Supplier corrective action plans in China: move from a failed inspection to verified control.

A supplier corrective action plan in China should begin with control of the affected shipment. Buyers often jump directly from a failed inspection to a discount negotiation or a promise that workers will be more careful. Neither action proves that defective units were contained or that the same problem will not recur.

The response should be proportionate to product risk and contract terms. Safety, regulatory or serious functional issues require appropriate specialist review. Use the existing quality inspection process to connect the report, specification and release decision.

Pause release and define containment

Identify the purchase order, SKU, lot, quantity, locations and shipment status. Tell the supplier which goods must not ship and how they should be marked or segregated. If part of the order may be unaffected, require evidence before splitting it from the held quantity.

Containment is immediate risk control, not the final correction. Examples include stopping packing, isolating suspect cartons, screening all units for a clearly detectable issue or preserving samples for investigation.

Describe the nonconformity with evidence

Use the approved requirement and actual result. Include defect photographs, measurements, sample counts, carton references and inspection method. Avoid descriptions such as “quality is bad.” A useful statement tells another person what failed, where it was found, how often and against which specification.

Separate correction from corrective action

Correction addresses the current defect: rework, replacement, sorting or repacking. Corrective action changes the system that allowed it: a fixture, parameter limit, material check, work instruction, training control, first-piece approval or in-process inspection. Both may be needed.

A 100% sort can remove visibly defective units from this lot, but it does not automatically prevent the next lot from producing the same defect.

Investigate both creation and escape

Ask why the defect occurred and why the supplier’s controls did not detect it. Root-cause tools such as five-whys or a cause-and-effect review can organize thinking, but the final cause needs evidence. “Operator carelessness” is incomplete unless the process design, instruction, tooling, workload and detection controls are examined.

Confirm whether the problem affects other variants, dates, components or orders. A material batch or shared process may extend the scope beyond the inspected SKU.

Assign actions, owners and dates

Each action needs one responsible owner, a completion date and a deliverable. Link actions to the identified cause. If dimensional variation came from a worn fixture, retraining alone is unlikely to be sufficient. If artwork was wrong because an obsolete file remained available, remove the file and improve revision access.

Control rework carefully

Approve the rework method before it begins when the activity can damage appearance, function or compliance. Define tools, acceptance checks, handling and traceability. Ask for counts: units checked, units repaired, units rejected and units remaining. Rework samples should be reviewed when the process changes the product.

Verify effectiveness before release

Choose verification that matches the failure: repeat measurements, functional tests, packaging checks, process evidence or a reinspection of the corrected lot. A supplier’s completed spreadsheet is not effectiveness evidence by itself. For sampled workmanship decisions, follow the defined AQL inspection plan without changing defect classification after seeing results.

Make the commercial decision separately

After technical verification, decide whether to accept, reject, rework, replace, ship partially or revise timing under the contract. Consider customer impact, compliance, inventory need, reinspection cost and future risk. A price concession does not make an unsafe or unusable product acceptable.

Carry learning into the next production run

Update the specification, approved sample notes, control plan and supplier performance record. Add a focused during-production checkpoint if late detection created high rework cost. Review the corrected risk before the next purchase order and monitor whether the action remains in use.

Set escalation rules for repeated failures

A single corrective action should not reset the supplier record to zero. Track whether the same defect, process or detection gap returns in later lots. Define escalation thresholds such as increased inspection, earlier production checkpoints, management review, temporary order limits or supplier replacement analysis. The response should reflect severity and recurrence rather than frustration.

Close the CAPA only when actions are completed, evidence is reviewed and effectiveness is demonstrated over an appropriate period or production cycle. Keep open risks visible on the next order checklist so the team does not rely on memory.

FAQ

Should every inspection defect require a formal CAPA?

No. The response should match severity, recurrence and business risk. Major, systemic or repeated failures deserve stronger documented action.

Is reinspection always required?

Not universally, but the buyer needs appropriate verification. Reinspection is often useful when the lot was sorted or reworked after failure.

Who pays for correction and reinspection?

That depends on the contract, responsibility and negotiated remedy. Define these expectations before production whenever possible.