Community health
Paid community audit — the three-layer evaluation most operators skip
Most paid community audits stop at the activation layer. The operator pulls new-member post rates, identifies that 40% of joiners didn’t post in week one, and starts working on the welcome DM. That is not a community audit — it is a week-one diagnostic. A complete audit adds the retention layer (why activated members cancel at months two, three, and five) and the economics layer (whether the LTV of a retained member makes the fixes worth doing at all). Operators who run all three layers discover that the activation fix is almost never the only fix, and that the retention and economics layers often reveal structural problems that week-one work alone cannot solve.
TL;DR
A paid community audit has three layers: activation (week-one post rate, root causes of non-activation), retention (cohort cancellation distribution across months one through six, programming calendar audit, relationship diagnostic), and economics (member LTV, acquisition cost vs. retention cost, break-even new-member count). Run activation weekly. Run all three layers quarterly. Sequence fixes in order — no point improving month-three programming if month-one activation is below 40%.
Why the three-layer structure matters
The three layers of a paid community audit address different failure modes, and each layer’s diagnostic tools are entirely different. Mixing them produces confusion and bad intervention decisions.
The activation layer covers the period from day zero (join) through day seven. Its primary metric is 7-day activation rate: the percentage of new members in a cohort who completed the primary value-exchange action (posting an introduction, answering a thread, or making a commitment) within seven days of joining. A community with below-50% activation is losing members at month one for structural reasons that no amount of month-three programming will fix — those members are already gone before the month-three content calendar ever reaches them.
The retention layer covers months one through six after activation. Its diagnostic tool is a tenure-segmented cancellation table: cancellations broken down by how many months the member had been paying when they cancelled. The distribution tells you which tenure window is your primary problem, and the tenure window tells you the root cause. A month-one spike means activation-quality failure. A months-two-to-three spike means programming void. A months-four-to-six spike means relationship-thin — the member never formed peer connections that made leaving costly.
The economics layer covers the business question underneath both: is the community generating enough LTV per member to justify what it costs to acquire and serve them? This layer produces the numbers that make the other two layers make sense as investment decisions. A community with $180 member LTV and $60 acquisition cost has very different intervention economics than a community with $1,800 member LTV and $60 acquisition cost — even if the activation and retention patterns look identical.
Sequence rule: Fix the activation layer before the retention layer. Fix the retention layer before the economics layer. A low-activation community that improves its month-three programming will produce slightly better retention among the minority who activated — while doing nothing for the majority who cancelled at month one. The earliest layer is always the binding constraint on every later layer’s effectiveness.
Layer one: the activation audit
Activation — who joined and didn’t engage in week one
What to pull: For each of the last three cohort months, count total members who joined and how many posted at least once within seven days. The ratio is your 7-day activation rate for that cohort. Run the same calculation for 30 days to get the lagging complement.
What the numbers mean: For paid communities at $50–$150/mo, a healthy 7-day rate is 55–65%. Below 45% is acute. For communities above $150/mo, shift those thresholds up by 10 percentage points. If your rate is declining across the three cohorts (say, 58%, 51%, 44%), you have a trend problem — something changed in the onboarding sequence, the acquisition channel, or the community’s perceived value in the last three months.
Root cause triage: Every activation shortfall has one of three primary causes. (1) No goal-specific Day 0 DM: without a direct message within two hours of joining, the member’s motivation decays before they know where to act on it. This is the most common gap and the highest-leverage single fix. (2) Channel sidebar overwhelm: a new member opening a 20-channel workspace faces analysis paralysis. Hiding the full sidebar from new members using Slack’s channel sections is a configuration change, not a content change, and typically produces 8–14 percentage-point activation improvements within one cohort. (3) No conditional Day 3 nudge: members who didn’t post after the Day 0 DM are not lost — they are available for a second, lower-barrier ask on Day 3. The nudge should go only to non-posters and should lower the activation bar (a specific thread reply, not a full introduction).
Activation audit checklist:
- Pull 7-day and 30-day activation rates for each of the last three cohort months
- Check whether a Day 0 DM fires within 2 hours of join (test with a new workspace invite to yourself)
- Count channels visible to a new member on first login (target: 4–6)
- Verify whether a Day 3 conditional nudge exists and fires only to non-posters
- Compare activation rates by acquisition channel if you have multiple (referral vs. list vs. direct search) — a low-activation channel inflates the aggregate problem
Layer two: the retention audit
Retention — why activated members cancel at specific tenure milestones
What to pull: From your billing tool, export all cancellations from the last six months with the member’s original join date. Calculate tenure-at-cancellation (months from join to cancel) and build a frequency table. You should see which tenure month has the highest cancellation concentration.
