Non-Conformance Report

A warehouse receives 100 components and finds five outside the specified dimension, so staff create an NCR with the lot, measurement, requirement and quantity affected, and the units should be controlled while a disposition is decided. ASQ's eight-disciplines problem-solving guide separates describing a problem, containing it, verifying root causes and validating corrective actions, and an NCR can feed those later steps without pretending they are complete at first report.

ASQ's root-cause guidance discusses identifying underlying causes instead of treating symptoms alone, so a guessed cause in an NCR should be labelled a hypothesis until tested. Cite the requirement, because a drawing number, contract clause, service target or procedure version establishes the standard and "looks wrong" is too vague for repeatable action.

Describe the observed difference, including measurement, date, location, photographs or test results where useful, keeping facts and opinions distinct. Identify affected items by lot, serial, order, customer or process period to trace what is at risk, since a sample defect may require checking the rest of a batch.

Contain immediately if necessary by blocking stock, pausing a process or notifying an owner so unapproved material does not continue downstream; containment is temporary risk control, not final correction. Choose a disposition through authorized roles, since rework, repair, return, scrap or acceptance by concession may be options under specific requirements and a line worker should not silently accept an out-of-spec item.

Record approvals, because a customer waiver or engineering concession may be needed, especially where safety or regulated specifications matter, and approval scope and expiry should be clear. Assess cause proportionately, as a one-off label error and a repeated critical defect need different investigation depth, and root-cause analysis should use evidence.

Distinguish correction from corrective action: replacing five bad components fixes those units, while changing a faulty machine setup may prevent recurrence, and the report should show both if needed. Assign an owner and due date, because an NCR with no responsible person can remain open while affected stock occupies space or customers wait, and verify the fix by sampling later output or reviewing a trend, since closing because a training email was sent does not prove improvement.

An illustrative defect incidence is nonconforming inspected units divided by inspected units under the same inspection method, so five of 100 gives 5% for that inspected batch, which is not the supplier's long-term defect rate. Keep cost separate, since scrap, rework, returns and downtime can be tracked by the NCR ID but a defect count alone is not a cost measure.

Link external claims by connecting any customer complaint, shipment and NCR to avoid duplicate investigations with conflicting conclusions, and review repeated categories, because several NCRs for the same machine and dimension may point to a systematic cause and trend analysis should not erase individual records. Protect traceability: if a lot is released after rework, the record should show what changed and which checks passed, without overwriting the original failure.

Respect local or industry rules, since medical, food, aviation and other regulated settings can require particular handling and a generic NCR template is not enough. Separate detection from responsibility, because finding a defect in receiving does not prove the supplier caused it (transit damage, handling and measurement error are possible), make the record easy to retrieve with a stable identifier tied to purchase orders and batches, and remember that for owners an NCR is a reliable problem record and decision trail that helps contain harm now and supports evidence-based improvement later.