How much margin are returns quietly taking?
A returns impact calculator shows what your return rate costs in contribution margin per order. A returned order refunds its full revenue, loses the outbound shipping and return processing, and recovers only the resellable share of its COGS, so each return finishes below zero. Averaged across all orders, a 10% return rate can quietly tax 14% or more of per-order margin.
Revenue per order, before any deductions.
Cost plus freight-in plus duty: the product cost landed at your warehouse, not ex-factory.
Pick/pack plus outbound shipping.
Payment and transaction fees, as a share of order value.
Orders returned for a refund, divided by orders shipped.
The fraction of returned units that goes back into sellable stock at full value.
Inspection, restock handling and the return label, if you pay it.
A 10% return rate is taxing 14.3% of your per-order margin: $7.00 of the $49.00 every kept order earns.
The returns tax as a share of kept-order margin, at your return rate and shifted two and four points either way. Computed from your inputs, nothing assumed.
Seven inputs, straight from your stack
Every number already exists somewhere in your operation. Pull them from the same month so the blend is honest.
A month of revenue divided by a month of orders, from your store admin's sales report.
Unit cost plus freight-in plus duty, from your supplier and freight invoices. If you only know ex-factory cost, build the landed number first with the landed COGS calculator.
Pick/pack plus outbound shipping, from your 3PL or carrier invoices divided by orders shipped.
Payment and transaction fees as a share of order value, from your payment provider's statement.
Orders returned for a refund divided by orders shipped, from your store's refund report. Use a period long enough for the returns to have actually arrived back.
From your warehouse or 3PL restock records: of the units that came back, how many went back on the shelf as sellable, rather than damaged, worn or written off.
The per-return handling or inspection fee on your 3PL invoice, plus the return label if you pay for it.
A returned order does not earn zero
That is the whole trick of this number. A return refunds all of its revenue, then keeps costing you: the shipping is already spent, someone has to process the return, and part of the product never sells again. So the blend of kept and returned orders sits further below your headline margin than the return rate suggests.
Same definitions as The Math: landed COGS is cost plus freight-in plus duty, and the margin here is contribution margin before marketing. The stated assumptions sit under the inputs: full revenue including variable fees comes back on a return, shipping and processing are lost, and only the resellable share of COGS is recovered.
A kept order at $100.00 gives up $40.00 in landed COGS, $8.00 in shipping and fulfilment and $3.00 in payment and platform fees, leaving $49.00 of contribution.
A returned order refunds its full revenue, fees included, so the sale nets zero - then the costs land. Outbound shipping of $8.00 is already spent, processing the return costs $5.00, and with 80% of returns going back into sellable stock, $8.00 of COGS is never recovered. The returned order finishes at -$21.00.
Blend them: 90% of orders earn $49.00 each and 10% finish at -$21.00, so the average order really earns $42.00. The gap is $7.00 an order - 14.3% of the margin. That is the returns tax.
These are your live inputs from the calculator above, not a canned example. Change a number up there and this paragraph follows.
The averages upstream assume nobody returns
Contribution margin, break-even ROAS, CAC targets: every blended number you steer by quietly assumes each order keeps its margin. Effective margin per order is the honest version, and it changes three decisions.
A break-even ROAS computed on kept-order margin is too generous for any product line with a real return rate. Run it on effective margin instead, and a campaign that looked just profitable shows its true colour.
Free returns, exchange-first offers, a stricter resale grading: each one moves the return rate, the resellable share or the processing cost. This page turns every one of those moves into dollars per order before you commit to it.
A product that fits badly or photographs better than it wears carries a higher return rate, and this math says exactly how much margin that habit burns. Price it in before the launch budget is committed, not after.
What a blended average cannot see
Because return rates hide inside averages. Most reporting shows returns as refunded revenue, and refunded revenue understates the damage: it counts the sale you gave back, not the shipping you lost, the processing you paid or the stock you cannot resell. Costing the return in margin terms is what makes a 10% return rate show up as the 14%+ margin tax it can really be.
It is also fast. Seven numbers you already have, and the whole chain from kept order to effective margin is on one screen, with the sensitivity to your return rate directly under it.
- It is one blend. Return rates cluster by SKU, size, category and customer. The average cannot tell you which product or which buyer is doing the damage, and that is the only information you can act on.
- It assumes fee recovery. If your payment provider keeps its fees on refunds, every return costs more than this page shows.
- It treats every return as a refund. An exchange keeps the revenue and swaps the unit; its economics are different and better.
- Resellable share is not one number. It moves by category, season and condition grading, so a single input is an estimate.
When you need to act on returns rather than measure them, you need row-level truth. Blufire S7 Products / Inventory / Returns carries returns economics and refund-risk scoring with real return reasons, so the SKUs driving margin-destroying returns are named, not averaged. And S4 re-scores customer value net of returns - fake-VIP detection for the buyer who looks great on revenue and returns half of it.
When to run it, and what triggers action
Refresh the return rate and resellable share from the closed month's refund report and restock records. Watch the returns tax line, not the refunded-revenue line: refunds can look flat while the tax climbs.
Launching into a category where fit or expectation drives returns? Put its expected return rate through this page before the budget is committed, and set the launch ROAS floor on the effective margin it produces.
If the returns tax takes a bigger share of margin than last month, find which input moved - rate, resellable share or processing cost - before touching ad budgets. If effective margin moved, recompute your break-even ROAS on the new number.
The trigger is the share, not the dollars: refunded revenue can hold steady while a worsening resellable share quietly deepens every return.
Questions operators ask
Because a returned order does not earn zero, it goes negative. It hands back all of its revenue and still pays outbound shipping, return processing and the share of COGS that never resells. With the example numbers on this page, a kept order earns $49.00 and a returned one loses $21.00, so 10% of orders returning removes 14.3% of the margin, not 10%.
The outbound shipping you already paid, the cost of processing the return, and the slice of landed COGS that cannot go back on the shelf. The revenue itself washes out: it came in and went back. That is why the returned-order line on this page is built only from those three losses.
No, keep the rate to orders refunded. An exchange keeps the revenue and swaps the unit, so its cost is closer to shipping plus processing alone. Blending exchanges into the refund rate makes the tax read worse than it is; if exchanges are a big share of your returns, run the two populations separately.
Then your real returns tax is higher than this page shows. The calculator assumes a returned order refunds its full revenue including its variable fees; if the processor keeps its cut, add that kept fee to the processing cost per return and the math is honest again.
The fraction of returned units that goes back into sellable stock at full value. Units that come back damaged, worn, opened where that matters, or past season are not resellable, and their landed COGS is gone. Your warehouse or 3PL restock records are the source; grading standards move this number more than most operators expect.
Not with a blended calculator, and that is this page's honest limit. Returns cluster by SKU, size and buyer, so acting on them needs row-level data: Blufire S7 Products / Inventory / Returns carries returns economics and refund-risk scoring with real return reasons, and S4 re-scores customer value net of returns.
Which SKUs, which customers, which reasons
This page runs one blended average and cannot say who is doing the damage. Blufire S7 Products / Inventory / Returns carries returns economics and refund-risk scoring with real return reasons, and S4 re-scores customer value net of returns, so the fake VIP who returns half of what they buy stops looking like your best customer.