| Cancellation spike at… | Primary root cause | Diagnostic question |
|---|---|---|
| Month 1 | Activation quality — member technically activated but never found value in the first post | What % of month-1 cancellations have a message count > 0? If below 50%, this is an activation problem, not a retention problem |
| Months 2–3 | Programming void — no regular touchpoints after the onboarding window closes | Is there at least one response-requiring content touchpoint per week (a prompt thread, a live Q&A, a member feature) in months two and three? |
| Months 4–6 | Relationship-thin — member has no peer connections that make leaving costly | Do month-four-to-six cancellations show DM activity (i.e., did the member have any direct peer conversations)? No DM activity = relationship-thin signal |
| Month 12+ | Value drift — the community’s core value proposition is no longer matched to where the member is now | Did members who cancelled at 12+ months reduce posting frequency in the three months before cancelling? A quiet period before cancellation is a value-drift signal, not a programming problem |
Programming calendar audit: For months two and three specifically, review your content calendar. Count the number of response-requiring touchpoints per week: a question thread, a live session, a resource request, a member spotlight. A healthy programming calendar in months two and three has at least one response-requiring touchpoint per week. Broadcast posts (announcements, links shared without a question) do not count — they require no response and do not build the engagement habit that drives renewal at month three.
Relationship-thin diagnostic: The months-four-to-six cancellation spike almost always reflects members who never formed peer-to-peer connections inside the community. They engaged with the operator-produced content but never had a direct conversation with another member. Check this in Slack: go to workspace admin and look at direct message counts for members who cancelled in their fourth through sixth months. A member who cancelled with zero direct messages is a relationship-thin signal. The fix is not more content; it is facilitated peer connections: small-group sessions, peer matching on stated goals, or direct introductions from the operator.
Retention audit checklist:
- Export cancellations from the last six months with join date; build tenure-at-cancellation table
- Identify the tenure month with the highest cancellation concentration
- For month-1 cancellations: check what % had message count > 0 in Slack
- For months 2–3 cancellations: count response-requiring content touchpoints per week in that period
- For months 4–6 cancellations: check direct message counts for each cancelled member
- Note whether overall monthly churn rate has increased, decreased, or held steady over the last six months
Layer three: the economics audit
Economics — whether the LTV math justifies the fix investment
What to pull: ARPU (average revenue per user per month = MRR ÷ paying member count), average tenure in months (= 1 ÷ monthly churn rate expressed as a decimal), and your variable cost to serve one member per month (Slack fees allocated per member, tool costs, operator time at an hourly rate).
LTV formula: Member LTV = ARPU × average tenure − (cost-to-serve × average tenure). For a community at $99/mo ARPU with 8-month average tenure and $12/mo cost-to-serve: LTV = ($99 × 8) − ($12 × 8) = $792 − $96 = $696 per member.
Acquisition cost vs. retention cost comparison: If your acquisition cost per member (paid ads, referral incentives, or time spent on outreach at an hourly rate) is higher than your monthly cost to retain a member, you are in the standard community economics position: it is cheaper to retain than to replace. The audit question is by how much. If acquisition cost is $80 and monthly retention cost is $12, you need a member to stay for at least 7 months before their LTV exceeds what it cost to acquire them. Below 7-month average tenure, every cancelled member is a loss on the acquisition investment — and every improvement to retention that extends average tenure from 6 to 9 months converts a losing investment into a profitable one.
| Activation rate | Avg tenure (mo) | LTV @ $99/mo | Net LTV (after $80 CAC) |
|---|---|---|---|
| 30% (acute) | 3.5 | $303 | $223 — barely covers 2.8× CAC |
| 50% (typical) | 6.0 | $522 | $442 — 5.5× CAC |
| 65% (strong) | 9.5 | $826 | $746 — 9.3× CAC |
| 80% (exceptional) | 14.0 | $1,218 | $1,138 — 14.2× CAC |
The table illustrates the compounding relationship between activation rate and LTV: moving from 50% to 65% activation increases net LTV by 69% — not because you’re charging more or reducing acquisition cost, but because activated members stay longer and produce more months of revenue. The activation investment pays for itself in an accelerated way that programming improvements or acquisition efficiencies cannot match.
Break-even new-member count: How many new members per month does the community need to break even on its operating costs (tool costs plus operator time)? This is total monthly cost ÷ ARPU. For a community with $500/month in total costs and $99 ARPU, you need 6 new members per month at 100% retention — or 10 new members per month at 60% first-year retention (because 40% will cancel before their LTV exceeds the serving cost). The break-even analysis tells you whether your current acquisition rate is structurally sustainable or whether you need to either grow faster or reduce costs.
Economics audit checklist:
- Calculate current ARPU, monthly churn rate, and average tenure in months
- Calculate member LTV using the formula above
- Estimate average acquisition cost per new member (time + paid + referral costs)
- Check whether LTV ÷ CAC is above 3.0 (the healthy threshold for subscription businesses)
- Calculate break-even new-member count at current churn rate
- Model the LTV impact of a 10-percentage-point activation improvement at current ARPU
Running the three layers as a quarterly routine
A quarterly audit covering all three layers takes under four hours for a community with clean data. The first audit takes longer because it requires building the tenure-segmented cancellation table from a raw billing export — but once that template exists, subsequent audits are primarily a matter of updating numbers and re-running the analysis.
