What normal actually looks like
Spend a week reading about Skool and you will come away with an impression of the platform: six-figure months, thousands of members, traffic spikes in the tens of thousands. That impression is drawn entirely from the communities big enough to be worth posting about.
Here is the other 99%.
More than 52,000 communities
The typical community is small, and quiet
The full spread:
| Percentile | Members | Reading |
|---|---|---|
| 25th | 18 | A quarter of all communities are this size or smaller |
| 50th | 60 | The median community |
| 75th | 225 | Larger than three quarters of the platform |
| 90th | 802 | Top ten percent |
| 95th | 1,720 | Top five percent |
| 99th | 7,671 | Top one percent |
If you have 225 members you are in the top quartile of every community on the platform. At 802 you are in the top ten percent. The numbers that circulate publicly — five thousand members, twenty thousand — are the 99th percentile and above.
Most communities do not fade. They never start.
The more useful measure is not size but whether anything is happening.
| Posts, all time | Communities | Share |
|---|---|---|
| Under 10 | 11,998 | 23.0% |
| 10 – 49 | 14,784 | 28.3% |
| 50 – 999 | 21,940 | 42.1% |
| 1,000+ | 3,436 | 6.6% |
Just over half of all Skool communities have fewer than fifty posts in their entire history. Not fifty posts a month — fifty posts, ever.
That reframes the usual advice about engagement. The dominant failure on this platform is not a community that grew and then went quiet. It is one that was created, given a classroom and a cover image, and never developed a pulse at all. Around 8% of all communities have more than a hundred members and fewer than fifty posts — an audience assembled in a room where nobody speaks.
Time is not a strategy
Older communities are bigger, and the progression looks encouraging:
| Age | Communities | Median members | Median posts |
|---|---|---|---|
| Under 3 months | 2,651 | 21 | 12 |
| 3 – 6 months | 6,142 | 26 | 19 |
| 6 – 12 months | 16,494 | 36 | 29 |
| 1 – 2 years | 17,318 | 83 | 69 |
| 2+ years | 9,553 | 192 | 161 |
Median membership rises roughly nine-fold between the first quarter and the second year. But the median hides the thing worth knowing.
Of the 9,553 communities that have been running for two years or more, 35% still have under 100 members, and 28% have fewer than fifty posts. More than a third of the communities that lasted two years never got to a hundred people. Persistence is necessary and it is visibly not sufficient.
This is the number to hold onto if you are deciding whether to keep going. Not “does it get better with time” — it does, on average — but “does time on its own fix this”, to which the answer is no, roughly a third of the time.
Does building a classroom help?
Almost every owner builds one. Only 3% of communities have no courses at all; the median has five courses and twenty-six modules. So the question is whether the effort pays.
In the raw data, it looks like it does — emphatically. Restricting to communities with at least 100 members, so the empty ones do not distort it:
| Courses | Communities | Median members | Median posts | Posts per member |
|---|---|---|---|---|
| 0 | 1,036 | 265 | 78 | 0.29 |
| 1 – 2 | 2,581 | 295 | 85 | 0.29 |
| 3 – 5 | 4,304 | 314 | 127 | 0.40 |
| 6 – 10 | 5,989 | 320 | 177 | 0.55 |
| 11 – 20 | 4,840 | 326 | 287 | 0.88 |
| 21+ | 1,974 | 400 | 466 | 1.17 |
Communities with twenty-one or more courses have four times the posting activity per member of communities with none. The relationship is clean and monotone.
It is also almost certainly backwards. Nobody builds twenty-one courses for an empty room. Course count is what accumulates when an owner keeps showing up over a long period for people who are responding — which means it is measuring sustained effort and a working community, and reporting the result as evidence that content produces growth would be a straightforward causal error.
Note too what the first two rows show: going from zero courses to two moves median membership from 265 to 295 and changes posts per member not at all. The gains only appear well beyond the point most owners stop. Whatever this table is measuring, it is not “add a course, get members.”
The honest summary is that a deep classroom is a symptom of a healthy community far more reliably than it is a cause of one — and that building your fourth course before you have ten people talking to each other is very likely the wrong order.
What to do with this
- Recalibrate what failure means. At 40 members you are near the middle of the platform, not behind it. The comparison set in your head is the 99th percentile because that is who writes posts.
- Measure posts, not members. Half the platform has under fifty posts ever. Activity is the variable that separates a community from a mailing list, and it is the one most owners are not tracking.
- Do not wait it out. A third of two-year-old communities are still under a hundred members. If nothing is happening, more time is not the intervention.
- Stop building content to fix silence. The classroom data does not support it, and the first two courses move nothing. Ten people talking beats ten modules nobody opens.
The shape of the thing
The single most distorting fact about this platform is that the largest 1% of communities hold nearly 44% of all memberships. That is not a ladder most communities are climbing. It is a different population, and the distance between the median community and the ones that generate the public conversation is not a gap of effort or months — it is two orders of magnitude.
None of which makes a small community a failed one. Sixty members who show up is a real thing, and on this platform it is the ordinary thing. The reason that needs saying at all is that the feed makes it feel otherwise, and a lot of people quit at forty members believing they are the only ones there.
Caveats, including one that matters
This is a large sample, not a census. These figures cover more than 52,000 communities observed on the platform. That is a substantial slice, but it is not every Skool community and should not be read as one.
The sample is size-biased, and it flatters the platform. Communities become observable largely through their members, which means a large community is far more likely to be captured than a tiny one — a group with ten thousand members has ten thousand chances to be noticed; a group with four has four. So the quiet, small and isolated end of the platform is systematically under-counted here.
The direction of that error is worth being explicit about: the true median is probably lower than 60, not higher. Every figure above is, if anything, the optimistic version. If it reads as a bleak picture of platform-wide activity, the real one is bleaker.
Other limits. Member and post counts are lifetime totals rather than current activity, so a community that was busy last year and is silent now still looks healthy. Age is measured from creation, not from launch, and some communities sit idle before opening. And every relationship described here is a correlation — the classroom section in particular should be read as an argument against a causal story, not as one.