What does an original order earn after returns?
A refund reverses revenue; a return also changes costs and inventory. Treating the refund amount as the entire economic loss can overstate or understate the damage. Compare two complete outcomes: the customer keeps the product, or returns it under explicit assumptions.
This is a hypothetical 100-order cohort, not a real store's return rate, carrier quote or payment-provider policy. Replace the assumptions with your own records. Fixed overhead and taxes are outside the example.
Outcome one: the customer keeps the product
Revenue is $60. Goods cost $20, outbound delivery $6, packaging $1, processing $2.04 and acquisition $10. Processing uses an assumed 2.9% plus $0.30, not a universal fee.
Kept-order profit = $60 − $20 − $6 − $1 − $2.04 − $10 = $20.96.
If all 100 orders were kept, they would produce $2,096 of contribution under those assumptions.
Outcome two: a full refund and physical return
Assume a full $60 refund, a $5 return label and $2 inspection/restocking. Original processing, shipping, packaging and advertising are not refunded in this example. The returned unit restores $16 inventory value, 80% of its $20 cost.
| Component | Returned-order profit effect |
|---|---|
| Revenue after full refund | $0.00 |
| Original goods cost | −$20.00 |
| Recovered inventory value | +$16.00 |
| Original delivery | −$6.00 |
| Packaging | −$1.00 |
| Processing retained by provider | −$2.04 |
| Acquisition | −$10.00 |
| Return label | −$5.00 |
| Inspection/restocking | −$2.00 |
| Returned-order outcome | −$30.04 |
Inventory recovery is a value assumption, not cash received or another sale. Do not add the item's full retail resale price as well. A future resale has its own revenue and costs.
Weight both outcomes
Let r be the proportion of original orders returned:
Expected profit = (1 − r) × kept profit + r × returned outcome.
At 10% returns, 0.90 × $20.96 + 0.10 × (−$30.04) = $15.86 per original order. Ninety kept orders and ten returns produce $1,586 contribution and $5,400 retained revenue, about 29.37% margin on that retained revenue.
| Return rate | Expected profit / original order | Reduction versus no returns |
|---|---|---|
| 0% | $20.96 | $0.00 |
| 5% | $18.41 | $2.55 |
| 10% | $15.86 | $5.10 |
| 20% | $10.76 | $10.20 |
| 30% | $5.66 | $15.30 |
Dividing the same $1,586 by the original $6,000 orders placed gives 26.43%, not 29.37%. Name the revenue denominator whenever presenting margin.
Derive a return allowance
The difference between outcomes is $51: $20.96 − (−$30.04). At 10% returns, the allowance against an otherwise kept-order model is $5.10 per original order.
A shortcut of 10% × $60 deducts $6. Here it overstates the modeled loss by $0.90 because recovered inventory exceeds additional return handling by $9. Different recovery and handling costs can reverse that conclusion.
Use the full weighted-outcome method or the equivalent allowance, not both. When accounting revenue and goods cost already incorporate refunds and returned stock, reconcile those figures before adding a reserve for the same losses.
Model replacements and partial refunds separately
For a partial refund without physical return, reduce revenue but omit inventory recovery and return labels. For a free replacement, add replacement goods and shipping while checking whether any revenue was refunded. These events may require separate outcomes when they can overlap.
A recent cohort may not have completed its return window. Compare consistent observation periods rather than declaring the newest week unusually profitable before losses arrive. Track reasons too: damage and fit-related returns can need different operational changes.
The widget below weights the two outcomes. Profit Margin handles a simpler price/cost view, while the monthly guide explains why restored stock value is not cash. This model does not establish customer refund rights or tax treatment; it exposes the consequences of the assumptions entered.
Try different assumptions
Defaults reproduce the hypothetical example. Edit the inputs and recalculate. This is a separate teaching scenario, not a live platform quote.
Replace the example assumptions with your own costs. Check field definitions before comparing results.