The National Retail Federation estimates that about 9 percent of returns are fraudulent, and 93 percent of retailers say they are concerned about it. Return fraud is not one thing. It is wardrobing, empty-box returns, receipt fraud, and serial abusers who return most of what they buy.
You cannot stop all of it. But you can stop treating every customer the same. A first-time buyer and someone who returns 80 percent of orders should not get identical rules. That is where per-customer logic earns its keep.
What ecommerce return fraud actually is
Return fraud is any return that games your policy for gain rather than a genuine problem with the product. Sometimes it is deliberate and organized. Often it is a customer quietly bending the rules because nothing stops them.
The line is not always clean. A shopper who buys three sizes intending to keep one is normal. A shopper who wears a dress once and returns it as new is not. Fraud sits at one end of a spectrum that starts with ordinary behavior, which is exactly why blunt rules catch the wrong people.
The main types: wardrobing, empty box, serial returns
Different schemes need different defenses. Here are the common ones.
| Type | What happens | Signal to watch |
|---|---|---|
| Wardrobing | Item used once, returned as new | Returned "unworn" but shows wear |
| Empty box | Box sent back without the item | Weight mismatch on return |
| Receipt fraud | Fake or reused proof of purchase | Order details do not match |
| Serial returns | Buys a lot, returns most | High return rate per customer |
Wardrobing hits apparel hardest. Empty-box returns are rarer but expensive when they land. Serial returns are the quiet one, because each return looks fine on its own. Only the pattern gives it away.
Why blanket policies punish good customers
The instinct after a fraud spike is to tighten the policy for everyone. Shorter windows. Restocking fees. No returns on sale items. It feels decisive, and it backfires.
Most of your customers are honest. A stricter policy taxes them to stop a small minority. Worse, it shows up before they buy, so it drags on conversion. You lose sales from good customers to block fraud from a few. The fraudsters, meanwhile, adapt fast. Blanket rules are the bluntest tool in the drawer.
Per-customer and per-country return rules
The better approach is rules that flex by who is returning and where. A trusted, long-term customer can get instant refunds and easy exchanges. A brand-new account returning a high-value item can get inspection-first refunds until they build a track record.
Per-country logic matters here too. Fraud patterns and shipping costs differ by market, so the rule that fits one country may be wrong for another. The goal is not one strict wall. It is a set of rules that match the actual risk in front of you, so honest buyers stay happy and abusers hit friction.
Balancing fraud control against experience
Every fraud control adds friction, and friction costs sales. The question is always where to spend it. Spend it on the risky returns, not the routine ones.
A simple frame helps. Low-value items from trusted customers should flow through with almost no checks, because the cost of inspecting them beats the cost of the occasional loss. High-value items from unknown accounts justify a closer look. When you match the check to the risk, you catch more fraud with less customer pain. Treating every return as suspicious does the opposite.
Fraud signals worth tracking
You do not need a fraud team to start. You need to watch a handful of signals and act on the clear ones. Here are the ones that pay off.
| Signal | Why it matters |
|---|---|
| Return rate per customer | Flags serial returners early |
| "As new" returns showing wear | Points at wardrobing |
| Weight mismatch on arrival | Catches empty-box returns |
| Refund requested before item ships back | Common abuse pattern |
| High-value item, new account | Higher baseline risk |
None of these is proof on its own. Together they build a picture. The point is to move from gut feel to something you can measure and set rules against.
How Pango fits
Pango's live returns, exchanges, and claims module supports per-country label and refund logic, so the rules that apply to a return already flex by market. Exchanges go to any product in your catalog, which channels genuine returners toward keeping revenue instead of walking with a refund.
Analytics are live too, so return rate by customer and return reason by SKU are visible rather than buried. Branded tracking and proactive notifications round out the standing modules, normalizing the many statuses carriers return so return transit is never a black box.
Per-customer scoring rules tuned to your specific fraud patterns are build-to-fit. Pango reads your data and policies and builds that logic to match your risk, rather than offering a fixed fraud toggle. See how it connects on the Pango return management platform.
To put the rules in writing, read our guide on how to write a return policy. To understand what each abused return actually costs, see the true cost of a return. And for cases where getting the item back costs more than it is worth, look at returnless refunds.
