Repeat purchase rate shows how many customers come back to buy again. It looks like one number. It is at least three, depending on the window and on who counts as a customer.

What is repeat purchase rate?

Repeat purchase rate is the share of customers who place two or more orders in a set period. The usual formula divides customers who bought more than once by all customers, then multiplies by 100 (Shopify). Shopify offers a variant for any purchase count: customers who bought a set number of times, divided by all customers. BS&Co, an agency, calls 365 days the typical window (BS&Co).

Our guide to measuring customer loyalty gives the same formula. It adds that the window should match the buying cycle, such as 90 days for groceries or a year for travel.

Retention is a different measure. Repeat rate counts who bought more than once in a period. Retention rate tracks how many customers stay active over time (Shopify Enterprise). A member can stay active without a second buy.

How do you calculate repeat purchase rate?

For illustration, a program has 10,000 members who bought in the past 12 months. Of those, 4,300 bought on two or more days. The repeat rate is 43%. Of 2,500 members who made a first purchase last quarter, 600 bought again within 90 days. The second-order rate is 24%. Both numbers are right. They answer different questions.

MeasureFormulaIt answersReal example
Period repeat rateCustomers with 2 or more purchases in the window ÷ customers with 1 or moreWhat share of active customers buy againEtsy’s repeat and habitual buyers
Second-order rateFirst-time buyers who order again within N days ÷ first-time buyersHow well a first purchase convertsBS&Co: 18.8% within 365 days
Renewal rateMembers who renew ÷ members due to renewWhether the relationship continuesCostco’s renewal rate
Recurring sales shareSales from subscribers ÷ net salesHow much revenue is predictableChewy’s Autoship share

Public companies show how much the definition matters. Etsy counts purchase days. Its habitual buyers spent $200 or more and bought on six or more days in 12 months. Its repeat buyers bought on two or more days (Etsy 10-K). Costco counts renewals. Its rate is a trailing calculation of renewals seven to eighteen months before the reporting date (Costco 10-Q). Chewy tracks the share of net sales from Autoship customers (Chewy 10-K).

What is a good repeat purchase rate?

BS&Co studied 156,110 customers of direct-to-consumer brands. Only 18.8% placed a second order within 365 days. Of those who came back, half did so within 30 days, and 77% bought the same product again (BS&Co). The brands span consumables, fashion and durables.

Etsy gives a larger-company view. At the end of 2025 it had 86.5 million active buyers. Habitual buyers were 5.9 million, about 7% of buyers and about 40% of 2025 gross merchandise sales. Repeat buyers, who bought on two or more days, were 34.6 million. Etsy says roughly half of active buyers shop once a year (Etsy 10-K). Add the two groups and 40.5 million of 86.5 million bought on two or more days. That is 47%.

Costco shows the top end for a paid membership. At the end of its third quarter of 2026, its renewal rates were 92.2% in the U.S. and Canada and 89.7% worldwide (Costco 10-Q). So a good rate depends on the model. A fee-based club and a fashion shop cannot share a target.

Size matters too. Ehrenberg, Goodhardt and Barwise describe double jeopardy. In any period, a small brand has far fewer buyers, and its buyers tend to buy it less often (Ehrenberg et al., 1990). A smaller brand’s lower repeat rate can be normal for its size. Compare with brands of your own scale.

How do you raise repeat purchase rate?

The payoff of retention is large if the claim holds. Frederick Reichheld’s research, as HBR reports it, shows that a 5% rise in retention raises profits by 25% to 95% (HBR, 2014). Treat that range as a headline, and test it on your own margins.

BS&Co’s data point to timing. Half of the customers who came back did so within 30 days. So the first month after a first order is the window to work (BS&Co). Because 77% bought the same product again, a replenishment reminder fits many categories.

Etsy shows another lever. In 2025 it reactivated 30.0 million lapsed buyers, meaning buyers with no purchase in a year or more. It says they have about 40% higher first-year lifetime value than new buyers (Etsy 10-K). A win-back offer can cost less than a new customer, so test it. Our entry on win-back campaigns covers the design.

How do you predict who will buy again?

A fixed window misses timing. Schmittlein, Morrison and Colombo built a model on the number and timing of past transactions. It gives the probability that each customer is still active (Schmittlein et al., 1987). Fader, Hardie and Lee later built the BG/NBD model, a variant that is much easier to run. They show its parameters can be found in Microsoft Excel (Fader et al., 2005). The forecast then feeds customer lifetime value, since managers want to predict future purchasing as an input into lifetime value.

How do loyalty programs use repeat purchase rate?

The evidence says they move some customers and not others. Yuping Liu followed a convenience store franchise. Heavy buyers claimed their rewards but did not change how they bought. Light and moderate buyers gradually bought more and became more loyal (Liu, 2007). A French retail study tested the market as a whole. Double jeopardy was present, and loyalty programs did not substantially change market structure (Meyer-Waarden and Benavent, 2006).

Identity comes first. Kroger serves about 63 million households a year. Because of its rewards program, over 95% of customer transactions are tied to a loyalty card (Kroger 10-K). A repeat rate needs that tie. Without it, a second visit looks like a new shopper.

Then pick the window and the unit. Count purchase days, not baskets, if members can split a trip. Report the second-order rate for new members and the period rate for the base. Chewy shows the subscription end of the range. Autoship customer sales were 83.3% of net sales in its 2025 fiscal year, up from 79.2% a year earlier (Chewy 10-K).

Finally, split the rate by member type. Liu’s result says to look at light buyers separately, since they are the ones a program moves. Heavy buyers will buy anyway.

How GRAVTY handles repeat purchase rate

Repeat rate depends on knowing who the shopper is. The lji.io grocery page says identity ties the shopper to the basket across email, phone, card and app (grocery loyalty). GRAVTY then records each earn, burn, tier change and offer response as a timestamped event tied to a member.

You can count any window from that record, such as the repeat rate or the second-order rate (retention economics).

Frequently asked questions

Is repeat purchase rate the same as customer retention rate?

No. Repeat purchase rate shows what share of customers bought more than once in a period. Retention rate tracks how many existing customers stay active over a time frame (Shopify Enterprise). A customer can be retained, for example by holding a membership, without buying again.

What time period should you use for repeat purchase rate?

Use a period that matches the buying cycle. Our guide suggests 90 days for groceries and a year for travel. BS&Co calls 365 days typical for ecommerce (BS&Co). Keep the window fixed so month-to-month trends stay comparable.

Why is my repeat purchase rate lower than the benchmarks I see online?

Most online benchmarks use different definitions, categories and brand sizes. Double jeopardy also means a small brand has fewer buyers who buy less often (Ehrenberg et al., 1990). BS&Co found 18.8% across 156,110 direct-to-consumer customers (BS&Co). Compare with your own past cohorts first.

Do loyalty programs increase repeat purchases?

They can, for some members. In one franchise study, light and moderate buyers bought more after a program began, and heavy buyers did not change (Liu, 2007). A French retail study found programs did not substantially change market structure (Meyer-Waarden and Benavent, 2006). Test with a holdout group.

How do you treat new customers who have not had time to repeat?

Count a customer only after the full window has passed, or group customers by the month of first purchase and compare each group at the same age. A rate that mixes old and new customers understates repeat buying. Schmittlein, Morrison and Colombo give a model that estimates the chance each customer is still active (Schmittlein et al., 1987).