subscription on the platform (3.85 billion of 12.56 billion) on 25.4% of all active items. Normalised per item, a portfolio creator's mod is worth about 21% more than a single-game creator's.
The ladder is monotonic, and steeper than you'd expect:
| Portfolio breadth | Creators | Active items | Subs per item | Avg subs per creator |
|---|---|---|---|---|
| 1 game | 1,876,485 | 5,326,999 | 1,635 | 4,641 |
| 2 games | 146,628 | 1,209,053 | 2,036 | 16,788 |
| 3–5 games | 36,376 | 588,578 | 2,121 | 34,312 |
| 6–10 games | 1,415 | 56,244 | 2,394 | 95,153 |
| 11+ games | 40 | 4,120 | 1,278 | 131,604 |
A creator working in 3–5 games averages 7.4x the subscriptions of a single-game creator. At 6–10 games it's 20.5x.
That number looks like the strongest argument for breadth you could make. It isn't — it's mostly survivorship. But before we get to that, look at the one thing breadth genuinely does change: the odds of a breakout.
Breadth multiplies breakout odds
ModSignal's breakout score flags mods accelerating faster than their size predicts — we filtered to the honest early-risers (breakout_score >= 60, under 20,000 current subscriptions, positive 7-day growth) so the list isn't just mega-mods echoing their own install base.
Sorting creators by whether they have ever run one:
| Portfolio breadth | Creators | Share with ≥1 live breakout | Breakouts per creator |
|---|---|---|---|
| 1 game | 1,876,485 | 4.4% | 0.10 |
| 2 games | 146,628 | 13.3% | 0.34 |
| 3–5 games | 36,376 | 24.3% | 0.77 |
| 6+ games | 1,455 | 46.7% | 2.06 |
A creator in 3–5 games is 5.5x more likely to be running a breakout than a single-game creator. At 6 or more games, it's 10.5x — nearly half of them are running one right now.
The same effect shows up in the all-time rankings. Portfolio creators are 8.95% of the creator base, but they hold:
- 36 of the top 100 most-subscribed mods (36.0% — a 4.0x over-representation)
- 160 of the top 500 (32.0% — 3.6x)
- 1,496 of the top 5,000 (29.9% — 3.3x)
The twist: the advantage is not from being prolific
The obvious objection is that portfolio creators just have more items, so of course they break out more. So we controlled for it — holding item count constant and varying only breadth:
| Creator's active items | Single-game breakout rate | Multi-game breakout rate | Creators (multi / single) |
|---|---|---|---|
| 20–49 | 32.6% | 40.2% | 12,737 / 22,339 |
| 50–199 | 47.9% | 57.6% | 4,299 / 7,227 |
| 200–999 | 63.0% | 62.2% | 563 / 768 |
| 1,000+ | 81.1% | 70.6% | 17 / 37 |
The portfolio premium is real — and it evaporates with scale. Among small and mid-sized catalogues (20–199 items), working across games adds roughly 8 to 10 percentage points of breakout probability. At 200–999 items the two groups converge. Past 1,000 items, the single-game creator is ahead.
That is the actual mechanic, and it isn't "breadth as a lottery ticket." A creator with 30 mods in one game has one discovery pool. A creator with 30 mods across three games has three — three tags, three communities, three chances at a front-page impression. But that advantage is about number of independent audiences, and a 1,000-item single-game catalogue has already bought its way into more audience than a scattered portfolio can assemble. Breadth substitutes for scale; it doesn't beat it.
The bottom of the breadth ladder confirms the point brutally. The creators with the widest portfolios are the least successful of anyone:
| Games spanned | Creators | Items | Total subs | Subs per item |
|---|---|---|---|---|
| 15+ games | 8 | 1,724 | 488,693 | 283 |
Eight creators publish across 15 or more games. Between them they hold 1,724 items and less than half a million subscriptions total — an average of 283 subs per item, roughly one sixth of the platform average. The 40 creators spanning 11+ games average 1,278 subs per item, below single-game creators. Breadth taken to its extreme is spray-and-pray, and it reads exactly like it.
