Written by DROPS.ST.
Repeat purchase rate measures how many customers bought more than once under a defined set of rules. Count distinct customers, define a completed purchase, and state the observation window before doing the division. Otherwise a healthy-looking percentage can hide a shrinking business or a changing customer mix.
DROPS customer accounts and order history give shop staff the records needed to investigate repeat business. The commercial value comes from asking the right question of those records. A returning customer, a repeat order and a customer who returns within 60 days are three different measures.
Choose the question before the formula
For a period-based repeat purchase rate, divide customers with at least two qualifying orders during the period by customers with at least one qualifying order during that period. This tells you about buying frequency inside that window.
For a first-purchase cohort rate, start with customers whose first qualifying purchase was in a particular period. Measure how many make another qualifying purchase within an equal number of days after their first order. This answers how effectively you turn new customers into repeat customers.
A third metric, returning-customer share, counts people buying this period who also bought before it. They may place only one order this period. Give that metric its own name rather than mixing it into either formula.
Work through the same shop three ways
These figures are illustrative. A shop records 150 qualifying orders from 100 distinct customers in June. Within June, 30 customers place at least two orders. Fifty customers have purchase history before June, including some who purchase only once in June.
| Measure | Calculation | Result |
|---|---|---|
| June customers buying at least twice in June | 30 ÷ 100 | 30% |
| June customers who had purchased before June | 50 ÷ 100 | 50% |
| Orders per June customer | 150 ÷ 100 | 1.5 |
None of these numbers is the share of first-time June customers who return within 60 days. That requires a separate cohort and a later review date.
Suppose 50 customers made their first qualifying purchase in June. Once all 50 have had a full 60 days, 14 have purchased again. Their 60-day repeat purchase rate is 14 ÷ 50 = 28%. It is entirely possible for all four results to be correct at the same time.
Keep the counting rules on one worksheet
Record the following beside the result so next month's comparison means something.
- Customer identity: use a stable customer record where available; document how guest purchases and duplicate accounts are treated.
- Qualifying purchase: state the paid/completed status required and how cancellations, full refunds and test orders are excluded.
- Date basis: choose the relevant order date or completion date and one timezone.
- Window: specify the calendar period or days after first purchase.
- Cohort maturity: exclude customers who have not yet had the full observation time from a final cohort result.
- Channel coverage: identify which shops or sales channels the records cover.
Do not merge two customers because their names look alike. Do not quietly treat missing marketplace history as proof that someone is new to the entire business. The result can accurately describe your own shop without describing every sale you have ever made.
Read the rate beside the business result
A rise from 25% to 30% is five percentage points, or a 20% relative increase. It is not evidence of 20% more revenue. Customer count, order value, discount cost and margin may all have moved differently.
Compare counts alongside rates. Twenty repeat customers from 50 customers gives 40%; 60 from 200 gives 30%. The first group has a higher rate while the second contains three times as many repeat customers. Your next decision depends on whether you are trying to grow the base, improve retention, or protect contribution.
The weekly order review helps place retention beside everyday trading. When evaluating a campaign, add the distinctions in attribution and incremental sales: a campaign-linked purchase is not automatically an extra purchase caused by that campaign.
Run repeat business on numbers you can explain. See DROPS and inspect how customer accounts and order history support your chosen measurement. Bring your definition to the demo, then follow one customer through the records.