A part fails inspection. Someone fills out a form, a disposition is assigned, and the nonconformance disappears into a database. The product moves, the line keeps running, and the problem comes back next Tuesday.
This is the reality of nonconformance management (NCM) in most manufacturing organisations. The system designed as a frontline defence against defective product has become a rubber-stamp machine for disposal decisions. The corrective action you were supposed to link to remains an empty column because linking it meant doing actual work.
Nobody notices the degradation because the NCM reports still go out every month. The dispositions happen within 24 hours, the auditors check the boxes, and everyone assumes the system works. Right up until the customer calls. ISO 9001, AS9100, and IATF 16949 all demand that you identify, record, evaluate, disposition, and learn from nonconformances. The keyword is learn. If your system is not extracting information to prevent recurrence, it is just counting defects.
The Form-First Disease
The NCM form exists in the QMS software or as a paper traveller attachment, and people fill it out because they cannot close the job without one. But the typical entry reads: Description: 'Part out of spec.' Root cause: 'Operator error.' Disposition: 'Rework.' Corrective action: 'Retrained operator.' Verified by: [signature].
That is not nonconformance management; it is storytelling. 'Operator error' is a lazy categorisation that absolves the system of responsibility. 'Retrained operator' is a ritual, not a corrective action. Nobody verified anything; they signed because the form has a line that demands it.
This form-first disease is the most common failure mode in quality management. People optimise for completing the form rather than understanding the failure. The metric becomes 'NCM closed within 24 hours' instead of 'NCM prevented from recurring.' Because the metric rewards speed over depth, the system fills with shallow, meaningless records.
The Material Review Board Treadmill
In many plants, the Material Review Board (MRB) has devolved into a rubber-stamp committee. They meet daily or weekly, review the stack of nonconformances, assign dispositions, and move on. The problem is that the MRB has become a disposal system rather than an investigation system.
They ask 'what do we do with this part?' when they should also ask 'what does this failure tell us about our process?' I have seen MRB meetings where 40 nonconformances are dispositioned in 30 minutes. You cannot conduct root cause analysis in 45 seconds. You can barely read the defect description in 45 seconds. You can only look at the part type, say 'rework,' and move on.
This pace forces a false choice between production speed and technical rigour. The MRB's legitimate job is to make a time-sensitive disposition decision: rework, repair, scrap, or return to supplier. But that tactical decision must trigger a strategic investigation, not replace it.
The Use-As-Is Trap
'Use-as-is' is a legitimate disposition. Sometimes a part that fails drawing requirements is perfectly functional, and the schedule impact of scrapping it outweighs the technical risk. The concession process exists for exactly this reason. But use-as-is has a dark side.
When use-as-is becomes the default disposition because it is faster than arranging rework, the NCM system has failed. I have walked into plants where the use-as-is rate was over 60 per cent. Six out of ten parts that failed inspection were shipped anyway.
When I asked how they justified it, the answer was always 'engineering reviewed and approved.' When I asked to see the engineering review, it was a signature on a form. No analysis, no risk assessment, no data. Just a signature from an engineer under pressure to keep the line running. When use-as-is becomes routine, your engineering tolerances are wrong, your process capability is inadequate, or your quality culture is broken.
Disposition Health Indicators
Recurrence and the Database Nobody Opens
Here is the test that tells you whether your NCM system works. Pick any nonconformance from six months ago. Look at the part number, the defect type, and the root cause. Now search for the same combination in the last month. If you find it, your system has failed. You detected it, dispositioned it, closed it, and learned nothing.
Recurrence is the ultimate metric of nonconformance management. I have never visited a plant that tracked recurrence rate. They track NCM count, closure time, disposition breakdown, and scrap cost. They never track the one metric that tells them if any of it is working. Tracking recurrence would mean confronting the truth: most of their nonconformance management is theatre.
Meanwhile, your NCM data sits in the QMS database, untouched. The patterns — recurring failures, problematic part numbers, problematic suppliers — remain invisible because nobody looks for them. You are collecting enormous amounts of data and doing nothing with it. This data should be your roadmap for improvement.
If the same failures keep coming back, your system is a disposal pipeline, not a learning system.
Separating Disposition from Investigation
Stop trying to do root cause analysis in the MRB meeting. The MRB's job is to decide what to do with the nonconforming product right now. That is a legitimate, time-sensitive decision. Root cause investigation is a different activity requiring different people, timeframes, and methods. It must happen after the disposition.
Assign the investigation to someone who has the skills and time to do it properly. Use 5 Whys, Ishikawa, or fault tree analysis. Track the investigation as a separate activity with its own deadline and owner. 'Operator error' must be banned as a root cause entry. It is the beginning of an investigation, not the end.
Ask why the operator made the error. Was the work instruction unclear? Was the fixture inadequate? Was the gauge unreliable? If the person entering the NCM lacks the skills for root cause analysis, route it to a quality engineer who does.
The Post-Disposition Investigation Cycle
- 01MRB DispositionTactical decision: rework, scrap, or use-as-is. Timeframe: hours.
- 02ScopingAssign investigation owner, define the 5 Whys scope, set deadline.
- 03Root Cause AnalysisTechnical investigation beyond 'operator error' using Ishikawa or FMEA logic.
- 04CAPA ImplementationExecute systemic corrective action and update PFMEA and control plan.
- 05Recurrence VerificationMonitor the same part and defect code for 90 days to confirm zero recurrence.
Activating the Data and Leadership
Your NCM database is a goldmine. Run Pareto analysis on defect types. Track nonconformance rates by part number, supplier, process, and shift. Feed these findings back to engineering, supplier quality, and production. The data you already have is more valuable than any new quality initiative you are about to launch.
Nonconformance management is not a quality department activity; it is a business activity. The trends, costs, and recurrence patterns must be on the operations review agenda every month. Not as a passive report-out, but as an active discussion: What are our top three recurring nonconformances, what corrective actions are in progress, and when will we see results?
Leadership attention is the most powerful driver of NCM performance. When the plant manager asks about recurrence rates, people investigate root causes. When nobody above the quality manager cares, the system atrophies. Every manufacturer has a nonconformance system. The difference between a learning system and a disposal system is the commitment to investigate failures, implement corrective actions, and verify they worked.
That commitment costs time up front. It means slower dispositions on complex nonconformances and more investigation hours. It means holding product while the root cause is determined instead of shipping it and hoping. The return is fewer defects, lower costs, and a quality culture people believe in.