“The continuous quest for quality” is one of those phrases that sounds unarguable and means almost nothing. A quest is a journey with no end — and in a plant, a journey with no end and no defined destination is just spending. Too often it looks like this: a shop builds a scrap rate into its costing, beats that number on a good run, and pats itself on the back for “improving quality.” It missed the point entirely.
Quality’s real goal is not to catch defective parts. It’s to make a higher percentage of saleable parts per run — which is to say, quality’s real goal is to improve profit. Catching defects is damage control after the fact; making good parts in the first place is the win. Once you see quality as a profit lever rather than a virtue, a sharper question replaces the slogan: not “are we pursuing quality?” but “how much quality is worth paying for here?”
This guide doesn’t reproduce historical WJT Associates material. It uses the same applied logic — quality is an investment that must clear an ROI hurdle — to think about the cost of quality and where enough is enough.
Quality is a profit decision, not a virtue
Every quality program is a profit-improvement program, and every profit-improvement program has to survive a return-on-investment look. That sounds obvious and is routinely ignored, because “more quality” feels morally unassailable — who argues against quality? But the cost of quality can climb so high that it erodes the very profit it was supposed to protect. Spending more to inspect, sort, document, and chart than you save in fewer failures isn’t quality; it’s overhead wearing quality’s clothes.
The useful reframe is that quality has an optimum, not a maximum. Below the optimum you’re shipping problems; above it you’re burning money chasing a perfection the part doesn’t need. The job is to find that point, and it’s different for every part.
The four costs of quality
To find the optimum, it helps to see that quality cost isn’t one thing — it’s four, and they trade against each other:
| Cost category | What it is | Trend you want |
|---|---|---|
| Prevention | Stopping defects before they happen — training, process control, robust design, good tooling | Spend here; it’s the cheapest place to buy quality |
| Appraisal | Finding defects — inspection, testing, gauging, audits | Necessary, but more of it doesn’t make parts better |
| Internal failure | Defects caught in-house — scrap, rework, downtime before shipping | Lower by prevention, not by inspecting harder |
| External failure | Defects that reached the customer — returns, sorting, freight, lost trust | The most expensive by far; the one prevention exists to avoid |
The leverage in this table runs one direction: money spent on prevention reduces the other three. Investment in training, process discipline, or better tooling buys down scrap, inspection load, and customer returns at the same time. The shops with the lowest total cost of quality aren’t the ones inspecting the hardest — they’re the ones preventing the most.
Inspecting harder doesn’t make parts better
This is the trap appraisal sets. When rejects climb, the reflex is to add inspection — more checks, more gauges, more sorting. But inspection doesn’t create good parts; it only separates the good from the bad after both have already been made. You’ve added cost to find the problem without removing the cause that makes it.
A process that needs 100% inspection to ship acceptable parts isn’t a quality success because the inspection catches the bad ones — it’s a quality failure that the inspection is papering over. The real fix is upstream: a stable, documented process that makes saleable parts without being sorted. Appraisal is a cost you want to reduce by prevention, not a cost you proudly increase.
Internal versus external failure: the cost asymmetry
Not all failures cost the same, and the gap is enormous. A defect caught in-house is an internal failure — you scrap or rework a part, lose some material and time, and move on. A defect that reaches the customer is an external failure, and it carries costs the scrap bin never sees:
- Sorting and return freight, often of a whole lot
- The customer’s line disruption, which they remember
- Your credibility as a reliable supplier — the expensive, slow-to-rebuild one
- In the worst case, the account itself
This asymmetry is why “ship it and let the customer find the few bad ones” is such a costly gamble. The modest bit of inspection you skipped can turn into a returned lot and a customer who now audits every shipment. Prevention and in-house appraisal are cheap precisely because they keep failures on the internal side of that line.
It also reframes what a “reject” even is. A righteous reject — a lot that genuinely shipped wrong and couldn’t pass function or appearance — points to a real gap to close. But a reject pulled from a hostile audit, where one out-of-hundreds dimension was flagged without anyone checking whether the part actually worked, is a different animal: it’s expensive, but the cure is a clear, agreed acceptance standard, not more inspection. Knowing which kind you’re dealing with decides whether you spend on prevention or on the specification.
How much is enough? Where the data stops informing
So where’s the line? Two principles locate it.
First, the ROI line. Add quality cost as long as each increment saves more than it costs — fewer failures, more saleable parts, protected accounts. Stop when the next increment costs more than the failures it prevents. That point is real, and disciplined shops find it instead of assuming “more is always better.”
Second, Deming’s rule on data. The point of being “in control” and gathering data is to learn something you can act on. When the charts and measurements stop telling you anything new — when the process is stable and the data just confirms it, run after run — continuing to gather and chart it is no longer quality work. It’s busywork with a clipboard. At that point you stop, hold the control you’ve achieved, and put the effort where it still buys something.
| Symptom | What it usually means | The move |
|---|---|---|
| Rejects high, inspection rising | Prevention is underfunded | Spend upstream — process, training, tooling |
| 100% inspection just to ship | A process being sorted, not controlled | Stabilize the process; reduce appraisal |
| Charts that never reveal anything new | Data past the point of informing | Stop gathering; hold control; redeploy effort |
| Tolerances tighter than function needs | Over-quality — paying for unused margin | Match the spec to the part’s actual job |
The part sets the bar
Finally, “enough” isn’t a universal number — the part decides it. A safety-critical, regulated, or high-consequence part justifies heavy prevention and appraisal because the cost of an external failure is catastrophic. A low-criticality throwaway part does not, and holding it to the same regime is pure over-quality. Matching the quality investment to the cost of failure for that part is the whole discipline. Quality at an optimized level isn’t the most quality you can buy — it’s the right amount for what the part has to do.
FAQs
What is the “cost of quality”?
It’s the total cost of achieving (and failing to achieve) quality, usually split into four parts: prevention (stopping defects up front through training, process control, and good tooling), appraisal (finding defects through inspection and testing), internal failure (scrap and rework caught in-house), and external failure (defects that reached the customer — returns, sorting, lost trust). Seeing it as four interacting costs, rather than one inspection budget, is what reveals that spending on prevention lowers the other three at once.
Doesn’t more inspection mean better quality?
No. Inspection finds defects; it doesn’t prevent them. A process that needs heavy inspection or 100% sorting to ship acceptable parts is being papered over, not fixed — the parts were already made bad before anyone checked them. More appraisal adds cost without removing the cause. The way to actually improve quality and lower its total cost is to invest upstream in prevention so the process makes saleable parts without being sorted, which then lets you reduce inspection, not increase it.
How do I know if we’re spending too much on quality?
Two tests. Apply an ROI check: if the next increment of quality spend costs more than the failures it prevents (and the saleable parts it gains), you’ve passed the optimum. And watch your data — if your control charts and measurements have stopped revealing anything new because the process is stable, continuing to gather and chart them is busywork, not quality work. Over-quality also shows up as tolerances held tighter than the part’s function requires, which is paying for margin nobody uses.
Why is a customer reject so much more expensive than an in-house reject?
Because of where it’s caught. An internal failure costs you scrap, rework, material, and time — bounded and over with. An external failure — one that reached the customer — adds return freight, lot sorting, disruption to the customer’s line, and damage to your reputation as a reliable supplier that can take years to rebuild, and sometimes the account itself. That asymmetry is why skipping prevention or in-house inspection to save a little is such a poor gamble: it risks pushing a cheap internal failure across the line into an expensive external one.