Restaurant Retention Rate: How to Calculate Repeat Business Correctly
Calculate your restaurant retention rate with comparable guest groups, a fixed return window, and a visit-by-visit curve that shows where guests drop off.
A restaurant retention rate is the percentage of identified guests who return within a stated period after a completed visit. The most useful version is a visit-by-visit curve that shows how many guests move from visit one to two, two to three, and so on.
There is no universal “good” restaurant retention rate. Your definition, customer identity coverage, return window, service model, and guest mix all affect the number.
“We have a lot of new customers, but not repeat customers.”A restaurant owner
The restaurant retention-rate formula
Choose a group of identified guests who completed the same visit number, give all of them the same amount of time to return, and calculate:
Visit-to-visit retention rate = guests who complete the next visit within the return window ÷ eligible guests whose full return window has passed
For a clearly hypothetical example, if 200 identified guests completed their first visit and had a full 60 days to return, and 50 completed a second visit during that window, the first-to-second-visit retention rate is 25%.
Do not include newer guests who have not yet had the hypothetical 60 days. Keep them pending until their window closes. Also keep guests with unresolved identity or visit records outside the rate instead of counting them as failures.
Why one overall repeat-customer percentage hides the useful answer
An overall repeat-customer rate tells you how many people came back at least once. It does not show where the restaurant loses them.
A visit-by-visit curve can reveal that the hardest step is earning the second visit, while guests who reach a later visit are much more likely to continue. One historical POS-derived restaurant example reported 21% returning after visit one, 47% after visit two, and 64% after visit three. Those are observations from one restaurant, not industry benchmarks.
That shape changes the strategy. A restaurant with its largest loss after the first visit should concentrate on the reason to come back next. A restaurant that earns the second visit but loses people later may need a different experience, frequency cue, or loyalty milestone.
“That retention chart is huge. No one's been able to give that to us.”A restaurant marketer
Use industry numbers as context, not your target
Paytronix's 2026 Loyalty Report says 95% of guests who reach four visits keep returning. Industry coverage described a dataset of more than 225 million guest profiles across more than 800 brands, with comparisons across nine restaurant and convenience-store concepts.
That does not mean a first-time guest has a 95% chance of returning. It also does not prove that the fourth visit causes loyalty. Guests who naturally like the restaurant may be more likely to reach four visits in the first place.
Use the finding as a reason to inspect the early part of your own curve. Your restaurant's actual transition rates should determine where the biggest opportunity exists and whether an earned loyalty milestone makes sense.
A visit needs a clear definition
Retention becomes unreliable when a reservation, reward claim, order, check, and visit are treated as the same event.
Define a completed visit using the records your restaurant can support. Ideally, the visit has:
- a stable guest identifier;
- evidence that the guest actually arrived or received the completed order;
- a paid, non-voided check; and
- refunds and reversals reconciled.
Cash payments, different people paying, split checks, duplicate profiles, and POS migrations create blind spots. State those limits. If your system identifies only the payer, measure identified purchasers rather than claiming to identify every diner.
Also state the earliest reliable history date. A “first visit” in a two-year database means first recorded visit within those two years, not necessarily the person's first visit ever.
Give each guest the same chance to return
The most common retention mistake is comparing mature and immature customers.
Suppose one guest visited months ago and another visited last weekend. If the older guest's return window has closed and the newer guest's has not, only the older guest is ready to be judged. Counting the newer guest as not retained depresses the rate simply because time has not passed.
For every step in the curve, keep:
- the visit number being measured;
- cohort dates;
- the return window;
- eligible guests whose window closed;
- guests who returned inside it;
- newer guests still pending; and
- unknown or unresolved records.
Those details make the percentage reproducible without turning the calculation into a technical exercise.
Retention rate does not diagnose the whole business
A restaurant can have healthy retention and still be slow because too few new guests arrive. It can also have strong new-customer traffic and weak retention. Average order value, visit frequency, capacity, and tracking coverage can create other problems.
Compare retention with first-time guest volume before choosing the campaign. The acquisition-versus-retention guide explains how to decide which one is holding back growth.
Once retention is clearly weak, use the largest early drop to choose one next-visit test. Give that guest group a specific reason to return, keep the return window consistent, and compare mature groups again. A change in the observed rate is useful evidence, but a simple before-and-after comparison does not prove the message or reward caused it.
Use the curve to design loyalty around behavior
Points issued and members enrolled do not show whether people came back. A useful loyalty system measures visit frequency, average spend, and referrals, then uses the restaurant's own curve to decide where progress or earned status should matter.
Book a demo of Feast to connect guest history, loyalty activity, and POS-linked sales around the return behavior your restaurant needs to improve.