Cost of Poor Quality: finding the money that is already in your ledger

Ask a finance team what poor quality costs the business and you will usually get a shrug, or a number somebody made up in a workshop. That is not because the money is invisible. It is because it is already recorded, correctly, under headings that have nothing to do with quality. Overtime. Freight. Credit notes. Stock write-offs. Warranty. Training. Each one sits in a different account, owned by a different manager, and nobody adds them up.

Cost of Poor Quality is the discipline of adding them up. It sorts every cost into four buckets and then asks one question: how much are you spending on failure compared with how much you are spending to stop it. The answer is almost always uncomfortable, and that is the point. A number that makes a steering committee uncomfortable is a number that funds a project.

This walkthrough uses a fictional business throughout. Kettleford Service Group does not exist. The numbers are invented, but they are internally consistent, and every screenshot below is the real Ci Flow tool running on exactly these figures.

Prevention is what you spend to stop defects happening. Training, maintenance planning, better instructions, design reviews. Appraisal is what you spend to find defects before the customer does. Inspection, testing, audits. Internal failure is what defects cost you when you catch them yourself. Rework, scrap, re-diagnosis. External failure is what they cost once the customer has seen them. Return visits, credits, warranty, lost accounts.

The total tells you the size of the prize. The split tells you what kind of business you are running. Spend heavily on failure and lightly on prevention and you are paying to fix the same things over and over. Shift money from the right of that list to the left and the total falls, usually by more than you moved.

The tool enforces one discipline that matters more than the arithmetic. Every line needs an owner and an evidence source. A cost with no source is an opinion, and an opinion will not survive the first finance review.

Kettleford Service Group maintains commercial coffee equipment for cafes and hospitality groups. Around 21.5 million in annual revenue, a workshop, and a field team. The complaint that started the project was not about cost. It was that technicians were busy and customers were still unhappy.

The team spent two weeks pulling eleven cost lines out of five different systems. The job management system gave them repeat visits. Finance gave them credit notes and the warranty ledger. Payroll gave them penalty rates. The workshop gave them job cards and stock write-offs. The training budget and the planner allocation gave them the prevention side.

Nothing here was new spending. Every one of these numbers was already being paid, every month, by somebody who was not thinking about quality when they approved it.

The total came to 173,950 a month, which is 2,087,400 a year. Against 21.5 million of revenue that is 9.71 per cent. Just under one dollar in every ten the business takes.

That percentage is what moves a sponsor, and it is worth being careful with it. It is not profit, and it is not all recoverable. Some of it is the cost of running a service business at all. But a tenth of revenue sitting in failure and inspection is a long way above where anyone expects to be, and it changes the conversation from whether to fund an improvement project to which one to fund first.

Add the annual revenue figure if you have it. Without it the tool still gives you the total and the split, but you lose the one number most likely to be repeated back to you in a board paper.

Failure cost is 89 per cent of the total. Prevention and appraisal together are 11 per cent. That is roughly eight dollars spent on failure for every dollar spent avoiding or finding it, and it is the classic profile of a business that is good at firefighting.

The sharper detail is that external failure alone is 66 per cent. That is failure the customer saw. Internal failure, the kind you catch yourself, is 23 per cent. When the external share is nearly three times the internal share, your inspection is not catching things before they leave, and the first project almost picks itself.

The largest single line is return visits at 573,600 a year. Not the warranty disputes, not the credits, which are the lines people usually name first. The quiet, recurring cost of going back is the biggest one, and it only became visible once it was separated out and annualised.

The tool can read the profile back to you. It names the dominant costs, points at the data gaps, and suggests where projects would go. Used well, it is a second pair of eyes on a table you have been staring at for a fortnight.

It is worth reading the data gaps section carefully, because that is where it is most useful. In this case it flagged three things the team had not recorded: management time spent resolving disputes, any link between failures and specific machine types, and the cost of customers who simply leave. All three are real, all three are missing, and none of them would have shown up by looking harder at the eleven lines that were there.

What it cannot do is decide for you. It has the eleven numbers you gave it and nothing else. It does not know which of your managers will cooperate, which costs are contractually fixed, or that the warranty line is inflated by one bad batch last March. That context is yours.

It will not give you a precise number. Cost of Poor Quality is an estimate built on allocations and judgement, and two competent teams will produce different totals from the same business. Chasing accuracy past the point where the ranking stops changing is wasted effort. You need to know that return visits are the biggest line, not that they are 573,600 rather than 561,000.

It will not find the costs nobody records. Lost customers, damaged reputation, the good technician who resigned because they were tired of apologising. Those are usually larger than anything in the table and they will not appear in it. Say so when you present, or somebody else will.

It will not survive without owners. A cost line with no name against it gets argued away in the first review. The owner column is not administrative tidiness, it is what makes the number defensible.

And it will not tell you the answer, only where to look. A high external failure share says your defects are reaching customers. It does not say why. That is what the Define and Measure phases are for.

The ranked list is a project shortlist. The largest failure lines, with their owners already attached, are the candidates. Take them into project selection and weigh them against effort and strategic fit rather than starting with the biggest number automatically.

The baseline figure belongs in the project charter. If you open a project on first time fix, the 573,600 is the benefit case, and the same table is what you come back to in Control to show the number moved.

Re-run it annually, not monthly. Cost of Poor Quality is a strategic picture, and the effort of pulling eleven lines out of five systems is not worth repeating every quarter. What is worth tracking monthly is the one or two lines you have opened a project against.