The cost of poor quality is everything you pay because something was not right the first time: scrap, rework, sorting, special freight, credit notes, warranty, and the orders that quietly went elsewhere. Almost all of it is already in your P&L — booked under production, logistics and service, almost never under "quality".
A plant manager once told me his cost of poor quality was 1.2 percent of revenue. He had the number from his scrap account, and the scrap account was clean. That week his logistics team had booked three special freight runs to keep a customer line going, his second shift had spent two evenings sorting, and his sales colleague had issued a credit note for a late delivery. None of it was in the 1.2 percent. None of it was wrong, either. It was simply booked somewhere else.
That is the whole problem in one paragraph. The cost of poor quality is not hidden because somebody hides it. It is hidden because nobody ever gave it a name.
Quality costs nothing. Failure does.
Philip Crosby made the distinction in 1979: quality is free, what costs money is the unquality. The cost of quality model splits spending into four categories. Two are the price of getting it right: prevention (process design, training, mistake-proofing) and appraisal (inspection, testing, audits). Two are the price of getting it wrong: internal failure, caught before delivery, and external failure, caught by the customer.
The first two have budgets. Somebody approved them, somebody owns them, and they show up in the quality department's cost centre. The second two have no budget, because nobody plans to fail. So they land wherever the work was done.
The four categories of the cost of poor quality
| Category | Example | Usually booked as | Visible as quality cost? |
| Prevention | process design, training, mistake-proofing | quality or engineering budget | yes |
| Appraisal | incoming inspection, final test, lab | quality cost centre | yes |
| Internal failure | scrap | scrap account | yes — the one everybody knows |
| Internal failure | rework hours | production hours | no |
| Internal failure | sorting shift | temporary labour or overtime | no |
| Internal failure | re-inspection, retest | appraisal, by mistake | no |
| External failure | special freight | logistics | no |
| External failure | credit note, price concession | sales deduction | no |
| External failure | warranty, field return | service | no |
| External failure | customer audit, escalation meetings | salaries | no |
| Opportunity | the next order, the churned customer | nowhere | no |
Read the right-hand column. The two categories with budgets are visible. The two categories without budgets are visible exactly once, in the scrap account. Everything else is in the P&L and carries another label.
Why most of it stays unlabelled
It is not carelessness. It is how cost accounting works.
A booking needs an account, and the account was decided years ago by whoever set up the system. Rework happens in production, so it is production time. The sorting shift is staffed by temps, so it is temporary labour. The freight is a freight invoice. Each decision is locally correct. Nobody configuring a cost centre structure was asked “and which of these are failure costs?”, because that question belongs to a different department.
The result is a cost that is fully recorded and completely distributed. You cannot add it up because there is no line to add. And what has no line gets no owner, and what has no owner never gets reduced.
Lorem ipsum dolor sit amet, consectetur adipiscing elit. Ut elit tellus, luctus nec ullamcorper mattis, pulvinar dapibus leo.The pizzeria version. The owner knows what a burnt pizza costs: flour, cheese, oven time. That is his scrap account. He does not know what the Friday evening shift spends remaking them, because that is wages. He does not know what the refund on the delivery app cost, because that is a platform fee. He does not know how many customers ordered elsewhere the following week, because nobody books that anywhere. Four failure costs, one of them visible. He is not careless. Nobody ever gave the other three a name.
Three states: in view, partly in view, blind spot
Every cost block in your company is in one of three states.
In view. The cost has a line, an owner and a monthly figure that survives a follow-up question. Scrap usually lives here.
Partly in view. A figure exists. It is pulled when somebody asks, by somebody who knows where to look, and it is different each time. Rework, sorting and special freight usually live here. This is the dangerous state, because a number exists and looks solid.
Blind spot. The cost exists and nobody books it. Firefighting hours, the customer who went elsewhere, the audit your customer now runs on you. Almost never from carelessness. Nobody ever gave it a name.
