People search for a Skool community health score hoping for a single number — 78 out of 100, green, keep going. It would be lovely. It also would not survive contact with reality, and it is worth understanding why before you go looking for a tool that promises one.
A composite score has to pick weightings. Is a 3% participation rate with excellent retention healthier than an 8% participation rate with people churning at month two? There is no universal answer. It depends on what you charge, what you promised, and what kind of community you are running. Any product that hands you one number has quietly made that judgement on your behalf and hidden it.
To be explicit, because this article is on a vendor's site: Skoolgrades does not have a health score feature. There is no health-score page, no letter grade, no dial. What follows is the composite an owner can genuinely assemble, along with which pieces Skoolgrades actually produces.
How do I work out my participation rate?
Participation rate is the proportion of your members who did anything at all in a given window — posted, commented, voted in a poll — and it is the closest thing a community has to a vital sign.
Do it as arithmetic on your own numbers, and be careful with the denominator. If you have 600 members and 42 of them contributed something in the last 30 days, that is 7%. Whether 7% is good is a genuinely open question that depends on your context, which is the point of what counts as a good engagement rate.
Two disciplines that make this number honest:
- Use distinct people, not total actions. Ten comments from one enthusiast is not ten participants.
- Watch the rate, not the count. As a community grows, a flat participant count means a falling rate. Growth flatters raw counts and hides decay. This is the trap engagement rate exists to catch.
- Insist on participation, not visits. Participation rate counts true interactions with posts — comments, comment replies and poll votes. A visit-based engagement rate counts a member who loaded the page once and did nothing. As of July 2026 Skool's own dashboard reports engagement on a visit basis; check yours, since Skool ships changes often. The visit-based number is always the bigger and more flattering one, so a community can show a healthy engagement percentage while almost nobody is talking. Participation is the version Skoolgrades reports, and it is comments plus poll votes — per-post likes are not captured at all.
Are the people who just joined actually doing anything?
New-joiner activation is the signal most owners underweight, and it may be the most predictive thing on the list.
Of the people who joined in the last fortnight, how many did anything? Introduced themselves, commented once, voted in a poll?
A member who contributes nothing in their first two weeks has learned that this is a place they watch rather than a place they participate in. That habit is set early and it is hard to undo later. If your community looks fine overall but nobody who joined this month has spoken, you are looking at a healthy-looking present and an unhealthy future.
Concretely: of the 30 people who joined in the last 14 days, how many appear anywhere in the activity? If the answer is 3, that is your real problem, regardless of what the overall number says.
Why should I split churn by how long members have been here?
Aggregate churn is a blunt instrument, because losing people at week two and losing people at month eight mean completely different things.
- Week two loss is an onboarding or expectation-setting problem. They did not find what they thought they were joining.
- Month two to three loss is usually a value-delivery problem. They found it, consumed it, and ran out.
- Long-tenure loss is often natural graduation, and it is the least alarming of the three — though it does mean your community has an end point you should plan for.
Splitting retention by tenure is the difference between "we lose people" and "we lose people at a specific, fixable moment". Skoolgrades resolves member lifecycle across six lanes and three tenure bands — newly joined at 14 days or less, mid at 15 to 90 days, and long-tenured beyond 90 days — precisely so that distinction is visible. More on the response side in improving retention.
How much of my community depends on me?
Owner dependency is the uncomfortable signal, and no tool will volunteer it, so you have to measure it yourself.
Look at last month's posts. What share were written by you? Of the comment threads, how many started with your comment? If you disappeared for two weeks, what would still be here when you came back?
A community where the owner writes 90% of the posts and comments first on everything is not a community yet. It is a newsletter with a comment section, and it is entirely dependent on your energy. That is survivable while you are energetic. It is fatal the first time you get ill, go on holiday, or simply get bored.
The healthy shape is member-initiated threads growing as a proportion of total threads over time. It does not have to be fast. It has to be moving in the right direction.
The honest test of a community's health is what happens in it during a week you are not there. Most owners have never run that experiment and are quietly afraid of the result.
How does my community compare with similar ones?
Comparing your community with similar ones is worth exactly one signal out of five, since everything above measures you against yourself, which is where most of the value sits. But there is one useful outside reference: how does your engagement compare with communities of a similar shape?
Skoolgrades computes a peer benchmark on matured 72-hour posts, so it does not swing every time you publish something fresh. It is in Reports, which is included on every Skoolgrades plan. What it is good for is settling the "is this normal or is it just me?" question that eats an unreasonable amount of owner mental energy. What it is not is a target — a community deliberately run small and intense should look different from a large free room, and that is fine.
How do I build my own health scorecard?
Build the scorecard as a five-row table you fill in monthly, rather than chasing a single composite score. Here is the shape, with the values left for you:
| Signal | How to compute it | Direction you want |
|---|---|---|
| Participation rate | Distinct contributors in 30 days ÷ total members | Flat or rising as you grow |
| New-joiner activation | Joiners in 14 days who did anything ÷ all joiners in 14 days | Rising |
| Early churn | Members lost within 14 days ÷ joiners in that window | Falling |
| Owner dependency | Threads you started ÷ all threads | Falling |
| Peer position | Your engagement vs comparable communities | Context, not a target |
Five numbers, recorded monthly, tell you far more than any composite would — because when one moves you know exactly which lever to pull. A composite score that drops from 72 to 68 tells you to worry without telling you what about.
One caveat worth being honest about: the value here is in the series, not any single month. A first reading is a starting point and nothing more. Skoolgrades deliberately withholds trend sparklines until enough snapshot history has accrued, for the same reason — a trend line drawn from a fortnight of data is decoration, not information.
What actually causes an unhealthy reading
If your numbers come back poor, the causes are usually mundane and fixable:
- Nobody knows what to do on arrival. No clear first action for a new member.
- Posts that cannot be answered. Broadcasts with no question, or questions so broad they are work to reply to.
- Posting when nobody is around. Fixable, and covered in why your community is quiet.
- Recognition going only to the top few. If the same three names get all the acknowledgement, everyone else concludes the game is decided.
- Silence being misread. A member quiet with you may be prolific elsewhere on Skool — that is a person with the habit who has not brought it to your room, not a lost cause.
For a full pass through the community rather than a monthly reading, see how to run a Skool community audit.
So is my Skool community healthy or not?
Community health is a set of five signals rather than a verdict, and there is no health score from Skool or from Skoolgrades, and you should be wary of anyone who sells you one. What exists is a handful of signals you can compute from your own numbers, and which get more useful every month you record them.
Participation rate. New-joiner activation. Retention split by tenure. Owner dependency. Peer context. Five rows, once a month, twenty minutes. That is a real health check, and unlike a score it tells you what to do next.