Reports
Retention Cohort Analysis
Last updated 19 May 2026
Retention Cohort Analysis
Monthly cohort table showing what percentage of new clients return in month 1, 2, and 3 after their first visit. The cleanest signal for whether your venue is building repeat business — or losing first-timers.
Where to find it — Web: Intelligence → Business Health zone → Retention cohorts panel. PRO only. Mobile: not yet ported.
TL;DR
- Six monthly cohorts at a time; current month excluded.
- Each row: New (count) + M1, M2, M3 return rates (%).
- Below the table: the three-cohort average — "Your 3-month retention averages X%".
- Needs 10+ billable bookings before it appears; web-only; PRO.
How cohorts are defined
A cohort = the group of clients whose first ever billable booking at your venue fell in a specific calendar month. Customer identity is keyed by email (lowercased), so the same person across two emails counts as two clients.
A booking is "billable" if its status is confirmed, completed, or arrived. Cancelled and no-show bookings are excluded — they don't seed a cohort.
What each metric measures
| Column | Meaning |
|---|---|
| Cohort | Calendar month label (e.g. "Mar 26"). |
| New | Count of brand-new customers in that month. |
| M1 | Percentage of those new customers who returned for a billable booking during their second month. |
| M2 | Percentage who returned during their third month. |
| M3 | Percentage who returned during their fourth month. |
Newer cohorts show blanks in later columns because calendar time hasn't elapsed yet:
- Last month: M1 only
- Two months ago: M1 + M2
- Three months ago and older: all three populated
Reading the table
The colour treatment helps you scan:
| Rate | Visual |
|---|---|
| ≥40% | Strong — green tint |
| 20–39% | Neutral foreground |
| 1–19% | Warning — amber tint |
| 0% | Muted grey |
Healthy salons typically see 30-50% M1 retention. Below 20% suggests something happened during or after the first visit that pushed clients away. Above 60% is exceptional — often a sign of a strong consultation flow or first-visit experience.
The averaged summary
Below the table:
"Your 3-month retention averages X% — clients who return once tend to become regulars."
X is the average of M1 rate across the three most recent cohorts that have data. It's the single number to track week-to-week — easier than scanning the whole table.
Data threshold
The panel needs at least 10 billable bookings total at your venue before it computes anything. Below that threshold, it shows: "Not enough booking history yet. Retention cohorts will appear after two months of bookings."
In practice, you'll see the first usable signal once you have 2-3 complete months of operation.
What's NOT in the table
- No per-staff breakdown — the table is venue-wide. For per-stylist retention signal, look at the Service Mix card and the Staff Performance section of Intelligence.
- No per-service breakdown — same as above. The Service Mix card shows which services drive return visits.
- No churn flag for individual clients — the separate Retention Intelligence card (also in Business Health zone) uses client health scores to flag specific at-risk individuals.
- No lifetime-value column — Cohort here tracks visit cadence, not revenue. LTV is in the Growth & Retention zone.
- No filtering or drill-down — the table is a static six-month view.
Caching
The data refreshes every 30 minutes (server-side cache) and is held client-side with a 30-minute stale-time. Changes to bookings in the last half hour may take that long to surface.
On the FREE plan
Retention cohorts are a PRO feature. FREE Intelligence shows a single "Unlock Pro Intelligence" CTA in place of the entire PRO block including this panel.
How it relates to other Intelligence surfaces
| Surface | What it tells you |
|---|---|
| Retention cohorts (this panel) | Are first-timers becoming regulars? Month-by-month percentages. |
| Retention Intelligence card | Which individual clients are at risk right now? Per-client health scores. |
| Service Mix card | Which services drive repeat visits vs one-offs? |
| AI narrative briefing | Plain-English explanation of the trend in the cohort table. |
Use them together — cohort table for the trend, Retention Intelligence for the people, Service Mix for the why.
Mobile parity
The retention cohort table is web-only today. Mobile Intelligence shows the at-risk/health-score Retention card but not the cohort table. On the roadmap.
Common mistakes
| Problem | What to check |
|---|---|
| Panel shows "Not enough booking history yet" | You need 10+ billable bookings. Cancelled and no-shows don't count. |
| Cohort row shows 0 new clients | That month was below the booking floor (no brand-new clients). Check your overall trend. |
| M1 rate dropped sharply month-over-month | Look for a service change, price change, or staff change in that month. Check feedback for the cohort. |
| Same person counted twice | Cohorts key on email (lowercased). The same person with two email addresses appears as two clients. |
| Want a cohort breakdown by service | Not supported today. Use the Service Mix card and per-staff Performance section together. |
FAQ
Where is the Retention Cohort table?
Intelligence dashboard → Business Health zone. PRO only. The table needs at least 10 billable bookings before it appears; until then, the panel shows 'Retention cohorts will appear after two months of bookings.'
What does each row mean?
Each row is a calendar month — clients whose first ever billable booking fell in that month. New = how many; M1 / M2 / M3 = the percentage who returned for another billable booking 1, 2, and 3 months later. Six months of cohorts at a time, current month excluded.
Can I filter the cohorts by staff or service?
Not today. The table is venue-wide only. Per-client retention signals live in the separate Retention Intelligence card (powered by client health scores) and per-service mix lives in the Service Mix card.
Why are my newest cohorts showing blanks?
Calendar time has to pass before M1/M2/M3 can be measured. A cohort from last month has M1 filled and M2/M3 blank; two months ago has M1+M2 filled and M3 blank. Older cohorts fill in over time.
What's a good retention rate?
Rough benchmarks for salons and barbershops: M1 over 40% is strong; 20-40% typical; below 20% needs attention. Returns at M2 and M3 should ideally hold steady (a falling pattern means people try once or twice then leave). Compare to your own past cohorts as the baseline that matters most.
Does this update when I refund or cancel an old booking?
Eventually yes — the next 30-minute refresh recomputes cohorts using current data. Cancelling a booking that seeded a cohort will remove that client from the cohort (and from any return-rate denominators).