Customer Retention Rate

The standard formula

Retention rate = (Customers at end − New customers acquired during the period) ÷ Customers at start × 100.

Subtracting new customers is the step people skip, and it matters: without it, growth disguises churn. A business that starts with 200 customers, ends with 220, and acquired 40 new ones along the way did not retain 110%. It retained (220 − 40) ÷ 200 = 90%.

Worked example

A med spa begins Q1 with 200 active clients. During the quarter it gains 40 new clients and ends with 220 active clients.

Retained clients = 220 − 40 = 180. Retention rate = 180 ÷ 200 × 100 = 90%. The 20-client shortfall (200 − 180) is the quarter's churned base. Check the arithmetic backwards: 200 started − 20 lost + 40 new = 220 ended. It reconciles, always reconcile; a retention figure that cannot be reconciled to start, end, and new counts is not a measurement.

Retention vs repeat purchase rate

Repeat purchase rate: the share of customers who bought more than once, is simpler but weaker: it ignores time. A customer who bought twice in 2019 and never again counts as “repeat” forever. Retention rate fixes a period and asks who is still with you, which is why it suits subscription and membership thinking. For appointment businesses, both are usually inferior to return-by-deadline measures (see rebooking rate), because “still a customer” is vague when no contract exists: the expected return window makes it concrete.

Retention vs churn

For a fixed starting cohort over a fixed period, churn rate = 100% − retention rate. Our example: 10% quarterly churn. The identity breaks the moment cohorts or periods shift, blended churn across overlapping cohorts is not simply one minus blended retention, so state the cohort and period every time you quote either number. Annualizing quarterly churn by multiplying by four is also wrong (churn compounds); convert properly or do not convert at all.

Choosing the period

The period should match how often customers naturally decide: monthly for memberships, quarterly for typical med spa cadences, annually for slow-cycle services. Too short a period and normal gaps look like churn; too long and real losses hide inside the average. When in doubt, compute two periods (e.g. quarterly and annual) and report both, disagreement between them is itself informative.

Cohort considerations

Blended retention mixes January joiners with June joiners and calls it one number. Cohort retention, grouping customers by when they started and tracking each group, reveals whether retention is improving or the mix is merely shifting toward newer, not-yet-churned customers. Any retention claim used to argue “we are getting better” should be cohort-based; blended numbers can improve while every cohort worsens, the classic plane-crash-survivorship pattern of growth-stage businesses.

Why appointment businesses are different

No contract, no cancellation event, no “still subscribed” flag: a med spa client simply stops booking. Retention therefore has to be inferred from return timing against expected windows, which is a modeling choice, not an observed fact. Be explicit about the choice: retained = returned within X days of the expected date, for service category Y. RetentionTek's score methodology publishes exactly these choices (180-day first-return window, per-client expected deadlines, 30-mature-customer minimums) so the number can be interrogated rather than admired.

Limitations of a single number

A single retention rate hides who is leaving (new vs established clients), why (experience, price, relocation, death-by-calendar), and whether the remaining base is healthy or merely inert. It also says nothing about value, retaining 90% of clients who visit half as often is a quieter business than the number suggests. Use retention rate as the headline, then immediately segment: by cohort, by service, by practitioner, by location. The headline tells you whether to worry; only the segments tell you what to do.

And never trust a retention figure without its companions: period, cohort definition, new-customer handling, and data coverage. A vendor or article that omits all four is sharing a vibe, not a metric.

How the Retention Score differs

A basic retention rate is one ratio over one period. RetentionTek's Retention Score is a composite built for appointment reality: 40% first-to-second visit conversion (new-client leakage, where med spas bleed most) plus 60% established return adherence (deadline-based return across the base), computed from completed visits and withheld below 30 mature customers per component. Think of retention rate as the vital sign and the score as the diagnosis, related, but the score says where to act.

Frequently asked questions