Marketing advice says to build a persona. The data says your roster does not have one — it has a flat spread across every level of activity, with no bulge anywhere for a persona to sit on.
How active is a typical Skool member?
The median member is active 26 days a year. The mean is 81.
A gap of more than three times between the two, which is the statistical signature of a long tail: a handful of people are enormously more active than everyone else, and a mean adds their totals into a pot divided by everybody. One member with 9,000 actions outweighs eighty members with five each.
Whenever a median and a mean diverge like that, the mean is the wrong number. Quoting 81 days would describe a member who barely exists.
Where do members actually cluster?
They do not. That is the finding:
Every bucket holds roughly a tenth of the membership, and the buckets are doublings — 2–3 days, then 4–7, then 8–15, all the way to 256+. A flat distribution across a logarithmic scale means members are spread across orders of magnitude of engagement with no clustering at all.
There is no hump. There is no typical case. Your community contains, in roughly equal numbers, people who have never shown up and people who are there almost daily, and every gradation between.
The same shape appears in time online, where the spread runs from under 10 minutes a week at the 10th percentile to over 11 hours at the 99th. How much time members spend on Skool has that full distribution.
What does this break?
Personas. "Our typical member is a busy founder who checks in twice a week" describes one slice of a roster that is evenly spread across ten slices. Content built for that person lands for a tenth of your community.
Averages in reporting. Any mean you compute across your membership is dominated by the top bucket. Average time on site, average posts per member, average sessions — all of them describe the 10.6% who are active 256+ days a year, who also generate 68.4% of all activity. Activity concentration is the platform-level version of that arithmetic.
Single-track onboarding. A welcome sequence assumes a rate of consumption. With members spread across three orders of magnitude of activity, no single pace fits — it is too slow for the top decile and far too fast for the bottom half.
What should you do instead?
Segment by behaviour, not by demographics. The useful divisions are behavioural and they are already in your data: never participated, commented only, posted once, posts regularly. Those groups need genuinely different things, and each has a defined next step. Member segmentation is the practical version.
Design for at least three speeds. Something for the people who are there daily, something that works if you check in weekly, and something that still lands if you appear once a month. That is not three times the work — it is usually the same material with different entry points.
Judge your content by which segment moved. A post that delights your core and reaches nobody else is not a failure, but it is not growth either. The number worth watching is whether anyone crossed a line: a first comment, a first post. Those are the transitions that change a member's trajectory, and they are measurable in a way "engagement" is not.
And measure the crossing you can most cheaply cause. Roughly 21.3% of members comment without ever posting — a group that has already decided the room is worth speaking in. Moving them is one nudge; moving a silent member is a campaign. The participation rate covers where that group sits.
The scope worth stating: these are platform-wide member activity records, measured in days active per year. "Active" means seen on Skool rather than active specifically in one community, so a member counted in the top bucket may be spending that activity somewhere else entirely — about 71.5% of a typical roster's Skool activity happens outside the community measuring it.
These figures come from The State of Skool 2026, an independent measurement of Skool communities. See the full report for the full member analysis and 21 other findings.