Hi,
I’ve processed a product return request from a customer. I generated a USPS return label for him to return the product from Florida back to California.
He dropped it off on May 9th, according to the USPS tracking. But the shipment took like one month to go from Florida to California. Then the package was in & out of the local USPS facilities for like a week but never actually delivered to me.
I went to a USPS office and asked them what’s going on. They told me that the shipment isn’t destined for my address. It’s being routed to Kentucky.
I showed them the USPS PDF label. They said even though the USPS tracking number on my USPS PDF label matches the physical label on the package, but the destination address on the package doesn’t match. They got a picture of the physical label but couldn’t show it to me due to privacy regulations. They said the system always respects the destination address written on the label.
I suspect the buyer wrote down a different destination address to cover my destination address on the package. So he can claim to have returned my package and get a full refund, but actually keep the package.
Has anyone been in similar situations?
Hi @sss1128 ,
This kind of return fraud where a customer manipulates the label or tracking to appear as though they’ve returned a product, has come up more often lately. USPS’s privacy policy makes it even harder to verify what actually happened without the package being delivered.
A few thoughts that might help in the future:
-
Use a returns portal with stricter control
Instead of manually generating return labels, you might consider using a returns app like ParcelPanel Returns & Exchanges, which issues pre-approved labels and tracks return reasons per item. That way, every return request goes through a controlled flow and you have records to refer back to. You can also set manual approval for certain products or high-risk orders.
-
Enable return fraud detection workflows
While not foolproof, apps like these let you flag or block returns based on order history, location mismatches, or refund abuse patterns. You could also require photos of the item before return is approved, something that deters fraudsters and provides an audit trail.
-
Always double-confirm return tracking info
It’s a good idea to log the generated label info on your end, both the tracking number and the embedded address. That way if there’s ever a mismatch, you can file a stronger claim with USPS or your return platform.
-
Manually inspecting return requests
If your return volume is still manageable, consider enabling “manual approval only” for all return requests. That gives you a chance to vet the reason, check past order behavior, and ensure customers don’t shortcut the system.
Happy to share more if you’d like to explore this kind of setup, there are definitely ways to prevent this sort of incident from happening again.
One pattern we started tracking recently is refund ratio per customer, not just individual return cases.
For example:
orders: 10
refunds: 6
refund ratio: 0.60
Customers with ratios above ~0.4–0.5 often end up being repeat returners or abusing policies across multiple orders.
Many stores only look at each return request separately, which hides the pattern. When you aggregate the customer history it becomes much easier to spot.
We’ve been experimenting with automatically calculating this from Shopify orders + refunds and flagging customers when the ratio crosses a threshold.
Curious if anyone else here is tracking something similar.