Every garment factory pays for poor quality. Few of them know how much. Rework is absorbed into the line, rejected pieces are written off quietly, and the inspection team is treated as a fixed overhead rather than as part of the cost of quality. The result is that quality is discussed as a technical problem — defect rates, AQL levels, inspection reports — while its real language, money, is never spoken.
The cost of poor quality, or COPQ, translates quality performance into that language. It is the money a factory spends because the work was not right the first time. This guide explains what counts as COPQ, how a garment factory can calculate it from data it already collects, and — the part most guides skip — how a buyer can read a supplier’s COPQ figures to judge whether a quote is built on solid ground.
The four categories — and a distinction that matters
Quality costs are conventionally split into four categories, in the prevention–appraisal–failure (PAF) model — the four-way split of quality costs that the American Society for Quality (ASQ) uses, and that the literature attributes to Armand Feigenbaum’s work on quality costs in the 1950s:
- Prevention costs. Money spent to stop defects happening: training, quality planning, process design, supplier qualification, preventive maintenance.
- Appraisal costs. Money spent checking whether the work conforms: inspection, testing, audits, calibration, supplier surveillance.
- Internal failure costs. Money lost on defects found before delivery: rework, re-stitching, re-cutting, scrap, re-inspection, downtime.
- External failure costs. Money lost on defects found after delivery: returns, chargebacks, complaint handling, warranty work, replacement shipments, lost business.
Here is the distinction that most garment-industry explanations gloss over. In the strict sense, COPQ is the cost of failure only: internal failure plus external failure. The broader cost of quality (COQ) adds prevention and appraisal on top. The Juran Institute defines COPQ as the costs that would disappear if the work were done right the first time and every time, and ASQ defines it the same way, as the costs associated with internal and external failure only. Routine inspection therefore sits outside strict COPQ — but that application of the definition is this article’s, not Juran’s own words, and the line is not absolute: reinspection and 100% sorting triggered by defects are themselves internal failure costs, and disappear when defects do. The broad usage also has a lineage inside the Juran tradition: Joseph A. DeFeo of the Juran Institute argued in “The Tip of the Iceberg” (Quality Progress, May 2001) for counting appraisal alongside failure costs. Neither usage is wrong, but they produce different numbers, and a reader who does not know which definition is being used cannot compare two factories at all. Whenever you see a COPQ figure, ask first: does it include appraisal cost? This article keeps the distinction visible throughout, and flags it again in the worked example below.
How a factory can calculate COPQ in eleven steps
The practitioner method below is adapted from Online Clothing Study’s step-by-step guide. It uses data a reasonably organised factory already records: DHU (defects per hundred units), repair times, headcount, wages, rejection rates and shipment values.
Step 1 — Measure rework and rejection levels. From DHU values, calculate the total number of defects that need repair each day (A). A factory running DHU 10 in stitching and DHU 15 in finishing on 1,000 pieces a day is generating 250 defects a day. Two cautions apply from the start. DHU counts defects, not defective garments — one garment can carry several defects — and this step follows the practitioner’s guide in treating every counted defect as a repair job. That is a simplifying assumption, not a measurement: some defective garments cannot be repaired at all, and a garment defective in both stitching and finishing is counted twice. Treat A as an upper-bound estimate of repair workload, not a counted figure.
Steps 2–3 — Measure repair time. Study how long it takes, on average, to remove stitches from a defective garment (T1) and to re-stitch it (T2). Repair time varies by defect type, so these are averages, not constants.
Step 4 — Convert repair time into labour. The number of workers needed for repair work is L = A × (T1 + T2) ÷ 480, where 480 is the working minutes in a standard eight-hour day. Multiply L by the daily wage to get the labour cost of repair work.
Steps 5–6 — Count excess material and re-cutting. If defective parts must be replaced, calculate the cost of the extra fabric issued for re-cutting (FC) and the manpower cost of the additional cutting (CC).
Step 7 — Price the rejections. Calculate how many garments are rejected as a percentage of the order, and value the loss at the cost already invested in those pieces: material, labour and apportioned overhead up to the last operation completed, less any salvage value. That is the standard basis for internal failure cost. Valuing rejected pieces at FOB instead measures something different — the revenue the factory never earns. It is a legitimate way to show management the opportunity loss, but it is not a cost, and it must never be added to cost-based failure figures: FOB includes a profit margin the factory never spent, and mixing the two bases inflates the total. If both views are reported, keep them on separate lines.
Steps 8–9 — Add up the monthly total. Internal failure cost is the sum of alteration labour, rejected garments (at cost, not FOB), excess material and re-cutting — with one control the method needs and the guide does not state: a garment that fails rework and is then rejected is counted once only, so rejected pieces are excluded from the repairable defect count, or the rework labour already booked against them is deducted from the scrap cost. Monthly COPQ on the broad definition is that figure plus the cost of the quality team (the appraisal element). Note what this method leaves out entirely: prevention cost has no step and no amount anywhere in the eleven steps, and external failure cost — claims, returns, chargebacks — is absent too. The total is therefore a partial figure even on its own terms.
