Product recalls: trace affected units and remedy completion
A product-recall announcement identifies a hazard and an affected population, but does not establish that customers have received an effective remedy. Trace units through notification, response and completed correction. The useful output is a cohort-level progress table, not a binary label that treats an announced recall as an eliminated operating exposure.
Start with the official campaign scope
CPSC’s monthly progress-report page describes reports that show the progress of recall remedies after an announcement. This article uses US consumer-product recalls in the CPSC setting. Vehicle, food and medical-product campaigns can follow different authorities and reporting systems; do not merge them under one assumed completion definition.
Copy the campaign identifier, models, serial ranges, manufacture dates, distribution geography and stated remedy. A product name is rarely a sufficient population key. A second campaign can cover overlapping models but a different defect or production period. Preserve the campaign relationship before adding affected-unit counts.
A hypothetical unit-status ledger
Assume a fictional appliance campaign covers 100,000 units. At the cutoff, 80,000 are with consumers, 15,000 are at distributors and 5,000 remain with the manufacturer. Internal stock is quarantined, and distributors confirm that 12,000 of their units are quarantined. Consumers have completed a remedy for 20,000 units.
| Hypothetical population | Units | Observed action |
|---|---|---|
| Manufacturer inventory | 5,000 | Quarantined |
| Distributor inventory | 15,000 | 12,000 quarantined |
| Consumer-held | 80,000 | 20,000 remedied |
| Total campaign population | 100,000 | Different action states |
The consumer remedy rate is 20,000 / 80,000 = 25% under this worksheet’s explicit denominator. Calling it 20% uses the whole campaign population and answers a different question. Quarantining 17,000 unsold units prevents further distribution but does not repair 17,000 consumer-held products.
Separate contact from correction
Assume 40,000 consumers received notices and 30,000 submitted valid requests, while the completed consumer remedies remain 20,000. Notification coverage is 50% of consumer-held units; valid-request coverage is 37.5%; completed-remedy coverage is 25%. These are stages, not additive percentages.
A request acknowledgment or a mailed repair kit may precede actual hazard removal. Follow the campaign’s stated completion evidence: installation confirmation, repair record, replacement delivery, refund plus required disposal verification, or another documented method. The example does not prescribe a universal legal completion rule.
Customers may dispose of low-value products without contacting the manufacturer. Such action can reduce hazard exposure while remaining outside an observed correction count. Preserve this possibility as unmeasured; do not simply assume that every nonresponding unit has been discarded safely.
Look for the bottleneck
Review when valid requests accumulate faster than completed remedies, when a repair becomes unavailable, or when distributor stock continues appearing for sale. The first suggests a remedy-capacity or parts issue; the second changes the promised action; the third calls for channel enforcement evidence.
Keep cumulative and monthly figures separate. If a later report repeats 20,000 cumulative completions, that is not another 20,000 corrected units. Record status at each cutoff and calculate incremental changes only after checking whether the campaign population was revised.
The worksheet should include affected units by location, contact attempts, valid requests, completed corrections, quarantined unsold units and unresolved reconciliation differences. Public datasets may omit some stages or rely on manufacturer submissions. Identify the evidence source and avoid inventing a complete operational funnel from one headline.
Bound the investment inference
This task evaluates recall progress, not ordinary warranty failure rates, accounting provision classification or litigation probabilities. Completion can reduce one hazard exposure while costs, customer claims or monitoring continue. A high correction percentage is therefore evidence about a defined remedy population, not proof that every commercial or legal consequence has ended.
FAQ
Does recall announcement mean the products are already safe?
No. Notification and effective correction are different events. Check the actual remedy progress for the affected units.
Should quarantined stock count as repaired consumer units?
No. It is a separate action on a different population. Report both results with their own denominators and avoid double counting.