Reporting worksheet · Synthetic example
Read repeat purchase with enough time for a second order
A synthetic cohort example shows why a repeat purchase comparison needs the same customer definition and time to reorder.
Define the customer group and the event
For this worksheet, a cohort is a group of distinct customers whose first qualifying order occurred in a stated period. A repeat purchaser is a customer with at least one additional qualifying order within the follow-up window. Define cancelled, test and refunded orders consistently before counting.
Count customers and orders separately. A customer placing three orders is one repeat purchaser in this measure. This is a customer repeat-purchase rate, not the percentage of this month's orders placed by returning customers.
A synthetic comparison at unequal ages
All figures below are invented for explanation. They are not RavenOS output, client results or a retail benchmark. Assume both groups contain 100 first-time customers and use the same qualifying-order definition.
| Customer group | Follow-up available | Repeat purchasers | Observed share |
|---|---|---|---|
| Older group | 90 days per customer | 30 of 100 | 30% |
| Recent group | 30 days per customer | 12 of 100 | 12% |
What the table cannot tell you yet
The recent group's 12% does not establish worse retention than the older group's 30%. One has had three times as long to place another order. Purchase cycles, product mix, acquisition source and promotions could matter too.
Suppose the older group had 10 repeat purchasers by day 30. The equal-age comparison is then 10% versus 12% at 30 days. That narrower observation still does not prove a causal improvement. The recent group's 90-day result remains unknown.
Set the clock for each customer
Use days since each customer's first qualifying order, not simply the calendar date when the report is opened. A monthly cohort contains customers acquired on different days. Include only customers who have reached the chosen horizon when calculating a complete 30-day or 90-day measure, or clearly mark the result as immature.
Keep the cohort size, eligible customer count, cutoff date and unresolved records next to the rate. Preserve a prior snapshot when later orders or refunds change it.
Copy the review record
Record cohort entry rule: ____. Qualifying order rule: ____. Follow-up horizon: ____. Data cutoff and timezone: ____. Eligible customers: ____. Customers with an additional qualifying order: ____. Rate and known exclusions: ____.
Add product mix, acquisition changes and the expected purchase cycle as investigation questions. Do not fill those gaps with a generic category benchmark. No universal reorder interval is assumed here.
Turn the comparison into a useful brief
A responsible next step may be to wait for cohort maturity, fix the definitions or investigate a difference that remains at equal ages. A second order is not itself evidence of customer profit; margins, returns and costs need their own review.
RavenOS supports Joshua's operator-led work. This worksheet does not establish a connected retention integration or an automatic campaign. Bring the question and the available reports to a fit conversation before sharing confidential records.