What Is Customer Retention? Definition, Formula, Benchmarks

If 7 out of every 100 loyalty members you had in January are gone by March, you have a 93% retention rate and a problem worth looking into. Customer retention is the share of existing customers who keep buying from you over a set period, rather than lapsing or switching to a competitor. You calculate it as a rate: customers at the end of the period, minus new customers, divided by customers at the start, times 100. It tells you whether the customers you already won are staying, which matters more to most retailers and QSR chains than winning new ones.

This page covers the definition, the formula with a worked example, a benchmark table by sector, what drives retention up or down, and how a loyalty programme fits into the picture. If you run loyalty or CRM for a brand with 50 or more locations, the numbers below should map onto what you see in your own data.

Customer retention, defined

Customer retention is a company's ability to keep existing customers buying from it over time. Retained customers come back for repeat purchases, tend to be less price sensitive, and often recommend the brand without being asked. The point is not one extra transaction. It is a relationship that makes revenue more predictable quarter to quarter.

Retention matters wherever switching is easy: grocery, fuel retail, fashion, quick service restaurants. A fuel retailer with a station on every corner competes on the same basis as the one two streets over, so the brand that keeps drivers coming back wins on margin, not just on volume. You track retention to see whether your loyalty programme, CRM and in-store experience are pulling in the same direction.

For the broader concept, see what a loyalty programme is and how it differs from the outcome you're measuring here.

Retention versus loyalty, and versus churn

Retention is the measurable outcome: the percentage of customers who keep buying. Loyalty is the attitude behind it, how strongly customers prefer your brand and whether they'd recommend it. You build loyalty through consistent experiences and rewards that feel fair; you see the result in your retention number.

Churn is simply the mirror image. If your retention rate is 93%, your churn rate is 7%, and that 7% is the group your CRM and store teams should investigate first, before they lapse quietly and for good.

How to calculate customer retention rate

The formula:

Customer retention rate (%) = ((E - N) / S) x 100

E is customers at the end of the period, N is new customers acquired during it, S is customers at the start. Subtracting new customers is the step teams most often skip, and skipping it is how a strong acquisition month hides a bad retention month.

A fashion chain starts the quarter with 10,000 active loyalty members (S). It ends the quarter with 10,800 members (E), of which 1,500 joined during the quarter (N). ((10,800 - 1,500) / 10,000) x 100 = 93% retention. So 7% of the members it had in January didn't make it to March. These figures are illustrative, not a benchmark, but the method is exactly what you'd run on your own loyalty data.

What counts as a good retention rate, by sector

There's no single good number, because purchase frequency differs wildly by category. A grocery shopper visits weekly out of necessity; a fashion shopper might buy four times a year by choice. The table below gives directional ranges, not hard benchmarks, to help you judge your own number in context.

Rather than chasing an industry number, compare your own rate quarter on quarter and by cohort, by sign-up month or loyalty tier. That tells you whether a recent change to your programme helped or hurt, which is the only comparison that actually changes a decision.

National survey data from the UK found that 97% of shoppers belong to at least one supermarket loyalty scheme, yet many still distrust how loyalty pricing works (NatCen, 2024). High enrolment doesn't automatically mean high trust, and trust is what keeps the retention number up once the novelty of joining wears off.

What actually moves the number

Deloitte's 2025 consumer loyalty survey found that members who feel a programme understands them personally are markedly more likely to keep buying from that brand (Deloitte Insights, 2025). That lines up with academic work on AI-driven personalisation, which strengthens the link between trust, satisfaction and loyalty in e-commerce settings (Future Business Journal, 2025).

Gamified mechanics, missions, progress bars, tier upgrades, also lift repeat purchase, but only within limits. A 2026 study in the Journal of Retailing and Consumer Services found gamification effects are non-monotonic: push the challenge too hard and engagement drops instead of rising (ScienceDirect, 2026). A field study of nearly 19,000 app users found the same pattern in practice, that well-paced missions beat long, demanding ones (Journal of Marketing Research, 2025). A handful of active missions per member, mixing a quick win this week with a longer-term tier goal, tends to outperform one long checklist.