Which games share creators
If breadth works by stacking independent audiences, the value of any second game depends on how well its audience correlates with the first. The creator-overlap graph measures precisely that. Across 351 games there are 20,764 game pairs sharing at least one creator, and the tightest pairs are startlingly tight:
| Games | Creators publishing in both |
|---|---|
| Trailmakers ↔ Scrap Mechanic | 2,010 |
| Planet Coaster ↔ Planet Zoo | 1,945 |
| Garry's Mod ↔ Left 4 Dead 2 | 1,809 |
| Garry's Mod ↔ Portal 2 | 1,808 |
| People Playground ↔ Trailmakers | 1,473 |
| Besiege ↔ Kerbal Space Program | 1,391 |
| Stormworks: Build and Rescue ↔ Kerbal Space Program | 1,389 |
| Portal 2 ↔ People Playground | 1,304 |
| Space Engineers ↔ Empyrion - Galactic Survival | 1,297 |
The pattern is legible. These are not random pairings — they're engineering-sandbox clusters (Trailmakers, Scrap Mechanic, Besiege, Stormworks, Space Engineers, KSP), Valve's Source engine cluster (Garry's Mod, Portal 2, Left 4 Dead 2, Counter-Strike 2, Half-Life: Alyx), and coaster-park simulation (Planet Coaster, Planet Zoo). Mod creators follow toolchains and engine conventions, not genres as marketed.
That makes some games structural hubs. Counting games that share 100+ creators with at least 100 other games:
| Game | Overlap partners (≥100 shared creators) | Largest single overlap |
|---|---|---|
| Portal 2 | 50 | 1,808 |
| Garry's Mod | 48 | 1,809 |
| Besiege | 42 | 1,391 |
| Space Engineers | 36 | 1,297 |
| Stormworks: Build and Rescue | 33 | 1,389 |
| Kerbal Space Program | 32 | 1,391 |
| People Playground | 31 | 1,473 |
Portal 2 leading the table is the surprise. A 2011 single-player puzzle game is the Workshop's best-connected ecosystem — because its level-editor tooling taught a generation of creators who then carried those skills into Garry's Mod, People Playground, and the sandbox wave. Its partner games aren't its competitors; they're its alumni.
The three archetypes
The named creators at the top of the portfolio ranking show all three outcomes, and they're worth knowing before you generalise:
- The anchor. Oskar Potocki (Poland) — 4 games, 143 items, 49.5 million subscriptions, the most of any multi-game creator on the platform. But 49.4M of it (99.8%) sits in one game: RimWorld. The other three games hold three items and 77,000 subs between them. This is a single-game giant with a footprint, not a portfolio play.
- The genuine two-pillar.
star— 9 games, 123 items, 14.1M subs, but the weight is concentrated in two: Project Zomboid (7.6M) and Don't Starve Together (5.7M), with the other seven games contributing 850,000 combined. Two strong audiences, deliberately farmed. - The adjacency play. Voikanaa (Finland) — 5 games, 10.4M subs, and 10.1M of it in Garry's Mod, with the remaining four games (Left 4 Dead 2, Counter-Strike 2, SteamVR, Half-Life: Alyx) all inside Valve's own ecosystem. This is the overlap graph in action: a creator who knows Source engine conventions moving only to where those conventions pay.
Notice what none of them did: none of them are the 38-games, 745-item creator with 212,635 subs. The most-broadened portfolios in the dataset are the least valuable ones.
What this means for creators and studios
- For creators, a second game is a real growth lever — up to a point. Held at constant catalogue size, adding games adds roughly 8–10 percentage points of breakout probability. But it works because it multiplies audiences, so pick the second game by toolchain adjacency, not by genre: the overlap graph says a Source creator should go to another Source game, not to a popular unrelated one.
- Breadth is a substitute for scale, not a replacement. If you have 20–200 items, spanning games meaningfully raises your odds. If you have 1,000+, you already own more audience than a portfolio can assemble, and scattering genuinely costs you.
- Past five games the returns go negative. The 11+ and 15+ game cohorts earn below the single-game average per item. There is no version of this where publishing one item into thirty games is the play.
- For studios, the overlap table is a recruiting map. If you want to know which other games' creators will understand your modding surface on day one, look at who already publishes across both — Planet Coaster and Planet Zoo share 1,945 creators; Trailmakers and Scrap Mechanic share 2,010. That is a warm audience you can reach before you ship a single tool.
- And for platform-side discovery, Portal 2's 50 partner ecosystems are the strongest signal in the graph that Workshop skill is transferable and that cross-game discovery is being left entirely to creators to figure out themselves.
How we measure
All figures come from the ModSignal live dataset of the Steam Workshop. Creators are identified by Steam64 account ID and counted only where they hold at least one active item; game count is the number of distinct games in which that ID has an active item. Subscription totals are current installs as of the 2026-09-22 snapshot (7,325,178 active items, 401 games tracked). Breakout figures use ModSignal's breakout_score >= 60 filter combined with growth_pct_7d > 0 and subscriptions_current < 20,000, so the list reflects mods accelerating ahead of their size rather than established giants. Game-pair overlap counts distinct creators publishing in both games of a pair; pair statistics cover 351 games with at least one shared creator.
This is the data ModSignal tracks daily to spot breakout creators and fading ecosystems early. Explore the live dataset to watch the portfolio effect move, or compare games directly.