Quick answer
For a clearly defined reporting window, repeat purchase rate can be calculated as distinct customers with at least two qualifying orders divided by distinct customers with at least one qualifying order, multiplied by 100. Count orders first. Three product rows from one basket are not three separate visits.
Learn the definition before the formula
The independent Dig Data video uses a simulated business case to explore repeat purchase rate and Google Sheets. Its public description emphasizes defining the measure before working with tools. We verified the title, description, creator and original February 25, 2026 upload metadata; this companion is not a transcript or a claim to have independently reproduced the creator's workbook. Our practical workflow also uses Google's pivot-table documentation.
Make a clean order list
Choose a period, such as one complete calendar month, and write down what qualifies: paid orders, how cancelled orders are removed, and how refunds are treated. Keep a stable customer ID, order ID and transaction date. Work from a protected copy of the export; do not upload customer contact details to a public teaching workbook.
If the export contains product lines, reduce it to one row per qualifying order before counting purchases by customer. A customer who buys a sandwich and drink in one order should contribute one order. Missing customer identifiers should remain an explicit unknown group, rather than being merged into a fictional customer who appears to purchase hundreds of times.
Calculate and interpret
Use a pivot table on the clean order list to count orders for each customer. Google documents choosing row fields, summarized values and filters in Sheets. Count how many identified customers have two or more orders, then divide by all identified purchasing customers in that same filtered population.
For a hypothetical café with 100 identifiable purchasing customers, of whom 24 place at least two qualifying orders, the rate is 24%. This is illustrative arithmetic, not a recommended target. Keep the unidentified-order share beside the result: better identification next month can change the metric even if buying behaviour has not improved.
A monthly repeat rate also differs from a new-customer cohort's second-purchase rate within 30 days. The latter follows people from their first purchase and needs enough observation time. Decide which question you want answered before comparing periods.
Use the loyalty ROI guide to connect repeat behaviour with reward cost, and the dashboard lesson to inspect available loyalty measures. You can explore Boomerangme to see whether its records and integrations support your reporting needs; a recorded loyalty action is not automatically a complete sales ledger.
Quick questions
Do two items on one receipt make someone a repeat customer?
No. Under this definition, the person needs two distinct qualifying orders. Deduplicate order lines before counting.
Should anonymous cash purchases be assigned to one customer?
No. Report them as unidentified activity. Combining them would produce a misleading repeat-customer count.
Does a higher rate prove the loyalty program caused more visits?
No. Seasonality, identification changes and customer mix can move the measure. Compare consistent definitions and use a suitable campaign evaluation before making a causal claim.
Sources & further reading
- Think Like a Data Analyst: Repeat Purchase Rate Case Study (Google Sheets)Independent educational case study; verified metadata, no transcript claim.
- Create and use pivot tablesPrimary Google Sheets instructions for rows, summarized values and filtering.
Sources reviewed .
Numerical examples are illustrative unless a cited source states otherwise.