Most communities are mostly silent. That is true of every online space anyone has measured, and it is true of yours. The useful questions are how mostly, and what happens on the posts themselves — because the answers to those two are very different in tone.
The definition
Untapped share is the proportion of a community's members who have never participated in any visible way — never posted, and never appeared among a post's listed contributors.
"Visible" is load-bearing. This is a measurement of what an outside observer can see, and it is bounded by what Skool exposes anonymously.
Why it is a range and not a number
Anonymous access shows post authorship exactly. Comment authorship it shows only partially: each post lists its top five contributors, and no more. So for any community there are two defensible counts.
- The lower bound on untapped share counts posters plus every listed contributor as having participated. It assumes we saw everyone who commented, which we did not.
- The upper bound counts only posters as having participated. It assumes nobody commented who did not also post, which is also false.
The true figure lies between them. Publishing a single number would mean picking one of those two assumptions and not saying which, so both are published instead.
| Community | Untapped share is at least | And at most |
|---|---|---|
| 50–199, free | 75.9% | 92.0% |
| 50–199, paid | 77.9% | 92.4% |
| 200–999, free | 83.3% | 92.5% |
| 200–999, paid | 78.5% | 91.2% |
| 1,000–5,000, free | 87.7% | 94.8% |
| 1,000–5,000, paid | 82.3% | 89.0% |
| 5,000+, free | 89.7% | 94.4% |
The modeled point inside the range
A point estimate of roughly 70–75% can be derived by taking the upper bound and subtracting the comment-only share measured in communities where full data is available — about 21 points.
That derivation imports a quantity measured in one kind of community into a bound measured in another, which is why it is always labelled modeled, always stated with its method attached, and never printed on its own. Those communities skew engaged, so their comment-only share is probably too high, which pushes the modeled figure lower than the truth rather than higher.
The number that changes how the range reads
A large untapped share tells you how many people are quiet. It says nothing about the people who are not — and that is where Skool looks unusual.
Using the same formula as the Rival IQ / Quid 2026 Social Media Industry Benchmark Report — interactions on a post divided by audience size — the communities we track run a median of 5.1% of members per post, across 24 communities and 3,110 posts over 90 days. The same report puts the Instagram median at 0.30% of followers per post, and organic Facebook between 0.02% and 0.23%.
The mismatch, up front. Those are brand accounts posting to followers; these are community owners posting to members. Organic reach on both platforms is also actively suppressed. The formula is identical; the situations are not. And this Skool figure comes from communities using Skoolgrades, with full engagement data available by consent — not from the public panel, and not a platform-wide claim.
The size confound is real, and here is the gradient. Dividing interactions by member count inflates the rate in small rooms. The bands show it plainly:
| Community size | Median per-post engagement | Communities measured |
|---|---|---|
| 25–99 members | 13.5% | 6 |
| 100–499 | 4.0% | 15 |
| 500–999 | 3.6% | 2 |
| 1,000+ | 0.61% | 1 |
The gradient is published rather than the headline alone, because the headline on its own would be doing the small communities' work for them. What survives the confound is that every band clears Instagram's median, including the single community above 1,000 members at 0.61%. So does every individual community measured: the quietest of the 24 came in at 1.63%.
Two things are true at the same time, and most owners only ever hear the first. Most of your roster has never said anything. And the members who are there respond at rates the platforms they came from do not reach.
The distinction this metric cannot yet make
Untapped share counts members who have never participated. It does not distinguish between two very different people:
- someone present on Skool and ignoring this community, and
- someone gone from Skool entirely, whom no action by any owner can reach.
Only the first group is addressable. The second is a permanent floor on what any engagement strategy can achieve, and treating them as the same population is how owners end up measuring themselves against a target that was never available. Splitting them is what the show-up rate is for.
How to use it
Compare against your own band, not against a headline. The bands above differ by size and by whether the community has a paid door, and the differences between them are smaller than the width of any individual range — which is itself the finding. A roster that is 85% untapped is an ordinary Skool community, not a failing one.
The number to act on is not the share. It is the subset of that share you can name, and the smaller behavioural groups inside it: members who posted once and stopped, and members who are active elsewhere on Skool while silent with you, which is what pull rate measures. Both groups are small enough to work through by hand. The untapped share never will be, and it does not need to be.