Steps 10–11 — Express it as a percentage. Divide monthly COPQ by the value of monthly shipments. A percentage, not an absolute figure, is what allows comparison across months and across factories of different sizes.
A worked example — with its limits stated
The following example comes from the same practitioner guide, and it is explicitly presented there as hypothetical data, not a real factory’s accounts. It is worth walking through because it shows how the arithmetic works — and where the judgement calls sit.
A factory produces 1,000 pieces a day with DHU 10 in stitching and DHU 15 in finishing, giving 250 defects a day. Assuming five minutes per defect to repair, that is 1,250 man-minutes, or 2.6 man-days of repair work a day (1,250 ÷ 480 minutes in an eight-hour day). Over a 25-day month that is 65 man-days; at an average worker wage of INR 10,000 a month, or INR 400 a man-day (10,000 ÷ 25), the repair labour cost is 65 × 400 = Rs 26,000 a month. Two cautions travel with this arithmetic. First, the guide’s own shorthand — “2.6 × 10,000” — is dimensionally wrong (man-days per day multiplied by a monthly wage); the number comes out right only because 25 × 400 happens to equal 10,000. Second, the 250 defects are an upper bound on repair jobs, not a count of them: DHU counts defects, not defective garments, and a garment counted in both stitching and finishing DHU is counted twice.
After finishing, 1% of garments are rejected — 10 pieces a day. The guide values them at an FOB of $5 a piece, with $1 equal to INR 80, giving Rs 4,000 a day, or Rs 100,000 a month. Read that line as the guide’s convention, not as a cost: FOB is the revenue never earned, and the guide does not publish a cost figure for the rejected pieces, so the cost actually sunk in those ten pieces cannot be stated from the guide’s data — and no strict-definition total can be computed from the guide’s figures at all, as explained below; the Rs 100,000 is retained here as the guide’s own illustration, flagged as lost revenue rather than failure cost.
The appraisal cost is the quality team’s salaries: one quality head at Rs 60,000 a month and 11 checkers — two fabric checkers, one cutting-room checker, four in stitching and four in finishing — at Rs 15,000 each, totalling Rs 225,000 a month.
Adding the three elements as the guide does gives a monthly COPQ of Rs 351,000. Monthly shipments of 25,000 pieces at $5 FOB are worth Rs 10,000,000, so COPQ is 3.51% of shipment value.
Label it plainly: this is an illustrative partial total — not a comparable COPQ benchmark: 3.51% of shipment value (Rs 351,000 / Rs 10,000,000), broad definition, FOB-valued rejections, prevention and external failure excluded. It is not a COPQ figure at all under either the strict (internal + external failure) or broad (plus appraisal) definitions — it is a partial, mixed-basis illustration: rework labour at cost, rejected pieces at FOB, and appraisal cost, with prevention and external failure missing entirely.
Read that number carefully. First, it is an illustration, not a benchmark: no published industry-wide COPQ benchmark for garment factories is cited in the sources consulted for this article, and any supplier quoting a “standard” percentage is inventing one. The closest published figures come from Bheda’s survey of 61 Indian apparel manufacturers, reported in his 2009 report for GTZ, which found average cost of quality — a broader measure than COPQ — of 14.05% of sales; its scope differs, so it cannot be used as a yardstick for 3.51%. Second, the 3.51% is sometimes described as a broad-definition figure because it includes the Rs 225,000 appraisal cost — but that description is too generous: as the label above states, it is not a COPQ figure under either definition, since the rejection element is priced at FOB, not at cost. But the guide’s own figures do not yield a clean strict-definition total either: rework labour is Rs 26,000, while the Rs 100,000 rejection element is priced at FOB, not at cost — so the often-quoted strict figure of Rs 126,000, or 1.26%, is itself a mixed-basis number, not a cost-basis one. The true cost-basis total is Rs 26,000 plus whatever the rejected pieces actually cost — a figure the guide never publishes — and no true cost-basis total can be computed from the guide’s figures. Even that mixed-basis 1.26% sits far below Bheda’s finding that internal failure alone averaged 9.86% of sales, but the two are not comparable: the guide’s figure is a mixed-basis illustration, and no cost-basis figure can be computed from its data at all.
Third, the Rs 225,000 appraisal cost counts the quality head’s full salary as inspection, although a quality head’s time also covers management, training and supplier work — only the checkers’ inspection hours are unambiguously appraisal cost, and the head’s salary should be apportioned, not counted in full. Fourth, external failure cost is absent entirely, and so is prevention cost: the example counts nothing for claims, returns or chargebacks, which in a bad month can dwarf everything else, and nothing for the training and planning that might have prevented the defects. The guide’s own author notes that the method is for reference and that a detailed analysis would be needed for an accurate figure.