Fashion retail is its own case. A 2025 systematic review of 201 articles on fashion loyalty found that paid-tier benefits only work if members can see the value constantly, not just at sign-up (ResearchGate, 2025). Compare that to examples from real fashion and retail programmes to see how brands keep that visibility up between seasonal purchases.

Setting up a programme without losing control of the cost

Buyers ask us a version of the same question on almost every call: who owns the member experience once it's live, and what does the total cost look like per location and per member, not just the setup fee. Those are fair questions, and the honest answer is that ownership should sit with your loyalty or CRM team, with IT limited to integration and security review, not day-to-day campaign changes. A platform that needs IT to approve every mission or coupon will slow you down exactly when a campaign needs to launch.

On cost: a workable approach is to set a three-year cost cap tied to expected retention or basket-size gains, reviewed every six months against actual member growth and redemption rates, rather than committing to a fixed multi-year spend upfront. Expiry rules on points or stamps, and caps on how much a member can earn per visit, should be decided before launch, not patched in after members start gaming the system. NeoDay's clients typically see programme launches in as little as 10 weeks, with minimal IT effort because the platform is API-centric and sits alongside existing POS and CRM systems rather than replacing them.

If you're building the internal business case for a new or renewed programme, that's the point to bring in actual numbers: current retention rate, cost per acquired customer, and what a 5-point improvement in retention would be worth over a year. We can help you build that case with a short working session rather than a generic deck; book a demo and bring your own numbers.

Where retention data actually comes from

Your CRM and loyalty platform are the two sources that matter. A narrative review of supermarket loyalty-card data found it reliably reveals shifts in purchase behaviour well before a customer formally churns, which is the whole point of watching cohorts rather than one blended average (PMC, 2025). If a cohort that joined in March is dropping off faster than the one that joined in January, that's a signal to check what changed in the onboarding flow or welcome offer between those two months.

Read more in what customer retention is and why it matters, or look at loyalty examples across industries to see how different verticals structure their retention mechanics.

Frequently asked questions

How often should I measure customer retention? Monthly if you have enough volume to make the number meaningful, quarterly if you're a smaller chain. Measuring too rarely hides the cohort that started drifting two months before the lapse shows up in the aggregate number.

Does a loyalty programme guarantee better retention? No. A meta-analysis spanning four decades of loyalty research found programmes reliably lift repeat-purchase behaviour but change attitudes less reliably, so a badly designed programme can still fail to build real preference (Journal of the Academy of Marketing Science, 2021). Design, not the existence of a programme, is what decides the outcome.

What's a realistic timeline to see retention improve after launching a programme? Most brands see early signals, redemption rates, repeat visit frequency, within the first full quarter after launch. A meaningful shift in the retention rate itself usually takes two to three quarters, once enough cohorts have passed through a full purchase cycle.

Should I track retention per channel, like app versus in-store? Yes, if your brand sells through both. Omnichannel research shows purchase patterns and switching behaviour differ meaningfully between channels, so a blended retention number can hide a weak app experience propped up by strong in-store habits (PLOS One, 2024).

Is a high enrolment rate the same as high retention? No. Nearly all UK shoppers belong to a supermarket scheme, but that says nothing about whether they keep engaging with it or trust how it prices offers (NatCen, 2024). Enrolment is a vanity number; cohort retention is the one that reflects whether people stay.

Sources: Deloitte Insights, Journal of the Academy of Marketing Science, NatCen, PMC: supermarket loyalty-card data, PLOS One / PMC, Future Business Journal, ScienceDirect: non-monotonic gamification, Journal of Marketing Research, ResearchGate: fashion retail loyalty SLR

More questions about loyalty programs? Find the answers in our loyalty FAQ.