Generous return policies have earned their place as a reliable growth lever. They reduce purchase hesitation, lift conversion rates and build loyalty that compounds over a customer’s lifetime. Whether the sale happens at a store counter, on a website or through a third-party marketplace, finance teams understand those outcomes well enough to model them directly into revenue projections. Flexible returns are a deliberate commercial strategy, and they work.
What is less visible is how systematically repeat abusers and organized fraud rings exploit the same policies, and how little of that cost reaches a margin forecast. Return abuse and refund fraud rarely get their own line in financial reporting. Operations, customer experience and fraud teams absorb the cost, which is why financial planning and analysis never really see the true number. Returns fraud now accounts for nearly 14% of all U.S. returns and costs retailers roughly $103 billion in 2024, a figure that’s climbed sharply since 2019. The result is a margin leak that scales with revenue itself, across every channel a retailer sells through.
What Fraud Teams Call “Abuse,” Finance Should Call “Margin Leakage”
Wardrobing and bracketing sound like fraud-team vocabulary, but they describe ordinary shopping habits pushed to their limits. Wardrobing — buying an item, wearing it once then returning it as unused — is the single most common form of return fraud retailers report, and it happens as readily with a dress carried out of a store as one shipped to a doorstep. Bracketing, ordering the same item in three sizes with no intention of keeping more than one, shifts the cost of a fitting room onto the retailer’s shipping and processing budget. Neither violates any rule in the returns policy.
The channel changes the tactic, but the economics stay the same. In stores, abuse shows up as receipt fraud, returns of shoplifted goods for gift cards and employee collusion. On marketplaces, it surfaces as empty-box returns, decoy returns of counterfeit or substituted items and exploited returnless-refund policies, where the buyer keeps the product and the refund both. Each drains margins the same way.
These behaviors aren’t rare edge cases. In SEON’s 2026 Fraud & AML Leaders Report, a survey of 1,010 fraud, risk and compliance leaders, return fraud and promo or discount abuse each ranked as a top threat for 18% of respondents, trailing only account takeover at 26%. Two of the three leading fraud categories in the industry are customers using the rules against the business, not criminals breaking in from outside.
That distinction, rule-following instead of rule-breaking, is exactly why the losses get buried instead of flagged. A fraud team built to catch bad actors circumventing controls has no clear owner for a customer who follows every posted policy. Finance priced shipping subsidies, restocking and return windows around an assumed abuse rate, and nobody revisits that assumption once real behavior overtakes it. Neither side can fix it alone.
Where the CFO’s and Fraud Team’s Numbers Diverge
Fraud teams track fraud rate and chargeback ratio. Finance tracks gross margin, contribution margin per order and inventory shrinkage. Neither metric captures a customer who wears an item once, returns it and then reorders three sizes up on the next order, or one who walks a used item back to the counter for full credit. Fraud calls it abuse. Finance calls it margin, and the gap between the two lets it compound quietly. A 1- to 2-point margin bleed rarely triggers a fraud review, but it shows up every quarter as a stubborn drag on gross margin that nobody can label.
Online, the Merchant Risk Council’s 2025 Global eCommerce Payments and Fraud Report found that refund and policy abuse remains the most prevalent fraud type merchants face. It’s a post-purchase loss that erodes margin one return at a time rather than showing up as a single flagged incident. If it’s the industry’s most common loss retailers report, it’s already inside your margin. You just haven’t measured it yet.
Static return policies can’t make that distinction; they either accept a return or they don’t. Real-time signals can: device and identity intelligence, email and phone reputation, loyalty and tender history and behavior patterns drawn from a customer’s order history record. Together they separate a legitimate serial returner from a bracketing or wardrobing pattern before the refund clears, at the counter, in the app or across a marketplace, while the merchandise and the margin can still be saved.
The Manual Review Gap
Gartner found that more than half of retailers still lack an automated way to catch returns policy abuse, relying instead on manual review of refund and return claims. That gap matters because manual review happens after the transaction, not before.
For a CFO, that timing is the real problem. A refund issued before anyone verifies the claim is cash out the door on an unconfirmed loss, legitimate or not. That’s a cash flow problem before it’s ever a shrinking number.
There’s a second wrinkle, and it widens every channel a retailer adds: no single function owns returns fraud end-to-end. Fraud sees the transaction. Store operations see the counter return. A marketplace platform and its third-party sellers each see a fragment. Finance sees the number weeks later, with no way to trace it back. Nobody assigns ownership, so nobody fixes the loss.
One Metric, Not Two Dashboards
Segmenting the return policy by risk signal, rather than applying identical terms to everyone, prevents bracketing and wardrobing from scaling. It works best when a retailer checks identity and device signals the moment the return request comes in, at the counter, in the app or through the marketplace, while there’s still time to stop a bad refund.
Finance and fraud can watch the same number instead: what each return reason actually costs in margin. Neither team has to guess whose budget the loss belongs to. Return fraud already sits inside every gross margin forecast, whether or not anyone modeled it there. Every unreviewed refund is a margin decision finance never got to make.
