Did-Not-Arrive Claims: Telling Real Lost Packages From Refund Fraud
The did-not-arrive claim is the hardest return type to judge, because the two explanations look identical from your desk: either the carrier lost it or the customer is lying. Most of the time it is the carrier. Industry experience consistently shows the large majority of non-delivery claims are genuine, which means the default posture has to be belief. The job is to find the small fraudulent share without treating every claimant like a suspect, because the process you build for the exceptions is experienced by everyone.
Start with the delivery evidence
Every investigation begins with the tracking record, read carefully rather than skimmed. Look for the full event chain: acceptance, transit scans, out-for-delivery, and the final delivery scan with its timestamp. A package that shows no out-for-delivery scan but a delivery scan deserves a carrier inquiry. A delivery scan at an unusual hour, or a photo showing the wrong doorstep, is evidence for the customer, not against them.
Carrier GPS data, where available, is the single most useful evidence type. Delivery photos settle most disputes in one glance. Address issues are the other common root cause: apartment buildings with insecure lobbies, addresses missing unit numbers, and rural routes with ambiguous drop points generate a steady stream of genuine claims. Fix the address data and a chunk of the claims disappear without any fraud process at all.
Patterns that suggest fraud
Fraudulent did-not-arrive claims have a recognizable shape. The claims repeat: three or more in a few months from the same account or address. They cluster on high-value items while low-value orders from the same account arrive without issue. The customer refuses a reshipment and insists on a refund, which is backwards from genuine behavior; someone who actually wants the item usually accepts the replacement.
New accounts with no order history filing high-value non-delivery claims deserve scrutiny, as do multiple accounts shipping to the same address with rotating claim patterns. None of these prove fraud individually. Two or three together justify a deeper look: carrier investigation, signature requirements on future orders, and a documented review before the refund is issued. The standard is evidence, not suspicion.
A fair process for every claim
The process should feel the same to honest customers whether or not they are ever suspected. Step one is acknowledgment and a timeline: we are investigating with the carrier, expect an answer within a set number of days. Step two is the carrier inquiry for claims above your low-value threshold. Step three is resolution: reship or refund, with reshipment as the default for first-time claimants because it is cheaper and because fraudsters prefer refunds.
Communication does most of the work. Customers who hear nothing for a week assume the worst and escalate; customers who get a clear timeline mostly wait. Document every step in the order record. If a claim later turns out to be part of a fraud pattern, the documentation is what lets you act. If it was genuine, the documentation is what proves you treated the customer fairly.
When to deny a claim
Denials need an evidence standard, written down, applied consistently. Examples that meet it: carrier GPS showing delivery to the correct address with a matching photo, combined with a repeat-claim pattern on the account. A signed proof of delivery for a high-value item. What does not meet it: a feeling about the customer, a single prior claim, or the item's value alone. Deny on evidence, never on instinct.
Deliver denials carefully and in writing, stating the specific evidence: our carrier's records show delivery to your address at this time with this confirmation. Offer the appeal path: if the customer can provide counter-evidence, such as building security footage, the case reopens. A small share of denied customers will produce exactly that and turn out to be genuine, which is why the appeal path exists. The goal was never zero fraud; it is fraud kept small without burning honest customers along the way.