The effectiveness-review gap — why 60% of CAPA audit findings stem from one missing step

A SCAR comes in. The quality manager opens it. The team picks a fix. The fix gets implemented. The CAR gets closed. Six months later the same defect signature shows up again, the customer auditor finds the original CAR was closed without verifying the fix held, and the shop loses a point on the scorecard.

This is the most common failure mode in IATF 16949 corrective-action workflows. It is not that fixes do not ship. It is that nobody verifies the fix stuck.

The IATF audit data is consistent on this. Roughly 60% of CAPA findings during certification audits trace back to insufficient effectiveness review: the step between "we did something" and "we proved it worked." In the IATF 16949 standard, clause 10.2.4 is explicit about it. Corrective actions must be reviewed for effectiveness, and that review must produce documented evidence.

Most shops handle this poorly for one of three reasons.

First, the CAR system has no concept of effectiveness review. The CAR log spreadsheet has columns for "issue," "action," "owner," "due date," and "status." It does not have a column for "verify the fix held three lots later." When the action ships, the row turns green. The auditor reads that row twelve months later and asks: how do you know it worked?

Second, the team is overloaded. The quality manager is the person who personally feels every missed corrective action and who is also the only person who would think to schedule a follow-up. There is no automation. There is no system. There is the quality manager and a calendar they forgot to add the reminder to.

Third, and this is the one that surprises people: the action itself was correct. Containment held. The root cause was diagnosed. The fix worked. But the proof that the fix worked never got documented, because the closure happened at "fix shipped," not at "fix verified."

The pattern Soltreya enforces is mechanical. Every corrective action gets a closure gate: a follow-up task scheduled at a date informed by the part's production volume (next three lots, next batch, or 60 days, whichever comes first). The gate is not optional. The CAR stays open until somebody attaches evidence (measurement data, inspection results, customer confirmation) to the gate.

When the auditor walks in twelve months later and pulls the CAR record, they see the issue, the action, the owner, the close date, and the effectiveness evidence stamped with a timestamp and a signature. The conversation goes from "let's talk about what you did" to "next finding."

This is the cheapest, highest-leverage thing a small quality team can systematize. Not because the action work is hard, but because the verification of the action is what an auditor is paid to find. Make that part automatic, and the most common audit finding stops being a finding.