Are new customers becoming repeat customers?
01 / THE APPROACH
Start with a defined repeat customer rate
Choose a consistent set of successful order statuses and a clear observation period. Count identified customers with at least two qualifying orders, then divide by all identified customers in that population. Report the period alongside the percentage. Exclude order-scoped anonymous identities from repeat-customer calculations because they cannot reliably connect purchases.
02 / THE APPROACH
Compare cohorts at the same age
Group customers by their first purchase month. A cohort acquired last month has had less opportunity to return than one acquired last year. Compare 30-day retention only after the relevant customers have had 30 days of follow-up; use the same principle for 60-, 90-, and 180-day windows. This avoids confusing incomplete history with poor retention.
03 / THE APPROACH
Investigate the first purchase
Products in a first order can be associated with different repeat rates. Associate a customer with every product in that first order, rather than assigning the whole relationship to one item. Check sample sizes and acquisition timing before interpreting the differences. The association does not prove that the product caused the customer to return.
Before you act on the analysis
- Define qualifying statuses and how refunds affect value.
- Use a consistent identity across guest and registered orders.
- Report eligible cohort size and follow-up duration.
- Test a retention intervention rather than assuming the pattern explains its cause.
WHAT WE’RE BUILDING
From the question to a useful next step.
Retail Science Co is developing cohort retention, reorder intervals, and first-purchase product analysis for WooCommerce. Early access is planned to present supporting metrics and ranked observations, using pseudonymous customer keys.
Join the early-access waitlist