Return Fraud Cost Calculator

False positives cost more than the fraud.

Written and maintained by Mohit PatelLast checked August 4, 2026How we build these

Currency

Returns per month

Share that are fraudulent

Average order value

Contribution margin

Share you currently detect

Detection with better controls

Monthly cost of the controls

Genuine returns wrongly blocked

Monthly fraud loss

$1,228

30 fraudulent returns

Loss with current detection$1,228

Loss with better detection$614

Cost of false positives−$772

Net gain from better controls−$1,058

False positives cost more than the fraud they prevent when detection is set too aggressively, a wrongly refused genuine return loses the customer, not just the order. Costing them at three times the order contribution is a conservative way to keep that trade-off visible.

How the Return Fraud Cost Calculator works

Detection set too aggressively costs more than the fraud it prevents, because a wrongly refused genuine return loses the customer rather than the order. Pricing false positives at several times the order contribution keeps that trade-off visible.

Also known as: wardrobing cost calculator · returns abuse cost · fraudulent return losses

The forms it takes

Return fraud is not one behaviour. Wardrobing is buying to use and returning, which dominates in occasionwear and electronics around events. Receipt fraud is returning goods obtained elsewhere or stolen. Switch fraud is returning a different, worse item in the box. Empty box returns are exactly what they sound like.

Then the grey area, which is larger than the criminal one: bracketing, where a customer orders several sizes or colours intending to keep one. It is not fraud, it is entirely permitted by most policies, and it produces the same handling costs. Lumping it in with fraud confuses the diagnosis and leads to policies that punish ordinary customers.

Industry estimates put return fraud and abuse at somewhere around 10% to 15% of total returns by value, with the range wide because definitions differ. What is not disputed is that it concentrates: a small number of accounts generate a large share of it.

Costing an incident properly

A fraudulent return costs more than a legitimate one because there is usually no recoverable inventory at the end of it. The refund goes out, the goods either do not come back or come back worthless, and you have paid the handling and the shipping regardless.

So the per-incident cost is close to the full retail value plus the operational cost, rather than the handling cost alone. On a £150 item that is perhaps £160 against a legitimate return of the same item costing £14. The multiple is roughly ten to one.

Multiply by frequency and it becomes a real line. A shop doing 20,000 orders a year at a 25% return rate handles 5,000 returns; if 3% are fraudulent at an average £120, that is £18,000, which for most small businesses is a material number that has never appeared anywhere in the accounts.

Detection that does not punish everyone else

Most fraud is visible in the pattern rather than the incident. Return rates far above the catalogue average, returns clustered just inside the policy window, high-value items returned repeatedly, mismatches between shipping and billing addresses, and multiple accounts sharing a delivery address.

Serial numbers are the single most effective control for electronics and anything with a unique identifier. Recording what shipped and checking what returned defeats switch fraud outright, and it costs a scan at dispatch.

Photography at packing does the same job for the empty box claim. A photograph of the packed carton on the scales, with the weight recorded, settles most disputes about what was in it. Both of these are cheap, both require discipline, and both are worth far more than any policy tightening applied across the whole customer base.

The cost of getting it wrong

False positives are expensive in a way that does not appear in a fraud report. A good customer accused of fraud does not come back, tells people, and may go to their card issuer, which turns a returns question into a chargeback. The lifetime value lost usually exceeds the fraud prevented.

Which argues for graduated responses rather than binary ones. A flagged account might lose free returns before it loses the ability to buy. It might move to inspection-before-refund rather than refund-on-receipt. It might be quietly excluded from promotions. Each of these reduces exposure without an accusation.

Account closure is the last step and should be rare, documented, and decided by a person. Automated closure on a pattern match will eventually close a legitimate account, and the first time it happens publicly it costs more than the policy saved.

Policy choices that reduce exposure

Shorter return windows reduce wardrobing directly, because the behaviour depends on using the item and returning it within the window. Cutting from 90 days to 30 removes a great deal of it at a modest conversion cost.

Tags that must remain attached work well in apparel and are cheap. A large, visible tag that cannot be reattached makes wear-and-return impractical, and honest customers barely notice because they try things on rather than wearing them out.

Inspection before refund is the strongest control and the one customers dislike most. Refunding on receipt is faster and better for everyone honest; refunding on inspection catches switch and empty-box fraud before the money leaves. Applying inspection selectively, above a value threshold or on flagged accounts, gets most of the protection without slowing down the majority.

Where to go next

The Return Fraud Cost question rarely arrives on its own. These are the ones that usually come with it:

Chargeback Cost Calculator — Even winning costs the fee and the time.
Return Cost Calculator — A multiple of the margin on a kept order.
Return Rate Calculator — By order and by item, which differ.
Etsy Fee Calculator — Every Etsy fee on one sale, itemised.

Not financial advice. Marketplace fees change, and they vary by country, plan and seller status. Every rate here is an editable default, not a quoted price, check the platform's current fee schedule before you price a product against it. This is not tax or business advice.

Frequently asked questions

What counts as return fraud?

Wardrobing, using and returning; returning a different or empty item; claiming non-delivery on a delivered parcel; and serial returning that exceeds any plausible genuine rate.

How common is it?

Estimates vary widely and self-reported figures from vendors selling detection tools should be treated carefully. What matters is your own data on serial returners.

How do I detect it without false positives?

Look at patterns across a customer's history rather than at single returns. One suspicious return proves nothing; a consistent pattern across many orders does.

What do I do about it?

Quietly restrict rather than confront: remove free returns, require original packaging, or decline future orders. Public accusations of fraud go badly wrong when the pattern turns out to be innocent.