Monthly between audits: Run the activation layer only. Every Sunday, check which members from the last seven days have not posted and add them to the Day 3 nudge queue. This 15-minute weekly review is the operational mechanism that keeps the activation layer from degrading between quarterly audits. A community that runs the quarterly audit but skips the weekly review will produce slightly worse numbers at each subsequent audit because activation problems compound: the new members who were in the unreviewed weekly queues become the month-one cancellations in the next quarterly retention table.
When to run an unscheduled audit: Three triggers should prompt an immediate audit regardless of cadence. First: a month-over-month increase in monthly churn rate of more than two percentage points. This usually signals an acquisition-channel shift that brought in lower-motivation members, a programming void that recently appeared, or a pricing change that altered member expectations at signup. Second: a month-over-month decline in week-one activation rate of more than five percentage points. This usually means a change in the onboarding sequence or a new acquisition source landing misaligned members in the community. Third: a significant change in average ARPU — either from a pricing experiment or a shift in plan mix — because the LTV math changes materially when ARPU shifts by more than 20%.
The most common audit mistake: Starting with the economics layer and working backwards. The economics numbers are outputs of the activation and retention health, not inputs to fixing them. Starting with “our LTV is too low” and jumping to pricing experiments skips the upstream root causes that are driving the low LTV. Run the layers in order: activation first, then retention, then economics. The economic diagnosis tells you how much the fixes are worth — but it can’t tell you which fixes to make.
For the specific metrics the activation and retention layers feed into on a weekly basis, see the Slack community health metrics guide — six numbers at three tracking frequencies. For the activation rate calculation in detail — cohort-based formula vs. the aggregate snapshot that most operators use by mistake — see paid community member activation rate. For what to do when the retention layer flags a months-two-and-three cancellation spike specifically, see the Slack community content strategy guide and the community membership cancellation rate breakdown.
Frequently asked questions
What is a paid community audit?
A paid community audit is a structured evaluation of a paid subscription community across three layers: the activation layer (how well new members progress through the first week, measured by 7-day and 30-day cohort activation rates), the retention layer (why activated members cancel at specific tenure milestones, diagnosed through a tenure-segmented cancellation table), and the economics layer (whether member LTV justifies acquisition and serving costs, calculated from ARPU, average tenure, and cost-to-serve). Most informal audits cover only the activation layer. A complete three-layer audit is what distinguishes a temporary fix (improving the welcome DM) from a structural understanding of why the community generates the revenue it does.
How often should you run a paid community audit?
Run the full three-layer audit quarterly. Between audits, maintain a weekly activation review: every Sunday, check which members from the last seven days have not posted and send a Day 3 conditional nudge to anyone still in the intervention window. The weekly review is the operational layer that prevents the activation metrics from degrading between quarterly audits. Run an unscheduled audit immediately if monthly churn rate increases by more than two percentage points month-over-month, if 7-day activation rate drops by more than five percentage points in a single cohort, or if ARPU shifts significantly due to a pricing change or plan-mix shift.
What numbers do you need to run a paid community audit?
For the activation layer: new-member count and post-count by join date for each of the last three cohort months (from Slack workspace admin). For the retention layer: cancellation date and join date for all cancellations in the last six months (from your billing tool — Stripe, Memberstack, or equivalent). For the economics layer: ARPU (MRR ÷ paying member count), average tenure in months (= 1 ÷ monthly churn rate as a decimal), and variable cost-to-serve per member per month. All six data sources are available to any operator with access to Slack admin and their billing tool. A complete first-run audit from raw exports typically takes three to four hours.
What is the difference between a community health check and a community audit?
A community health check is a point-in-time snapshot of current leading indicators — activation rate, weekly active poster rate, and current-month cancellation rate. It tells you whether things are healthy right now. A community audit is a retrospective analysis across multiple cohort months that identifies structural patterns: where members are exiting the retention curve, why, and whether the economics support fixing it. The health check answers “what is the current state?” The audit answers “why is the retention curve shaped this way, and what should we fix first?” For communities above six months old with real cohort data, the audit is the more actionable exercise — it distinguishes between activation failures, programming failures, and relationship failures that the health-check metrics cannot separate.
What is the retention layer in a paid community audit?
The retention layer covers the period after week-one activation through month six. It asks: of the members who passed the activation gate, why are some cancelling at specific tenure milestones? The primary tool is a tenure-segmented cancellation table: cancellations by month of tenure, read against root-cause patterns. Month-one spikes indicate activation-quality failure (the member activated technically but never found real value). Months-two-to-three spikes indicate programming void (no regular response-requiring touchpoints after the onboarding window closes). Months-four-to-six spikes indicate relationship-thin failure (the member never formed peer connections that made leaving costly). Each root cause has a different fix — which is why the tenure distribution is essential: it tells you which fix to prioritise before you spend time building it.