Measure the margin impact of returns and refunds

Direct answer

Compare the contribution expected from the original order with its contribution after the return. Reverse the refunded sale, include unrecovered shipping, fees, and handling, and account for the item’s recovered inventory value. Report the difference by SKU and channel; a refund alone is not the full margin impact.

Why this question comes up

The outcome differs when an item is resold, discounted, or written off. Refunds can also happen without a physical return, so sales reversals and returned units need separate counts.

Records to gather

• Returns by SKU, channel, and reason code
• Original sale, discounts, and refund amount
• Outbound and return shipping costs
• Payment and marketplace fees, including any fee credits
• Inspection and repackaging labor, if tracked
• Original item cost and disposition: restocked, discounted, refurbished, or written off

Review workflow

1. Link each return to its order. Separate refunds, cancellations, exchanges, and physical returns so one event is not counted twice. State whether rates use orders or units as the denominator.
2. Reconcile the sale. Start with original net revenue and item cost. Apply the actual refund and any shipping refund once; do not subtract the same reversal again from a net-sales figure that already includes it.
3. Add incremental costs. Include return postage, handling, and nonrefunded fees. Outbound shipping and original fees belong in the original order contribution and should not be added a second time.
4. Record what happened to the item. For a restocked unit, use the recovered inventory value; for damaged or written-off stock, use the value supported by the inventory records. Do not count the original item cost as both lost and recovered.
5. Compare outcomes. Show original versus post-return contribution by SKU, channel, and reason. Look for changes after a supplier, listing, or packaging update, but investigate before assigning cause.

What a useful answer should include

• Original and post-return contribution with each component shown once
• Refunds versus physical returns, and the rate denominator
• Item disposition and recovered inventory value
• SKU, channel, and reason breakdowns where records allow
• Missing cost or inventory data that limits the estimate

Common failure modes

• Double-counting. A net-sales report may already reflect the refund, and a restocked unit may already reverse cost of goods sold.
• Treating every refund as a returned unit. Some refunds are partial or issued without receiving stock.
• Assuming all costs are in the ledger. Return labels, handling time, and disposition may need order-platform or warehouse records.