Skip to content
Faisal Mahmud TusherDigital marketing, SEO & analytics
with Faisal Tusher

Metric

Repeat Purchase Rate

Share of customers who placed more than one order in the period.
BusinessmetricUpdated

Formula

Customers with two or more orders ÷ Customers with at least one order × 100

Unit: % · higher is better

Definition

Take the customers who placed at least one order in a window, count how many of them placed a second or later order within the same window, and divide. Every customer in the numerator also sits in the denominator, so the rate cannot exceed the whole.

Customers with two or more orders ÷ Customers with at least one order × 100(%)

  • Customers with at least one order are the base: distinct customers, not orders, identified by account, email or another stable key. Guest checkouts that cannot be matched to earlier orders will look like new customers and depress the rate.
  • Customers with two or more orders are members of that base who ordered again within the window.
  • The window is the time allowed for the repeat order to happen.

The window rule

A calendar-year rate counts a customer who first ordered in January as having almost twelve months to reorder and a customer who first ordered in December as having weeks. The rate then rises or falls with the timing of first orders, not with loyalty. The fix is a fixed window per customer: measure each customer from their first order for the same length of time, such as a set number of days, and report only customers whose window has closed. Choose the window to cover the product's natural repurchase cycle; a consumable bought monthly and a piece of furniture bought once a decade need very different windows, and some products barely repeat at all.

How to interpret it

Higher is better, read in three ways.

  • By first-order cohort. Group customers by the month of their first order and measure each group over the same fixed window. Comparing successive Cohort groups shows whether newer customers come back more or less than older ones, which a blended figure cannot.
  • By product category. Categories with short replacement cycles repeat more than durable ones, so a shop-wide figure moves with the category mix. Read each category against its own history.
  • By acquisition source. Customers won through heavy discounts or broad prospecting often repeat less than those who came through search or referral; the first order's channel predicts the second.

The repeat purchase and loyalty guide sets out the programmes that move these readings, and the ecommerce retention section places the rate among the other measures of a returning customer base.

Illustrative example — not a real result

A dog treat shop fixes its window

Pinecroft Dog Treats reports a calendar-year repeat purchase rate of 28% for 2025, down from 31% the year before, and assumes its post-purchase emails stopped working. But 2025 had a large December gift campaign: 2,400 of its 9,000 first-time customers first ordered in December, with only weeks left in the year. The shop switches to a fixed 90-day window per first-order cohort and reports only cohorts whose window has closed. Customers who first ordered before the post-purchase email launched in March reordered within 90 days at 22%; cohorts from April to September, who received it, reordered at 29%. The emails had been working all along.

What moves it

  • The product's natural cycle. Consumables repeat on their own schedule; marketing can bring the second order forward but cannot invent demand that does not exist.
  • The first-order experience. Delivery speed, packaging, product quality and how problems are handled decide whether a first-time customer considers a second order.
  • Post-purchase messages. Timely reminders, replenishment prompts and useful follow-up content bring customers back; the retention email guide covers how such sequences are planned.
  • Acquisition mix. Customers bought with deep first-order discounts tend to repeat less.
  • Loyalty and subscription options. Rewards and subscribe-and-save offers convert occasional buyers into regular ones, though subscription customers are then better measured on churn.

Purchase Frequency counts how many orders the average customer places in a period; the repeat rate counts how many customers placed more than one. They answer different questions, and a few very loyal customers can lift frequency while the repeat rate stays flat. LTV rests on both: repeat orders are where most of a non-subscription customer's value lies. Churn Rate carries the same information from the other side for subscription businesses, and transactional businesses often prefer the repeat rate because it avoids having to define when a silent customer has gone. The retention objective explains how the rate becomes a target, and the AARRR framework applied to sales uses it as the retention stage of the funnel. For the economics of retention in online retail more broadly, see the ecommerce industry guide.

What good looks like

There is no reliable universal benchmark, and the site does not cite one. The rate depends on the product's repurchase cycle, the price, the window chosen and the mix of customers acquired, and published figures seldom state their window. Judge it against your own cohorts at the same window length, by category, and against the economics: whether enough first-time customers return, often enough, for their lifetime margin to repay what it cost to win them.

Common mistakes

  • Using calendar periods instead of a fixed window per customer, so the rate tracks when first orders happened.
  • Choosing a window shorter than the product's natural cycle, which makes retention look poor.
  • Counting orders instead of customers in the numerator or base.
  • Including customers whose window has not closed, which drags recent cohorts down.
  • Failing to match guest orders to earlier customers, so repeat buyers look new.
  • Reading the blended rate only, so a change in the category or channel mix looks like a change in loyalty.

More terms to look up

Where this term is used

Related