SCOR DS · Return · caselet
Three times the returns, one disposition decision at a time
A firmware defect triples weekly returns. Blanket-scrap recovers ~10% of value in salvage; per-unit disposition recovers 71% — and cuts the customer's credit wait from 21 days to 6.
The situation. A connected-device maker ships a firmware release with a defect that surfaces in the field over weeks. Returns climb to 3.2× baseline — about 410 units a week — and the returns desk falls back to the historical default: scrap and credit.
The signal. The Returns & Service agent flags the volume anomaly, ties it to one firmware version from service records, and — the useful part — classifies the incoming stream: most units are healthy hardware with bad firmware. Systems consulted: RMA system (return reasons, serials), service/IoT telemetry (firmware versions), ERP (credit memos, inventory), and warranty entitlements.
The decision. Blanket-scrap is simple and destroys value — scrap-and-credit recovers salvage only, around 10%. The agent recommends disposition per unit: 62% qualify for reflash-and-test (~$14/unit against a $96 replacement cost) and return to stock as refurbished; 28% route to bench repair; 10% scrap. Credits under $400 with valid entitlement auto-approve rather than queueing for manual review.
The write-back. Disposition codes and refurb work orders in the ERP; auto-approved credit memos issued; the firmware-version linkage logged to the quality case.
The outcome.
| Metric | Per-unit disposition | Blanket scrap |
|---|---|---|
| Value recovered | 71% | ~10% (salvage) |
| Credit cycle time | 6 days | 21 days |
| Reflash cost per unit | ~$14 | — |
The takeaway. Return is where margin goes to die quietly. Treating disposition as a decision — not a default — is the difference between a cost center and a recovery operation.
Representative scenario; figures are synthetic. SCOR is ASCM’s framework.