In short
- Formula: customers of a cohort who order again within N days ÷ new customers of that cohort × 100. The cohort and the window make it comparable
- Benchmark: 13.45% of new customers order again within 90 days at the median D2C brand in our panel (June 2026 cohort)
- Not the same as retention rate (still active over time) or returning customer rate (the share of a period's customers or orders from people who ordered before)
- From rate to orders: if every order has the same chance of a next one, orders per new customer = 1 ÷ (1 − repeat rate), and the orders decide when a customer pays back their CAC
- Lags by design: a 90-day rate is final 90 days after the cohort month
How repeat purchase rate is calculated
Repeat purchase rate follows one group of new customers, a cohort, and asks how many of them order a second time within a fixed window:
Repeat purchase rate = customers of a cohort who order again within N days ÷ new customers of the cohort × 100
Example: in September, 1,200 customers place their first order with a supplement shop. Within 90 days of their first order, 150 of them order again. The September repeat purchase rate is 150 ÷ 1,200 = 12.5%.
Three choices decide the number, and a figure without them can’t be compared:
- The cohort. New customers grouped by the month of their first order. A rate over all customers mixes buyers from last month, who had no time to return, with buyers from three years ago, who had all the time in the world.
- The window. 30, 60 or 90 days, or 12 months. Our panel uses 90 days, so a cohort’s rate is final only 90 days after its month ends. The same shop as above could show 31% as “customers with two or more orders ÷ all customers ever”, which is a different metric.
- What counts as an order. Subscription renewals are orders, so a monthly subscription is a repeat purchase by design within 90 days. Whether cancelled orders and exchanges count depends on the report, and orders on a marketplace aren’t in a shop’s cohort at all.
Repeat purchase rate vs. retention vs. returning customer rate
Three metrics, three denominators, often used as if they were one:
| Metric | Counts | Out of | Answers |
|---|---|---|---|
| Repeat purchase rate | New customers of a cohort who order again within N days | The cohort’s new customers | Do new customers come back? |
| Retention rate | Customers of a cohort still buying or subscribed in a later period | The cohort’s customers | How long do customers stay? |
| Returning customer rate | A period’s customers or orders from people who ordered before | All customers or orders of the period | How much of this period’s business is repeat business? |
The returning customer rate is a mix metric, the mirror of the new-customer share, and the retention entry’s comparison of retention, repeat purchase rate and churn shows why it is no retention rate. The repeat purchase rate is the one that tells you whether customers come back, and a 90-day rate is final three months after the cohort month, when a 12-month retention rate for the same customers is still nine months away.
How your repeat purchase rate compares
The September cohort above, at 12.5% within 90 days, sits close to the median of the D2C brands in our panel: 13.45% (June 2026 cohort). Turn the rate into orders, which are what pay for a customer. If every order has the same chance of a next one:
Orders per new customer within a window = 1 ÷ (1 − repeat purchase rate within that window)
At 12.5%, that is 1.14 orders per new customer within 90 days, about 14 repeat orders for every 100 first orders. Say the cohort has ordered again at 20% after 6 months and 30% after 12: 1.25 and 1.43 orders per new customer. At €15 of contribution margin 2 per order, net of VAT, a customer brings €21.43 within the year; a 12-month rate of 40% would make it 1.67 orders and €25.00. Whether that covers what the customer cost decides the CAC payback period; the math the other way round, how many repeat orders a CAC needs, is in the entry on retention.
Why repeat rates differ between reports
- Different windows. A 90-day rate, a 12-month rate and an all-time share of customers with two or more orders can be far apart for the same shop.
- Young cohorts averaged in. A report that averages all cohorts includes customers whose window hasn’t closed and reads low; the cohort entry’s section on cohort maturity works through such a case.
- Who counts as the same customer. A second order with another email address, or as a guest, can look like a new customer instead of a repeat buyer. How tools recognize returning buyers: new vs. returning customers.
- Where the second order happens. A customer who reorders on a marketplace doesn’t appear in your shop’s cohort. Our panel measures repeat purchases within the same shop only.
“Which ad brings repeat buyers, and which brings bargain hunters?” is a question we hear in customer calls. Our view: split the 90-day repeat rate by the channel, campaign and discount code of the first order, and judge acquisition on it once the window has closed. Two channels with the same CAC are not worth the same when one brings customers who come back twice as often.
How adtribute measures repeat purchases
- Cohort reports. Journeys, funnels and cohorts are custom reports in Looker Studio, one click away under Analyses.
- A cohort table in your warehouse. On request, adtribute’s attribution and analytics platform shares a cohort table to your BigQuery dataset as a view.
- Who counts as returning. A second order is a repeat only if adtribute recognizes the buyer. How it labels new, returning, loyal and superfan buyers, and how far back it looks, is in the new vs. returning customers entry.
FAQ about Repeat Purchase Rate
Is a low repeat purchase rate a problem?
Only if your CAC counts on customers coming back. A brand whose first order covers its CAC can live with few repeat buyers; one that pays more for a customer than the first order earns needs the repeat orders, and its rate decides how long it waits for its money. To see where you stand against other D2C brands, hold a finished cohort’s 90-day rate against the repeat purchase rate benchmark.
Why is repeat purchase rate measured per cohort?
Because a rate over all customers mixes buyers from last month, who had no time to return, with buyers from years ago. A cohort’s rate stays fixed once its window has closed, so September can be compared with October, while a rate over all customers moves whenever acquisition does.
Why is our returning customer rate so much higher than our repeat purchase rate?
They have different denominators. The returning customer rate is the share of a period’s customers or orders from people who ordered before, so it grows with your customer base. The repeat purchase rate follows one cohort of new customers through a fixed window.
Which channel brings customers who buy again?
Group new customers by the channel, campaign or discount code of their first order and compare their 90-day repeat rates once the window has closed. Compare channels with similar volumes of new customers, because small cohorts swing from month to month.