The single most common mistake in reading community benchmarks is comparing yourself against a platform average. Nearly every figure here moves substantially with size, so here they all are, banded.
Contribution: who has ever posted
Mild drift, not collapse. Growth changes the number of people posting and leaves the participating fraction close to alone — a 5,000-member community runs roughly the contribution rate of a 150-member one. Full method in creator rate.
Attention: minutes online per member per week
The steepest gradient in the whole dataset. A member of a small community gives it four times the attention a member of a large one gives theirs. The platform-wide figure of 37 minutes sits near the bottom of this ladder because 83.1% of all members are in the largest band — how much time members spend explains why any single pooled figure misleads.
The engine: people carrying three quarters of the posts
The count scales with membership; the share stays near 2.15% throughout. Members grow 62-fold across these bands and the engine grows about 50-fold. Who actually writes the posts has the confound checks.
The owner's share of all posts
The number that most distinguishes a broadcast from a community. Below 200 members the owner writes over half of everything; the communities that grew are the ones where this fell. See how many admins a Skool community has.
Engagement per post, and the Instagram multiple
A factor of 22 across the range, because the rate divides interactions by audience. Every band still clears Instagram's 0.3% median — Skool vs Instagram shows the comparison honestly, gradient included.
The clock: does the community go quiet overnight
About one small community in five goes dark overnight. Of 101 communities over 5,000 members, not one does. That zero is a count, not an estimate — do Skool communities ever go quiet has the confound control.
Model: how common is a free door with paid rooms behind it
The plurality model at every size, and it leads show-up rates in all four bands. Free, paid or freemium covers the choice.
How to use this page
Find your band and read down. If you are at 300 members, your reference row is 200–999 everywhere: about 7.5% having ever posted, roughly 82 minutes of attention per member per week, an engine of about nine people, and an owner writing something like 39% of posts.
Treat a big miss as a question, not a verdict. A figure well below your band's median is worth investigating; one slightly below it is noise. These are medians across hundreds of communities, and half of all communities sit below each one by construction.
Then stop comparing outward. The only comparison you fully control is against your own numbers a month ago, and it is the only one that tells you whether what you are currently doing is working.
Scope. Contribution, engine, clock and model figures come from a public panel of 688 to 688-plus communities with complete crawls. Attention, per-post engagement and the owner's share come from communities that granted direct access — owners who bought analytics, who skew more engaged than a random sample. The two instruments are never blended into a single figure, and each chart above draws from one of them.
These figures come from The State of Skool 2026, an independent measurement of 18,412 Skool communities. See the full report for the complete method and 21 other findings.