The returns fraud glossary

Bracketing

Buying multiple sizes, colors, or versions of the same item with the intent to keep one and send the rest back. Common in apparel, where fit uncertainty pushes shoppers to turn their home into the fitting room.

Wardrobing

Wearing an item and then returning it as if it were new. Tags stay on, the item goes out once, and it comes back with wear that is hard to prove and costly to resell.

Serial returner

A customer whose return rate sits far above the norm over a long window. Most returns are honest; serial returners are the small group driving a large share of return cost.

Return abuse

Behavior that uses a returns policy in ways the merchant never intended, from wardrobing and bracketing at scale to returning used, swapped, or damaged items.

Friendly fraud

Disputes and claims from customers who did receive and keep their goods, such as chargebacks on legitimate orders or false claims that an item never arrived.

Return fraud

Deliberate deception in the returns process for financial gain, including returning stolen, used, or counterfeit items, and claiming refunds for items that were never sent back.

Refund fraud

Fraud aimed at the refund itself rather than the item, such as exploiting refund timing, tampering with return labels, or manipulating tracking to trigger an automatic refund.

Price switching

Returning a lower priced item in place of the higher priced one that was ordered, often by swapping tags, boxes, or labels before the return ships.

Returnless refund

A refund issued without requiring the item back, used when the cost of processing the return exceeds the item's value. Abusers probe for brands that grant them often.

Keep rate

The share of sold units a customer keeps rather than returns. Tracking keep rate per customer is the cleanest way to spot serial returners.

Return rate

The share of orders or units that come back. Apparel carries some of the highest return rates in ecommerce, which makes monitoring per customer behavior essential.

Return window

The number of days a customer has to send an item back. Long windows help honest shoppers and also give abusers more room to wear, use, and return.

Return scoring

Assigning each return request a risk score from signals like customer history, order pattern, timing, and item type, so CX teams can act on risk instead of gut feel.

Approve, warn, deny

The three decisions a modern returns flow supports. Approve honest returns fast, warn borderline customers with a policy notice, and deny returns that clearly break policy.

RMA

Return merchandise authorization, the approval step and reference that gates a return. RMA data is where return scoring and decisions get enforced.

Want these patterns scored in your own returns data?