What it means
Set the population as parcels actually returning or formally initiated for return under the chosen rule, since an unsuccessful delivery attempt alone is not necessarily a return-to-sender event. Use a taxonomy with address error, refusal, failed attempts, customs issue, damage and sender request, and keep unknown when the carrier provides too little detail.
UPS and FedEx describe undeliverable and return-to-shipper statuses, but available reason details differ by service and carrier. Match each event to the original label and change history, not just the customer's present account address, because the customer may have corrected the account after the parcel left.
Compare the address printed on the label with order data at the time of release, since a profile updated days later is not proof the carrier received the newer information. Compare detailed carrier scans with driver or broker notes where available, as a generic undeliverable label may hide a closure, wrong address or customs request.
Preserve a sequence of the first failed event, later contact and final decision, because a cancellation after failed attempts can be the last action without being the cause of the original failure. Where a destination had multiple delivery attempts, capture which attempt failed for which reason, so a later refusal does not erase an earlier address problem that the business could fix.
If evidence supports several contributing factors, document the primary coding rule and retain secondary factors. Avoid speculation and avoid blaming a customer for refusal from a broad carrier status alone; state the observed return fact and seek detail before attributing intent to a person.
For cross-border returns, seek the broker's specific reason before coding customs, since nonpayment, missing permit and wrong declaration can require different upstream fixes even when the carrier uses one broad code. Use an evidence-confidence state for broad or contradictory codes, so a known return status and an unknown cause can coexist without inventing certainty.
A return scan does not prove physical receipt at the sender's warehouse, so track arrival and restocking separately. An unverified reroute request should not overwrite the original address or return cause, and a sender-requested return should be classified only after verifying the request through the authorised account route.
The cause code does not itself settle refund, reshipment or carrier-fee rights, which follow the sale terms and verified physical status. Sample common codes against original carrier evidence, since a sudden rise in address errors could reflect a label export defect rather than more customers typing bad addresses.
Review the unknown category regularly, because missing detail from one carrier may signal a data-integration problem, and keep a dated log of corrections with the first code, reason for change and final reviewer. If a parcel is damaged during transport, preserve photographs and condition notes, since the operational cause can be damage while any compensation or customer remedy remains a separate decision.
In practice
Real-world examples.
Example
A returned parcel has an invalid postal code confirmed against the original label. The reviewer codes it as an address error and records that the order system, not the customer, introduced the mistake.
Example
A customer refusal is supported by the carrier record rather than assumed from a return scan. The note shows the recipient declined the parcel at the door, so the primary cause is coded as refusal.
Example
A vague return event stays unknown until the carrier supplies detail. The team raises a request with the carrier, and the code is updated, with the change logged, once the broker confirms a missing customs permit.
Formula
Calculation
Accuracy = Audited return-to-sender events with verified cause code / All eligible audited returns x 100. Show unknown cause separately.
Worked example. A fictional retailer audits 120 returns in a month against the original label, carrier scans and order data. The recorded cause matches the evidence for 102 of them.
- Accuracy = 102 / 120 x 100 = 85%.
- Of the 18 mismatches, 9 are coded unknown pending carrier detail, 6 were coded refusal but were address errors and 3 were coded address error but were customs holds, so the fixes are targeted at specific coding rules.Case study
Seen in the real world.
This entirely fictional case follows Green Dot Store. Its dashboard blamed customer refusal for several returned orders. A review showed the postal codes were missing on a label export. The team corrected the code and fixed the export rather than contacting customers with a false claim.
No real customer is attributed fault in this case. Before the audit, Green Dot's monthly report showed 45 of 100 returns coded as refusal. After checking each against labels and carrier scans, only 14 were genuine refusals and 22 were address errors from the export, so coding accuracy for the original codes was far lower than assumed. The fix cut address-related returns by more than half in the following quarter.
Watch out
Common mistakes.
- Treating every return scan as a customer refusal.
- Replacing an unknown reason with a plausible guess.
- Using the cause code alone to decide refund entitlement.
Questions
People also ask.
Can a return have two causes?
Yes. Keep contributing factors and a declared primary-code rule.
What if carrier data is vague?
Leave the cause unknown and investigate.
Does cause determine a refund?
No. Apply customer terms and law separately.
From the founder's library
Take it further with the book.
Build your financial confidence beyond this definition. Shihan's full-length guide, Accounting Fundamentals, takes the same plain-English approach and turns it into a complete, practical playbook for non-finance managers, business owners and students - with chapter-end quiz answers and presentation slides included.
25% off with code MMHQ25, applied at checkout. Priced in USD - checkout may show the equivalent in your local currency.
View the book and save 25%Related