Most companies believe they are in the first state and are in the second. Walking your five cost blocks through the free calculator takes ten minutes and usually surfaces the first blind spot — not because it names the state, but because you have to go looking for every figure.
How to find the first numbers
You do not need a project for this. You need five questions and an afternoon with somebody from controlling.
- Scrap accounts. Start with what is visible. Note the figure and the account number. This is your anchor.
- Special freight. Ask logistics for every freight invoice above the standard rate in the last three months, and the reason. Most reasons are quality reasons.
- Sorting and rework hours. Ask production for overtime and temp hours booked against containment, sorting or rework. If there is no such booking code, that is already a finding.
- Credit notes and concessions. Ask sales for every credit note in the last quarter and the reason behind it. Late deliveries and quality complaints are failure costs, whatever the booking says.
- Opportunity. Do not try to calculate it. Write down one customer you lost or one order that went elsewhere in the last year, and the reason if you know it. That single line is worth more than an estimate.
You now have five numbers of very different quality. That is fine. The point is not accuracy. The point is that each of them now has a name, and the gap between what you had before (one number) and what you have now (five) is the finding. If you want somewhere to write them down, the failure cost capture template has the five rows already in it.
If one of the five keeps coming back from the same cause, that is where a structured method such as the 8D report earns its place. Repeat failures are the most expensive kind, because every repeat carries the full cost again.
The honest part
You will find a figure in the literature that says the cost of poor quality is 15 to 40 percent of revenue. I had it in an earlier version of this article. I took it out, not because it is wrong but because it is unknowable in that form: it comes from surveys decades old, quoted forward from book to slide to consultant deck. A range that wide is not a finding, it is an admission that nobody measured. The only figure that means anything is your own.
You will also find the famous U-shaped cost of quality curve, with an “optimum” somewhere short of zero defects. It comes from the first edition of Juran’s Quality Control Handbook in 1951, and Juran himself revised it in the fourth edition of 1988 once processes had become capable enough to make the optimum move. Do not look for the optimum. Look for the labels.
And a limit to be clear about: a number is only as good as the system that produces it. If your rework has no booking code, the five questions above will give you an estimate, not a figure. That is not a reason to stop. It is the first thing to fix — and that is what cost of quality reporting is for.
Conclusion
The cost of poor quality is not a number you calculate. It is a set of costs you already pay, booked under other names. Scrap is the only one with its own account, which is why it is the only one anybody knows. The rest sits in production hours, logistics invoices, sales deductions and service budgets. Give them a name and they get an owner. Give them an owner and somebody reduces them. No curve, no benchmark and no consultant estimate replaces that first step.
If you want to know which of your five cost blocks you could put a defensible figure on today, start with the calculator: ten minutes, no sign-up. It will give you a number. The more useful result is the list of blocks where you had to guess.
Frequently Asked Questions
What is the cost of poor quality?
Everything a company pays because something was not right the first time: scrap, rework, sorting, special freight, credit notes, warranty claims and lost orders. It is the failure half of the cost of quality model, internal and external.
What is the difference between cost of quality and cost of poor quality?
Cost of quality covers four categories: prevention, appraisal, internal failure and external failure. Cost of poor quality is only the last two, the price of failure. Prevention and appraisal are the price of getting it right.
Where do I find the cost of poor quality in the accounts?
Mostly outside the quality cost centre. Rework sits in production hours, sorting in temporary labour or overtime, special freight in logistics, credit notes in sales deductions, warranty in service. Scrap is usually the only item with its own account.
How high is the cost of poor quality typically?
Published ranges exist, but they come from old surveys and vary so widely that they say little about your company. The only meaningful figure is your own, built from your scrap account, freight invoices, rework hours and credit notes.
Why does nobody see these costs?
Because every booking needs an account, and the accounts were set up by function, not by cause. A sorting shift is booked as labour because it was labour. Nobody was asked whether it was a failure cost. The cost is fully recorded and completely distributed.
