The Steam Workshop has a supply problem, and it isn't that people aren't making mods. It's that most of what gets published arrives without saying what it is.
Using ModSignal's live dataset of 7,278,533 active Workshop items, we measured the one field every single publisher controls — the description — and crossed it against the outcome that matters. The result splits the Workshop into two economies that barely resemble each other.
2,808,333 mods — 38.6% of the entire tracked catalog — carry no description at all. Not a short one, not a lazy one. An empty page. Another 1,816,138 mods (25.0%) have neither a description nor a preview image: a title and nothing else.
That is not a footnote to the Workshop. It is the plurality of it.
The empty shelf
Grouping every active mod by how much description it carries, and then measuring what that catalog actually earns, produces a staircase that should be uncomfortable reading for anyone who publishes here:
| Description length | Mods | Share of catalog | Reached 1,000 subs | Zero subscriptions | Share of all subs |
|---|---|---|---|---|---|
| None (empty) | 2,808,333 | 38.6% | 2.8% | 33.3% | 6.7% |
| Under 50 chars | 1,452,922 | 20.0% | 7.1% | 11.9% | 8.5% |
| 50–149 chars | 1,076,686 | 14.8% | 13.0% | 9.5% | 14.0% |
| 150–299 chars | 1,940,590 | 26.7% | 27.6% | 3.8% | 70.8% |
The mods that bothered to write a real description are 26.7% of the catalog but hold 70.8% of every subscription on the platform — roughly 8.9 billion of the 12.5 billion total. The blank ones are 38.6% of the catalog and hold 6.7%.
The ten-fold gap
The headline number is the ratio between the bottom and the top of that ladder.
A mod with a 150–299 character description reaches 1,000 subscriptions 27.6% of the time. A mod with no description reaches it 2.8% of the time. That is a 9.7× difference — and the gap only sharpens at the top: the descriptive band produces 125,996 mods at 10,000+ subscribers, while the blank band produces 14,109.
The same split shows up in the graveyard. Of the empty-description mods, 936,163 (33.3%) have zero subscriptions. In the 150–299 band, it's 74,438 (3.8%) — an eight-fold difference in how many mods got nothing at all.
To be clear about what this is: it is a correlation, and it is not that Steam magically rewards prose. A description is a proxy for whether a human actually built a storefront. Creators who write one also tend to set a preview image, name the mod meaningfully, tag it correctly and publish it once instead of a hundred times. The blank page is the visible symptom of a mod that was never presented to anyone — and the data says those mods are the ones nobody installs.
Where the blank pages pile up
The blank-page rate is wildly uneven across ecosystems, and it tracks almost perfectly in inverse to whether the game's modding scene has a culture of publishing.
| Game | Mods | % no description | % reaching 1k subs | Avg subs |
|---|---|---|---|---|
| Rust | 177,862 | 84.5% | 2.3% | 319 |
| Human Fall Flat | 216,632 | 81.9% | 2.6% | 276 |
| Besiege | 238,066 | 76.4% | 2.3% | 162 |
| Space Engineers | 176,031 | 61.6% | 6.8% | 492 |
| Counter-Strike 2 | 144,939 | 56.5% | 5.6% | 2,942 |
| Terraria | 149,513 | 52.8% | 3.8% | 417 |
| Trailmakers | 333,312 | 46.5% | 0.9% | 78 |
| Scrap Mechanic | 197,788 | 43.8% | 2.7% | 220 |
| Portal 2 | 179,270 | 39.1% | 2.5% | 223 |
| Brick Rigs | 176,448 | 20.5% | 13.2% | 635 |
| Cities: Skylines | 176,190 | 7.6% | 33.6% | 4,472 |
| Garry's Mod | 176,041 | 5.2% | 51.1% | 8,928 |
| Left 4 Dead 2 | 176,082 | 1.8% | 31.5% | 3,217 |
Two games of almost identical size sit at opposite ends. Garry's Mod has 176,041 mods, a 5.2% blank-page rate, and 51.1% of them clear 1,000 subscribers, averaging 8,928. Rust has 177,862 mods — nearly the same — a 84.5% blank-page rate, and just 2.3% clear 1,000, averaging 319. Left 4 Dead 2 is the cleanest ecosystem on the board at a 1.8% blank rate.
Nowhere is the pile-up more visible than in the count of mods that are both blank and dead — empty description, zero subscriptions:
| Game | Blank + zero-sub mods |
|---|---|
| Rust | 146,173 |
| Stick Fight: The Game | 122,006 |
| Human Fall Flat | 111,030 |
| Counter-Strike 2 | 80,802 |
| Besiege | 59,622 |
| Unturned | 41,119 |
| Terraria | 36,668 |
Rust alone has 146,173 mods that say nothing and installed nowhere — an entire silent city of content. These are overwhelmingly the output of bulk uploads: workshop-file dumps, repackaged skins, and auto-published template content that was never meant to be discovered. They inflate a game's mod count while contributing nothing a player could choose between.
The two exceptions prove the rule
Before concluding that a blank page is fatal, the data offers two counterexamples that are more interesting than the rule itself.
Wallpaper Engine is half blank — 53,289 of its 105,064 mods (50.7%) have no description — yet 93.7% of the whole catalog clears 1,000 subscriptions, averaging 18,247 each. The reason is structural: Wallpaper Engine's content is consumed in-place by a client that browses wallpapers visually, in a grid of animated previews. The preview image is the storefront, and an empty description costs nothing. Wallpaper Engine is where ModSignal's catalog is biggest, and it is a reminder that "description quality" is only a signal in ecosystems where browsing is text-driven.
Aimlabs shows the other edge. It has the platform's single most-subscribed mod with no description anywhere in the dataset: Oregon, at 6,396,975 subscribers — followed by Bind (6.26M), Ascent (6.13M) and Custom_rA100hz (5.85M). These are aim-training scenarios that ship as in-game codes, and their average view count is 38. Millions of players use them; almost nobody reads their Workshop page, because the download happens inside Aimlabs itself.
Both exceptions share a trait: the Workshop page is not the distribution channel. When a client or a game delivers the content directly, a blank page is harmless. For everyone else — the overwhelming majority of the Workshop, where discovery runs through the page — a blank page is a closed door.
What the blank page actually costs
Three things follow from the numbers, and they are the same three things whether you are a creator or a studio looking at your own ecosystem:
-
The Workshop's "supply" is softer than it looks. Trailmakers lists 333,312 mods and averages 78 subscriptions; Rust lists 177,862 and averages 319. Strip out the blank, zero-sub dumps and the real competitive field inside many games is a fraction of the headline count. Cities: Skylines at 176,190 mods and an average of 4,472 subscriptions is playing a different game entirely — it is a curated ecosystem dressed as a bulk one.
-
Listing quality is the cheapest available lever. The jump from the blank band to the 150–299 band is 2.8% → 27.6% on the single same-day action of writing a few honest sentences. No marketing budget, no follower base, no luck with an algorithm — just a filled-in field that a third of the catalog leaves empty.
-
Mod count is not a quality metric. Any ecosystem can inflate its mod number by tolerating bulk uploads. The description rate tells you whether the count means anything. A game whose catalog is 84% blank is not a thriving Workshop with a discoverability problem; it is a small Workshop buried under a large amount of nothing.
The Steam Workshop's real constraint in 2026 is not how many mods exist. It is how many of them ever bothered to be legible. 2.8 million of them never did — and the data shows exactly where they went.