Cost of quality

Cost of quality is a measurement model. It adds what a business spends getting quality right to what it loses getting it wrong, across four buckets: prevention, appraisal, internal failure and external failure. It runs as long as you make anything.

That four bucket shape is the whole difference between this and its nearest neighbour. Cost of poor quality counts only the two failure buckets. It is a strict subset of cost of quality — not a synonym.

It is a model, not an analysis technique and not a system. Cost center analysis measures one department’s spend. Cost of quality measures one thing across every department: what conformance and failure actually cost you.

And it has no end date. A cost transformation program closes when its targets land. Cost of quality keeps running, because the moment you stop measuring prevention, prevention is the first thing to get cut.

Key takeaways

• Four buckets: prevention, appraisal, internal failure and external failure.
• Cost of poor quality is the two failure buckets only, so it sits inside this model as a subset.
• It is a standing measurement model, not a repeatable analysis and not a program with an end date.
• Money moved into prevention usually buys a larger reduction in failure cost.

How it works

The model sorts every quality related dollar into one of four buckets, then watches how the mix shifts over time. Healthy operations spend more on the first two buckets — and much less on the last two.

The mix tells you more than the total does. A business whose failure buckets dwarf its prevention spending is paying for quality either way, just at the worst possible point in the process.

The first two buckets are conformance cost and the last two are failure cost. The point of the model is the trade — a dollar spent on prevention usually removes more than a dollar of failure.

Deciding what to measure comes first. Critical-to-quality characteristics name the few attributes a customer actually judges you on, and those are the ones worth inspecting.

Then you need a defect rate you can compare across processes. Defects per million opportunities puts complex and simple products onto one scale, which lets appraisal spending follow real risk.

The split between internal and external failure is the one to watch. Quality escape rate measures how many defects get past your checks and reach a customer, which is exactly that boundary.

Set a baseline before you set a target. Most operations underestimate the failure buckets badly, because rework absorbed inside a normal shift rarely gets recorded as a quality cost at all.

Prevention needs an owner. Quality assurance is the function that builds the controls, and its budget is the clearest single signal of how seriously a business takes the model.

Examples

Regulated manufacturing is where the four buckets are most visible, because prevention spending is written into law and audited. Software, professional services and food production all run the same model under a different vocabulary.

Medical devices are the clearest case. From 2 February 2026 the Food and Drug Administration’s Quality Management System Regulation incorporates ISO 13485:2016 into 21 CFR Part 820, making prevention a compliance obligation.

Self assessment frameworks push the same logic into services. The National Institute of Standards and Technology runs the Baldrige program, which publishes tools that identify gaps and compare you with other organizations.

Outsourced customer service applies the model without the factory language. Prevention is agent training and call design, appraisal is call scoring, internal failure is a repeat contact, and external failure is a churned customer.

Software teams renamed the buckets but kept the model. Code review is prevention, automated test suites are appraisal, a bug caught in staging is internal failure, and a production incident is external failure.

Construction shows internal failure at full scale. Rework on site burns materials, labor and schedule, and it is always cheaper to catch a drawing error in design review than on the third floor.

Food production shows the external failure bucket at full size. A single recall carries destruction cost, retailer penalties, regulatory reporting and a brand hit — the last of which outlives the product by years.

Related terms

• Quality assurance manager: the role that owns prevention and appraisal spending day to day.
• Defect density: defects per unit of size, a measure the failure buckets often run on.
• Quality assurance: the function itself; cost of quality is how you price what it does.
• Cost of poor quality: the two failure buckets only, a strict subset of this model.
• Critical-to-quality: names the attributes worth spending prevention and appraisal money on.

FAQ

What are the four cost of quality buckets?

Prevention, appraisal, internal failure and external failure. The first two are conformance cost, spent on purpose. The last two are failure cost, which nobody chooses to spend.

Is cost of quality the same as cost of poor quality?

No. Cost of poor quality covers only the internal and external failure buckets, which makes it a strict subset of cost of quality rather than another name for it.

Should you aim for zero cost of quality?

No. Prevention and appraisal spending is deliberate, and the target is the lowest total across all four buckets rather than the lowest figure in any single one. Cutting prevention to zero simply moves the money into failure.

Who owns the measurement?

Quality assurance usually owns the defect data and finance owns the valuation. Both are needed, because engineers count defects and accountants price them. Neither number is worth much on its own.

Does the model apply to outsourced work?

Yes. Build the four buckets into the service agreement so failure cost stays visible instead of being buried inside the provider’s margin.

Explore provider capability by sector in the Outsource Accelerator hubs.