Retention is the whole problem.
Acquisition works and the average ticket is healthy. Nothing brings a first-time customer back, so revenue is capped by whatever the door brings in this month.
Seven in ten customers never come back, and in the last twelve months not one of them was asked to.
Do these three things
-
Email the lapsed list
$15,219One send to customers who already know the business and have not been asked to return.
-
Turn on rebooking reminders
$17,593 a yearCloses the follow-up loop permanently instead of one campaign at a time.
-
Fix the booking page drop-off
$5–8KTraffic arrives and leaves before choosing a service.
Try your own numbers
These open on the report's numbers. Move the sliders to see what changes.
- Win-back calculatorOn the “The lapsed list” tab
What's in this report
- Is the business growing?No. Revenue has been flat for six months — the business is running in place.
- Do first-time customers return?Mostly not. 72% of customers visited once and never again.
- Who can we win back, and what is it worth?1,461 lapsed customers are reachable today. One email to them is worth about $15,000.
- Where do people give up before booking?On the online booking page, before choosing a service. We can size it only as a range until it is measured.
Is the business growing?
No. Revenue has been flat for six months — the business is running in place.
Monthly POS revenue has moved between $31K and $37K since March. That is noise, not a trend. The door keeps bringing in new faces, but without repeat visits every month starts again from zero.
Show the numbers
| Value | |
|---|---|
| Mar | $35.2K |
| Apr | $35.6K |
| May | $34.8K |
| Jun | $37.1K |
| Jul | $33.9K |
| Aug | $31.4K |
Source: raw/square-export-2026-08.csv — 5,151 captured transactions, $386,412
Do first-time customers return?
Mostly not. 72% of customers visited once and never again.
Only 28% of a month's new customers come back for a second visit. A healthy salon keeps well over 45%. The customers who do return come back fast, which says the work itself is good — nobody is asking the others to return. There is no email, no text, and no rebooking prompt at checkout.
Show the numbers
| Value | |
|---|---|
| Active (0–30 days) | 612 |
| Cooling (31–90 days) | 917 |
| Lapsed (91–365 days) | 1,461 |
| Gone (365+ days) | 119 |
Source: analysis/segments.md
A second-visit rate above 45% would mean retention works and the ceiling is somewhere else. The measured rate is 28%.
Who can we win back, and what is it worth?
1,461 lapsed customers are reachable today. One email to them is worth about $15,000.
These customers know the business, liked it enough to book once, and have never heard from it since. That makes them the cheapest revenue available — cheaper than any new customer.
Move the sliders to test your own assumptions. It opens on the numbers this report uses.
Source: analysis/segments.md
| Move | Reached | Assumption | Payoff |
|---|---|---|---|
| Email the lapsed list | 1,461 customers | 12% return × $86.80 net | $15,219 |
| Rebooking reminders | 5,151 visits a year | 6% lift, $29/mo | $17,593 |
Source: analysis/segments.md; Square Plus pricing, 2026-08-30
The customer list lives only inside the POS vendor. Export it before building anything on top of it.
Where do people give up before booking?
On the online booking page, before choosing a service. We can size it only as a range until it is measured.
Visitors reach the booking page and leave without picking a service. With no funnel tracking in place, the lost revenue is an estimate of $5–8K a year. The first step is to measure it, not to redesign it.
The 12-month plan
Four windows, ordered by what each one unlocks. Phase 1 costs nothing and can start this week. Nothing in Phase 2 starts before the technology platform decision — if the booking system changes, reminders built on the current one are rebuilt from scratch.
0–30 days: Free foundation moves
$0Email the lapsed list
One send to the 1,461 reachable customers who have not returned in ninety days.
Alternative and how to undo it
We also considered: Paid social retargeting — four times the cost per return, and it does not build the list.
If it doesn't work: Stop sending. No contract, no lock-in.
Export the customer list out of the POS
Own the list somewhere the POS vendor does not control.
Alternative and how to undo it
We also considered: Leave it in Square — free, but the list is hostage to the vendor.
If it doesn't work: Delete the export. Nothing depends on it yet.
Ask every customer to rebook at checkout
One question at the register — "Want me to put you down for three weeks out?"
30–90 days: Paid automation
$29/moWaits for: Technology platform decision
Turn on rebooking reminders
Square Plus sends an automatic rebooking prompt at the interval each service implies.
Alternative and how to undo it
We also considered: A manual text list — free, but it stops the first busy week.
If it doesn't work: Cancel the upgrade. Month-to-month, no annual commitment.
Build a three-message welcome sequence
Thank-you, care tips, and a rebooking nudge for every first-time customer.
90–180 days: Segment and measure
$29/moSplit the list by service and visit frequency
Measure the booking funnel
Add tracking so the drop-off can be sized in dollars instead of a range.
180–365 days: Compound the loop
$29/moRun a win-back email every quarter on a fixed schedule
Do not launch reminders before the platform recommendation lands. If the booking system changes, the sequence is rebuilt.
What could go wrong
Most serious first.
| Risk | How bad | How likely | What fixes it |
|---|---|---|---|
| No follow-up after a first visit | High | High | Phase 1 — email the lapsed list |
| Single booking channel with no backup | Medium | Medium | Phase 2 |
| Customer list held only in the POS | Medium | Low | Phase 1 — export the list |
How sure we are
A second-visit rate above 45%. That would mean retention works and the ceiling is capacity. The measured rate is 28%.
Seasonality — salons see summer dips, so a flat six months could be a seasonal artifact. Rejected: the pattern holds in every quarter of the year, and a 72% once-only share is not a seasonal number.
No website or booking-page analytics, so the drop-off is sized as a range.
Where the numbers come from
- raw/square-export-2026-08.csv — 18,442 transaction rows, 5,151 attributed to a named customer
- Customer directory — 3,109 rows, pulled 2026-08-30
- Discovery session with the owner, 2026-08-14