The CFO's Appraisal Haircut
Valmere Plastics runs a 320-person injection-molding plant. Last fiscal year the quality function recorded: prevention (training, SPC, design review) $60,000; appraisal (inspection, lab, audits) $240,000; internal failure (scrap, rework) $310,000; external failure (warranty, returns, and one lost account worth $180,000 in annual margin) $520,000. A new CFO has proposed cutting the appraisal budget by 30% "because inspection does not add value," using the savings to offset a freight overrun. The quality manager has countered by requesting $150,000 to expand prevention activities. The plant director wants a decision this week and has asked you, as quality lead, for a recommendation grounded in the numbers rather than in principle.
1. Classify the four cost categories, compute total cost of quality and the share attributable to failure, and state whether the CFO's proposal is likely to raise or lower total COQ. Show the arithmetic you would put in front of the director.
Read the worked answer
Classification and total
| Category | Amount | Share |
|---|---|---|
| Prevention (good) | $60,000 | 5.3% |
| Appraisal (good) | $240,000 | 21.2% |
| Internal failure (bad) | $310,000 | 27.4% |
| External failure (bad) | $520,000 | 46.0% |
| Total COQ | $1,130,000 | 100% |
Failure costs = $830,000, or 73.5% of total COQ. Prevention - the only category that reduces the number of defects created - is the smallest line on the page.
The CFO's proposal
The $72,000 saved (30% of $240,000) does not remove a single defect; it removes the screen that currently catches defects before they become warranty claims. Internal failure cost per defect is already roughly one-tenth the external cost per defect in this data ($310k internal vs $520k external). Tearing out appraisal converts some fraction of the $310k in scrap and rework into external failure at a multiple of that rate. Even a modest 15–20% escape rate wipes out the $72,000 several times over.
The defensible answer is not "inspection adds value" - it is "appraisal is a symptom of a process that still creates defects." The direction of travel should be prevention up, appraisal down over time, failure down hard. Cutting appraisal with no change to defect creation just moves cost to the least visible and most expensive bucket.
2. The $520,000 external figure includes a $180,000 lost account representing forgone future margin. Should that be in the analysis? Recompute the key ratios without it and say what changes - and what does not.
Read the worked answer
Recomputation
Excluding the lost account, external failure = $340,000 and total COQ = $950,000.
- Failure costs = $650,000 = 68.4% of total (was 73.5%)
- Prevention = $60,000 = 6.3% of total (was 5.3%)
- External failure = 35.8% of total (was 46.0%)
What changes: the size of the headline number and the drama of the story. A one-time account loss is an opportunity cost, not a recurring operating cost, and mixing it with warranty spend in a trend line will mislead next year's comparison.
What does not change: the structural finding. Failure costs still dominate at roughly seven dollars for every dollar of prevention, and appraisal-plus-prevention is still a small fraction of what nonconformance costs. The recommendation is identical.
How to present it: show the two versions side by side rather than choosing one. Keep the lost account in a separate line labelled "revenue at risk" so the operating COQ stays clean year over year, but do not let the director forget it - the plant has demonstrated that its external failure rate is high enough to cost a customer outright, which is the strongest argument in the room for the prevention spend.
3. The quality manager wants to justify the $150,000 prevention request using the "1-10-100 rule." How should you handle that, and what would make a stronger case to a CFO?
Read the worked answer
Handle the rule as a hypothesis, not evidence. The 1-10-100 ratio is a field heuristic drawn from other people's plants. Quoting it to a CFO who has just been handed an internal cost table invites the reply "show me that ratio in our numbers," and there is nothing to show - the plant has no measurement of what a defect costs at each stage of its own flow.
Stronger case, in three moves:
- Build the ratio from Valmere's own data for one product family: cost to prevent a defect at the press, cost to catch it at final inspection, cost to settle it as a warranty claim. Even a rough internal ratio is more persuasive than a borrowed one, and the exercise itself identifies where the money actually leaks.
- Pilot, don't promise. Ask for a defined slice - one cell, two presses, two quarters - with a pre-agreed baseline (scrap %, first-pass yield, warranty claims per 1,000) and a pre-agreed kill criterion. A $150,000 programme defended by a pilot's measured result is fundable; the same $150,000 defended by a rule of thumb is a budget fight.
- Frame in the CFO's units. Not "fewer defects" but "$X of scrap and warranty avoided against $150,000 of spend, break-even at Y% reduction." State the break-even explicitly so the director can accept the risk without accepting your optimism.
Watch for: prevention spend is easy to approve and easy to evaporate. Tie the funds to named activities (SPC on the top three scrap parts, design review gate, supplier development) with owners and dates, or the $150,000 becomes training that nobody attends.