Your customer return rate is the share of this month's buyers who had bought from you before. Divide repeat customers by total customers for the month. The figure tells you whether your shop is building a base or quietly replacing its customers month after month.
The formula, and what you actually need to run it
To calculate customer return rate you need no accounting system and no analyst. You need one thing: a way to know that this customer is the same person who came last month. A phone number does it, or a digital card serial, or even a name your cashier writes down. Without an identity that repeats there is no figure at all, and that is the first wall most small shops hit.
The formula itself is one line: customers who bought this month and had bought before ÷ everyone who bought this month × 100. Note that the denominator is all of this month's customers, not last month's. Mixing the two hands you an inflated figure that tells you everything is fine.
| Item | Count |
|---|---|
| Everyone who bought during the month | 840 customers |
| Of those, had bought from you before | 218 customers |
| First-time customers | 622 customers |
| Return rate | 218 ÷ 840 = 26% |
Twenty-six per cent means three quarters of the people through the door that month had never been through it before. The cafe looks busy and the bank balance looks healthy, but what it is really doing is buying fresh customers every month instead of keeping the ones it already met.
Why the number you feel is always higher than the truth
Because you remember the faces you see, never the ones that stopped appearing. Somebody who comes four times a week walks past you sixteen times a month; somebody who came once and never returned leaves no gap you will ever notice. That one bias makes almost every owner guess a return rate roughly double the real one.
There is a second and worse error: counting receipts rather than people. Twenty visits from five customers is not twenty returning customers. If your till counts transactions you are measuring how active your handful of loyalists are, not how wide your base is — and those two numbers sometimes move in opposite directions.
What counts as a good number in your trade?
There is no single right answer, because natural buying frequency varies enormously between trades. These ranges are a starting point to measure yourself against, not a verdict:
| Trade | Weak | Good | Strong |
|---|---|---|---|
| Cafe or bakery | under 25% | 35% to 45% | over 55% |
| Restaurant | under 18% | 25% to 35% | over 40% |
| Salon or barbershop | under 30% | 40% to 55% | over 60% |
| Car wash | under 20% | 30% to 40% | over 50% |
| General retail | under 12% | 18% to 28% | over 35% |
The working rule: the shorter the natural gap between two purchases, the higher the rate you should expect. A furniture shop that sells to a household once every three years should not be measuring this monthly at all — it measures annually, or not at all.
Four mistakes that flatter the number
Count it like this
- Count people, not receipts. One customer who visited ten times is one person.
- Make the denominator every customer of the month, new and returning together.
- Fix the window: a full calendar month every time, never four weeks then five.
- Exclude staff, test orders and your own account from the count.
Not like this
- Do not compare a peak season against an ordinary month — a season brings strangers who push the rate down through no fault of yours.
- Do not rely on your cashier's impression; memory keeps only the recurring faces.
- Do not merge branches into one figure when each one serves a different neighbourhood.
- Do not celebrate a rise caused by fewer new customers — the numerator never moved.
That last one is the sneakiest of the four. Stop advertising this month and your new-customer count falls, the denominator falls with it, and your return rate climbs on its own without a single extra person coming back. Which is why the percentage must always be read next to the absolute count of returners, never alone.
Three moves that lift the number within two cycles
- Give a specific reason to return, not a general invitation "We hope to see you again" is not a reason. "Two stamps left on your free coffee" is one, because it tells the customer exactly what they forfeit by not coming back.
- Close the gap between the first visit and the second This is where most customers leak away. Every new customer should leave with something waiting for them next time, and should hear about it again before they have forgotten you.
- Measure a cohort, not a total Separate the people who joined in January from those who joined in February and follow each group on its own. A total hides decay, because arriving newcomers numerically paper over everyone who left.
The second one pays far better than the other two combined. Moving the share who come back a second time from 20% to 30% doubles your standing base within a year, because anyone who has returned once is dramatically more likely to return again.
Common questions
How often should I measure this?
Monthly for a high-frequency trade such as a cafe or restaurant, quarterly where the natural gap is longer, as in specialist salons and retail. Consistency matters more than the interval: a figure calculated the same way every time is more useful than a precise one calculated differently each time.
How is this different from a churn figure?
They are two sides of one coin with different denominators. The return rate looks at this month's customers and asks how many are repeats. Churn looks at an earlier period's customers and asks how many vanished. The first measures the make-up of your current base, the second measures how fast it leaks.
I have no system that recognises a customer. What now?
Start with the simplest identity available. A digital loyalty card solves it by itself, because every card carries a serial tied to one phone, which turns each stamp into a dated record against a known person. That alone justifies launching one even if the reward was never your first motive.