Every Workshop creator eventually considers the same shortcut. You have thirty mods sitting half-finished in a folder, or a texture pack you can split into variations, or a script you can copy across a dozen items. Uploading them one at a time is tedious. So you write a loop, hit publish, and go to bed.

That shortcut has a measurable price, and until now nobody had put a number on it. Using ModSignal's live dataset — 6.81 million active Workshop mods and 12.50 billion subscriptions — we isolated every item published in the last 24 months and measured exactly what happens when a creator publishes in bulk instead of in sequence. 2,392,379 mods across 819,677 creators, bucketed by how many items they fired into the Workshop within the same hour.

The result is one of the cleanest dose-response curves in the dataset.

The batch penalty

Every mod in the cohort was classified by how many items its creator published inside the same clock hour — the tightest proxy for a scripted or one-sitting upload session. Then we measured how often each group reached 1,000 subscriptions:

Items published in the same hour Mods Avg subscriptions Reached 1,000 subs
Alone (1) 1,743,578 907 9.33%
2–3 372,704 528 8.24%
4–10 186,458 329 5.30%
11–30 66,439 164 3.00%
31 or more 23,202 82 1.04%

A mod published on its own reaches 1,000 subscriptions 9.33% of the time. A mod published as one of 31 or more items in the same hour reaches it 1.04% of the time. That is a 9.0x gap — and it holds in every direction you slice it: average subscriptions fall from 907 to 82, a 11x decline, monotonically, with no reversal anywhere on the ladder.

3.75% of all recently published mods (89,641 of them) fall into the "11 or more in an hour" band. Nearly one in a hundred (23,202 mods, 0.97%) was fired off as part of a 31-item burst or larger. That is not a fringe behaviour. It is a structural feature of how a meaningful slice of the Workshop gets supplied.

It survives every control we could run

The obvious objection: maybe batch publishers are just bad creators, and the batch is a symptom rather than a cause. We tested that four ways and it held each time.

Control 1 — the same creator, solo mods vs. batched mods. Restricting to creators who did both, and comparing only their own output:

Mods Avg subscriptions Reached 1,000 subs
Their solo-published mods 30,123 353 5.65%
Their mass-published mods (11+ in an hour) 45,725 192 3.20%

Across 1,277 creators who have both kinds of output in their own history, their batched mods land at 57% of the 1,000-subscriber rate of their solo ones — a 43% penalty applied inside the same account, with the same creator's own skill and audience held constant. It is not a creator-quality effect.

Control 2 — holding catalog size constant. Same total output, different release rhythm:

Creator's total active mods Published on 1 day Spread across 2 days Spread across 3+ days
8–12 mods 2.37% (8,105 mods) 2.90% (11,835) 9.16% (208,532)
13–20 mods 2.16% (4,492) 2.80% (6,004) 9.74% (189,858)
21–35 mods 2.99% (3,676) 2.90% (4,448) 11.16% (181,029)
36–60 mods 2.42% (1,858) 3.46% (2,051) 12.77% (141,580)

A creator with 21–35 mods who spreads them across three or more days hits 1,000 subs 11.16% of the time. A creator with the identical output released in one sitting hits it 2.99%. Same volume. Same creator tier. 3.7x difference, purely from the rhythm.

Control 3 — position inside the batch. If a flood of items is competing against itself, the later items should do worse:

Position within the day's batch Mods Avg subscriptions Reached 1,000
1st of the day 38,128 615 9.96%
2nd–3rd 76,256 497 9.38%
4th–10th 156,990 486 9.26%
11th–30th 57,252 375 7.68%
31st or later 21,272 119 1.86%

The first item out of a bulk session performs nearly as well as a solo publish (9.96% vs. 9.33%). Everything after the thirtieth collapses to 1.86%. The cost is not in starting a batch. It is in continuing one past a couple of dozen items.

Control 4 — the games stay the same. Pooling all 24 games with enough mass-published volume to compare, the 1,000-sub rate for bulk-published items is 1.14% against 8.62% for small-batch items. The big-batch group loses in 19 of the 24 games that have enough data to judge.

The same pattern shows up in game after game, with the effect strongest exactly where mass publishing is most common:

Game Mass-published items Share of its new mods that are mass-published Mass-published hit 1,000 Solo-published hit 1,000
Trailmakers 12,056 5.2% 0.07% 0.73%
Rust 9,165 46.2% 0.08% 0.12%
People Playground 5,295 4.1% 0.51% 2.42%
Space Engineers 4,670 5.9% 0.00% 1.82%
Cities: Skylines 4,599 14.8% 3.65% 11.28%
Stormworks 4,337 5.9% 0.14% 4.42%
Brick Rigs 4,286 4.6% 1.84% 6.69%
Dungeon Alchemist 4,202 25.0% 4.95% 2.42%
Left 4 Dead 2 1,446 2.4% 9.47% 20.37%

Rust is the outlier that explains itself: 46.2% of its newly published Workshop items arrive in a mass upload. It is the engine-sandbox genre's worst offender by far, and its mass-published items convert at 0.08%.

The part that should worry everyone else

Here is where the finding stops being advice for individual creators and becomes a platform problem.

Mass dumps do not only hurt the mods inside them. They occupy the new-releases feed of their game for the hours they occupy it — and the data says everything else published into that game on that day suffers for it.

