Empty-Box Returns: The Refund Scam That Starts Before the Package Arrives
The empty-box return is exactly what it sounds like: a customer claims they are returning an item, ships back a box with nothing inside, and collects the refund. It is one of the oldest refund scams in ecommerce, and it keeps working because most return operations are built for speed, not verification. The refund goes out when the carrier scans the return label, hours or days before anyone opens the box. By the time the empty box is discovered, the money is gone and the evidence is thin.
How the scam actually works
The basic version is simple. The customer initiates a return, receives the label, and ships an empty or nearly empty box. Some add weight: a brick, a book, old clothes, anything to make the package feel plausible if someone lifts it. The sophisticated version uses the carrier's own tracking as cover, because a delivered return tracking number is what triggers the refund in most systems, and the tracking number says nothing about the contents.
Variations exist. The partial empty box, where a high-value item is swapped for a low-value one, works the same way and is harder to catch. The "never arrived" claim on the outbound shipment is the mirror image: the customer keeps the item and claims the box was empty on arrival. All of these exploit the same gap, which is the gap between the tracking event and the contents verification.
The warehouse checks that catch it
The single most effective control is weighing every return on arrival and comparing it to the expected weight of the item. A dress that should weigh 400 grams arriving in a 900-gram box is an instant flag. Most warehouses already have scales; the missing piece is usually the expected-weight data, which means recording item weights at pack time. That one process change catches empty boxes, partial swaps, and weight-stuffed decoys in a single step.
Second, photograph high-value returns as they are opened, with the weight visible. The photo plus the weight record is the evidence package that wins chargeback disputes and deters repeat attempts. Third, delay refunds on flagged returns until contents are verified. Auto-refund on carrier scan is the policy that makes empty-box fraud profitable; moving high-risk categories to refund-on-inspection removes the profit without slowing down legitimate returns, because legitimate boxes pass inspection in seconds.
The evidence you need to deny the claim
Denying an empty-box claim without evidence is a customer service disaster. Denying it with evidence is routine. The package you need: the arrival weight versus expected weight, the opening photos, and the return history of the account. Present it plainly: "The return arrived weighing X against an expected Y, and the contents did not match the return." Most scammers abandon the dispute at this point. For the ones who escalate to a chargeback, this same package is what wins the representment.
Track empty-box attempts by account, not just by order. A single empty box can be a packing error or a misunderstanding. Two is a pattern, and the pattern is what justifies account-level action: manual review on all future returns, refund-on-inspection permanently, or closure for repeat offenders. And share the intelligence internally: empty-box scammers often probe multiple brands in the same portfolio, so a flag in one store should follow the customer, not the order.
The uncomfortable truth is that empty-box fraud persists because the refund process was designed around trust at a scale where trust no longer works. You do not need to rebuild it around suspicion. You need exactly two things: expected weights recorded at pack time, and refunds that wait for inspection on flagged returns. Everything else is commentary. Implement those two and the scam stops being profitable, which is the only deterrent that has ever worked.