Key takeaways
Short answer: A nonconformance is any departure from a specified requirement — even a cosmetic or paperwork deviation. A defect is a nonconformance that affects fitness for use, function or safety. Every defect is a nonconformance, but a minor nonconformance is not necessarily a defect. The distinction drives disposition, liability and how you read your quality data. See also scrap vs rework.
A nonconformance is any failure to meet a specified requirement, regardless of severity. A wrong label, a missing signature, a cosmetic blemish — all are nonconformances, logged for disposition even if the product works perfectly.
A defect is a nonconformance serious enough to affect fitness for use — function, safety or customer impact. It carries higher liability and urgency, and it is the subset of nonconformances that actually matters to whether the product is good.
A run produces 100 units with a faint scuff on the housing (a cosmetic nonconformance) and 3 units with an under-torqued safety fastener (a defect). Counting all 103 as "defects" makes quality look five times worse than it is and wastes effort; counting the 3 safety units as mere nonconformances understates a real liability. Graded correctly, the 100 scuffs are use-as-is or rework and the 3 fasteners follow a strict defect path. The wording decided both the data and the risk response.
Treating every nonconformance as a defect inflates apparent quality problems and wastes effort; treating a real defect as a minor nonconformance understates risk and liability. Clear criteria keep the data honest and the response proportionate.
Nonconformances get graded by severity — use-as-is, rework, scrap, reject. Defects, especially safety-related, follow stricter paths. The grading is only meaningful if the terms are used consistently across the plant and over time.
1. Counting all nonconformances as defects. Quality looks far worse than it is and effort is wasted.
2. Downgrading real defects. A genuine safety issue logged as a minor nonconformance hides liability.
3. Inconsistent criteria. Different graders call the same issue different things.
4. No link to OEE Quality. Only true defects should reduce the Quality rate.
Only true defects should reduce the OEE Quality rate as bad parts. Counting every minor nonconformance as a reject distorts OEE downward; ignoring real defects flatters it. Consistent grading keeps the Quality term honest.
Fabrico captures quality events with reason codes so defects and minor nonconformances are graded consistently and only true defects hit the Quality rate. Book a demo to see honest quality data.
No — only those affecting fitness for use, function or safety.
It drives disposition, liability and how your quality metrics read.
True defects (bad parts), not minor nonconformances.
Quality and engineering, against the specification.
Your quality data becomes unreliable and risk responses become disproportionate.
Login
Blog
