Define the return cohort and records
Records to reconcile
- Choose the order cohort, observation period and point at which the return window is complete. (not complete)
- Reconcile original orders, return events and completed customer refunds. (not complete)
- Record reverse delivery, inspection, handling and disposal costs per event. (not complete)
- Record product recovery only when the item is actually returned to usable stock or otherwise recovered. (not complete)
- Record fee credits only when supported by the current contract and settlement statement. (not complete)
- Keep exchanges, cancellations and failed deliveries separate unless the review defines their treatment. (not complete)
Reconcile each return event in sequence
- Measure return frequency from the defined original-order cohort.
- Record the refund and any customer shipping amount returned.
- Add reverse-logistics, inspection, handling and disposal costs.
- Subtract only realised product recovery and contract-supported fee credits.
- Calculate loss per return event before applying the return-frequency assumption.
- Reconcile expected loss per original order back into the complete order model.
- Review timing separately when refunds and recovery create a cash gap.
Expected return loss per original order = return frequency × (refund value + reverse-logistics cost - product recovery - fee credit)
- return frequency
- Return events divided by original orders in the matched cohort (returns per original order) — reconciled cohort record or user scenario
- refund value
- Customer refund attributed to one return event (currency units per return) — refund record
- reverse-logistics cost
- Return shipping, inspection, handling and disposal inside the boundary (currency units per return) — business record or user scenario
- product recovery
- Realised value recovered from the returned product (currency units per return) — inventory or recovery record
- fee credit
- Applicable fee amount actually credited after the return (currency units per return) — current contract and settlement record
| Component | Effect on loss | Evidence boundary |
|---|---|---|
| Refund value | Increases loss | Completed refund record |
| Reverse logistics | Increases loss | Invoice, handling record or explicit scenario |
| Product recovery | Reduces loss | Realised usable-stock or disposal recovery |
| Fee credit | Reduces loss | Contract and settlement evidence |
| Timing gap | Changes cash timing, not event loss | Dated refund and recovery records |
Work a bounded return scenario
Separate frequency from loss per event
These neutral figures are user assumptions in currency units and are not a market return benchmark.
| Line | Value |
|---|---|
| Refund value | 80 |
| Reverse logistics | 12 |
| Product recovery | (45) |
| Fee credit | (3) |
| Loss per return | 44 |
| Return-frequency scenario | 5% |
| Expected loss per original order | 2.20 |
Check common mistakes and next actions
- Applying a return percentage to revenue without modelling loss per event.
- Treating the original product cost as fully recovered before the item is usable or sold.
- Assuming every transaction fee reverses without checking the current contract.
- Counting recovery as a credit here and again as negative product cost elsewhere.
- Mixing open return windows with completed cohorts and calling the result comparable.
Methodology sources
- Returns Profit Impact methodology — Margin101: Product-owned return-event formula, cohort boundary and exclusions.
Treat return-rate tolerance as a scenario boundary, not a benchmark
A return-rate boundary depends on the contribution available before a return and the complete incremental loss created by one return event. Products with different starting contribution can tolerate different entered return rates even when reverse logistics, fee loss and recovery assumptions are identical.
Scenario return-rate boundary = pre-return contribution per order ÷ positive loss per return event
- Pre-return contribution per order
- Order revenue less the declared product, fulfilment, payment and other pre-return variable-cost boundary (CU/order) — completed cohort records or labelled scenario
- Loss per return event
- Refund-related contribution loss, reverse logistics, unreversed fees and value lost after evidenced recovery (CU/return) — completed return-event records or labelled scenario
- Scenario return-rate boundary
- Arithmetic point where expected event loss equals the stated pre-return contribution (percentage of aligned orders) — calculated scenario
The denominator must be positive. Use one completed cohort, currency, tax basis and return definition; the result is not a normal or acceptable rate.
| Product scenario | Pre-return contribution/order | Boundary calculation | If event loss rises to 72 CU |
|---|---|---|---|
| Higher-contribution product | 36 CU | 36 ÷ 60 = 60% | 36 ÷ 72 = 50% |
| Lower-contribution product | 12 CU | 12 ÷ 60 = 20% | 12 ÷ 72 = 16.67% |
Run the editable event-loss scenario
- Returns Profit Impact Planner
Test period contribution after returns and recovery assumptions
- Ecommerce Order Profitability Planner
Test order contribution after the full fee stack