Across the whole cohort, we separated every non-batched mod by whether another creator had run a mass dump (11+ items in an hour) into the same game on the same day:

Non-batched mods published... Mods Avg subscriptions Reached 1,000 subs
On a normal day 1,951,091 901 9.85%
On a day a mass dump landed in the same game 351,648 232 3.14%

A bystander mod published on a mass-dump day reaches 1,000 subscribers 3.14% of the time. On a normal day it reaches 9.85% — a 3.1x difference. Those 351,648 bystander mods are not part of any batch. They are innocent publishers who happened to open their storefront on the wrong afternoon.

We should be careful here: dump days are not random, and games that attract mass-publishing also tend to be games with weak per-item conversion for other reasons. Across the 43 games with enough volume to compare directly, dump days are worse in 20 and better in 23 — so the pooled 3.1x gap is inflated by which games get dumped into, not purely by the dump itself. The within-game effect is real but far from universal; the within-creator effect from Control 1 and Control 2 is the solid claim.

What it actually costs to be one of the 23,202

The top of the mass-publish table is a rogues' gallery, and the pattern is identical in every row: an enormous catalogue, and almost nothing to show for it.

  • Creator 76561198375523477 has 1,223 active mods in Rust. 1,168 of them — 95.5% — were mass-published rather than released individually. Their total across the whole catalogue is 10,205 subscriptions. Their hit rate is 0.3%.
  • Creator 76561199390136535 published 386 mods, 359 of them (93%) in mass batches, and zero individually. Total: 84,010 subscriptions — carried entirely by a handful of items in a catalogue that otherwise never converts.
  • Bob Kazamakis — 401 mods, 294 mass-published, one subscription in total.

And the counter-example is exact, not anecdotal. The highest-performing mid-sized creators in the whole two-year cohort are defined by the opposite rhythm:

Creator Mods Days published Mods per publishing day Total subs Reached 1,000
KI5 (Project Zomboid) 41 37 1.11 19,059,672 100%
76561199645851972 (Wallpaper Engine) 51 51 1.00 11,207,540 100%
76561198451543154 (Wallpaper Engine) 103 103 1.00 9,611,900 100%
76561199810711673 (Wallpaper Engine) 207 70 2.96 4,025,500 99.5%

KI5 shipped 41 mods across 37 separate days and hit 1,000 subscribers with every single one. 76561198451543154 published 103 mods on 103 distinct days — one per day — and hit 1,000 with all 103. Nobody at the top of this table publishes in batches. One item per day, for a year, beats three hundred items in an afternoon.

Why a batch destroys itself

The mechanism is not mysterious, and it is not a platform penalty. Steam's Workshop surfaces new items into a "recently added" feed that is per-game, chronological and finite. When thirty items from one creator land inside the same hour, they occupy thirty consecutive slots of a feed that has only so many visitors per hour.

Two things follow. First, the items cannibalise each other — the same finite audience is split across a batch instead of concentrated on a single release, which is precisely what the position-in-batch table shows: item 31 lands at 1.86% against item 1's 9.96%. Second, the whole batch is simultaneously visible and simultaneously forgotten, because the feed's next refresh pushes all thirty out at once. A creator publishing one item per day gets 365 separate windows of front-page presence. A creator publishing 365 items in one afternoon gets one.

That is also why breadth of games helps but breadth of timing hurts. ModSignal's earlier work on the portfolio effect found that spreading mods across multiple games adds 8–10 percentage points of breakout probability, because each game is a separate discovery pool. Spreading mods across time within one game does the same thing for the same reason — it multiplies the number of independent moments you are actually visible. A batch does the exact opposite: it collapses many possible windows into one crowded instant.

What to do about it

  • If you publish in bulk, stop batching past ten. The data does not forbid multi-item sessions — items 1 through 10 hold up at 9.26–9.96%. The cliff is after item 30, where the hit rate falls to 1.86%. Ten a day for a month beats three hundred in an afternoon.
  • One item per publishing day is the winning rhythm. The top of the mid-sized creator table is nothing but one-a-day publishers hitting 100% at 1,000+ subs.
  • If you have a backlog, meter it out. A backfill does not have to be published in one sitting to be published quickly. Schedule it. The same 41 mods KI5 shipped over 37 days would have earned almost nothing as a single dump.
  • For studios, mass dumps are a moderation-adjacent signal. Rust is at 46.2% mass-published supply and Dungeon Alchemist at 25.0%. If a game's new-releases feed is being colonised by scripted floods, its genuine creators are publishing into a degraded feed.
  • And be sceptical of the bystander damage claim specifically. The pooled 3.1x figure is contaminated by which games get dumped into. The within-creator 2x gap is the one to bet on, and it is quite enough on its own.

How we measure

All figures come from the ModSignal live dataset of the Steam Workshop, snapshot 2026-09-23: 6,812,501 active mods (file type 0) holding 12,503,470,864 subscriptions, alongside 368,274 other active items. The batch analysis covers the 2,392,379 mods published in the 24 months before the snapshot by the 819,677 distinct Steam64 creator IDs with at least one such item.

"Batch size" is the count of items a creator published within the same clock hour (publication timestamps, UTC), used as the proxy for a single upload session; the daily-batch tables use the same logic at day granularity. "Reached 1,000 subs" means current subscriptions ≥ 1,000 as of the snapshot, not lifetime. Comparisons holding catalog size constant are restricted to creators with 8–60 active mods published in the window. All within-game controls require at least 500 small-batch and 200 mass-batch items in the game; all within-creator controls compare only creators who have both solo-published and mass-published output.

This is the data ModSignal tracks daily to spot early-stage mods and fading ecosystems. Explore the live dataset to see the feed for any game, or see how Rust's Workshop compares directly.