Cost of Poor Quality
COPQ and total cost of quality, as money and as a share of revenue.
Ref: Juran’s Quality Handbook · Feigenbaum · BS 6143-2 · ASQ
IPrevention Good quality cost
IIAppraisal Good quality cost
IIIInternal failure Cost of poor quality
IVExternal failure Cost of poor quality
VScale
VIThe four categories
VIIQuality cost summary
| Measure | Value | Unit |
|---|
VIIIWhat the ratios say
| Indicator | Value | Reading |
|---|
IXConditions at analysis
XHow to read the number
The prevention-appraisal-failure model splits quality spending into money spent stopping defects, money spent finding them, and money lost because they happened. The first two are investments you chose. The last is a bill you were sent.
The single most informative output here is not the total — it is the prevention-to-failure ratio. A plant spending far more on failure than on prevention is paying to discover its defects rather than paying to not make them, and that is a spending mix you can change without any new equipment. The classic pattern in a plant that has never measured this is appraisal-heavy and prevention-light: a lot of inspectors, very few capability studies.
The external-to-internal ratio is the second thing to look at, because it is an escape-rate signal. Failures found in-house are expensive; failures found by your customer cost several times more for the same defect, and they spend your reputation as well as your money. If external failure is a large share of the total, the containment problem is upstream of the cost problem.
Two honest cautions. First, published benchmarks put total cost of quality somewhere between 5% and 30% of revenue depending on industry and maturity, which is a range so wide it should be used to sanity-check your arithmetic and not as a target. Second, this is an estimate assembled from accounts, not a measurement. It is exactly the right tool for sizing a business case and exactly the wrong tool for claiming a saving afterwards — for that you need the measure on the process itself, before and after, which is a different job than this page does.
This is one of the bench instruments, standing alone
The calculator above takes numbers you type, one set at a time. On the bench, the same solver runs against a dataset you keep — versioned, with the conditions held beside the result, so the number can still be explained to someone who asks six months from now.