What a buyer can do with COPQ
Most COPQ guides are written for factory managers. For a buyer, the value is different: COPQ data is one of the few windows into how a supplier actually runs its floor, and it is much harder to stage than a showroom.
Ask a supplier for three numbers: the DHU trend over the last three months, the rework hours logged per month, and the rejection rate by order. A supplier that can produce them quickly is measuring quality as a cost, which is a good sign in itself. A supplier that cannot produce them is pricing blind — its quote may still be competitive, but the buyer has no way of knowing whether the price depends on quality the factory cannot see.
Then read the numbers in combination — as hypotheses to test, not diagnoses. A low rejection rate with a high DHU could mean the factory is catching defects early and repairing them — expensive in labour, but controlled — or it could mean many defects are minor and cheap to fix while a few serious ones slip through; DHU alone says nothing about severity. A low DHU with a high rejection rate could mean defects are missed until finishing, when they are costlier to fix — or that the factory rejects rather than repairs. Ask for the defect mix before drawing either conclusion. A COPQ percentage that falls month after month while DHU stays flat is worth a question, not applause: it may mean genuine improvement, or it may mean rework is being booked elsewhere.
Used this way, COPQ becomes a negotiation tool with a specific purpose. When a supplier’s price looks too low to cover compliant wages — the check described in our guide to reading a factory quote line by line — a high internal failure cost is often the missing explanation: the factory is absorbing rework it cannot afford, and the buyer will eventually pay for it in delays, claims or a sudden price rise. Conversely, a supplier that can show a falling COPQ trend has a credible story for holding its price while others raise theirs.
Common mistakes when using COPQ
Counting only rework. Rework labour is the most visible element; whether it is the largest or smallest depends on the factory’s defect mix, rework time and rejection rate. Rejection losses and external failure costs are where the money goes.
Mixing definitions. Comparing a factory’s broad COPQ (including appraisal) with another’s strict COPQ (failure only) is meaningless. Always establish the definition first.
Valuing rejections at FOB and calling it a cost. FOB values the lost sale; cost values the wasted input. The two can differ by a factor of two or more, and only the cost basis belongs inside a COPQ total. Report FOB separately as lost revenue if management wants it — never add it to cost-based failure figures.
Counting a garment twice. A piece that fails rework and is then rejected appears in both the repair hours and the rejection loss unless the method excludes it. Agree the rule before the month starts, not after the numbers disagree.
Counting the whole quality payroll as inspection. Checkers’ inspection hours are appraisal cost; a quality head’s salary covers management and prevention too. Apportion it, or say plainly that the figure is overstated.
Reading a single month. COPQ is volatile: one bad order can double it. Trends over a quarter tell the truth; a single month tells a story.
Treating a low COPQ as proof of good quality. A factory that inspects little will record few defects — and a low COPQ. Low failure cost combined with low appraisal cost is not excellence; it is blindness. The healthiest pattern is a falling failure cost alongside a stable, deliberate appraisal cost.
Buyer checklist: five questions to ask a supplier
- What is your DHU trend over the last three months, by section — and can you show the underlying records?
- How many rework hours did you log last month, and how is that time costed?
- What is your rejection rate by order, and do you value rejected pieces at FOB or at cost?
- Does your COPQ figure include the quality team’s salaries, or is it failure cost only?
- What external failure costs — claims, returns, chargebacks — did you incur in the last six months, and what changed as a result?
COPQ will not tell a buyer whether a factory is good. It will tell the buyer whether the factory knows itself — and in garment sourcing, that is usually the more valuable answer.
Frequently asked questions
What is a good COPQ percentage for a garment factory? There is no published industry benchmark to compare against. The 3.51% in the worked example is one hypothetical illustration, not a standard. Compare a factory against its own trend over several months, not against an industry number nobody has published.
Does COPQ include the quality team’s salaries? Only on the broad definition. Strict COPQ is failure cost only — internal plus external failure. Always ask which definition a figure uses before comparing it with another factory’s.
Is DHU the same as the percentage of defective garments? No. DHU counts defects per 100 units, and one garment can carry several defects, so DHU is equal to or higher than the defective percentage. A DHU of 25 does not mean 25% of garments are defective.
Should rejected garments be valued at FOB? Not inside a COPQ total. Value them at the cost already invested — material, labour and overhead, less salvage. FOB measures the revenue never earned and can be reported separately as lost revenue, but it must not be added to cost-based failure figures.
Sources
- Online Clothing Study, “How to Calculate Cost of Poor Quality (COPQ) in Garment Manufacturing?” — practitioner method, 11 steps and hypothetical worked example (Rs 351,000/month; 3.51%). Published 21 October 2015; updated 18 May 2023; verified 29 September 2026. Read 6 November 2026. https://onlineclothingstudy.com/how-to-calculate-cost-of-poor-quality/
- Online Clothing Study, “What is DHU? How to Measure It?” — DHU as defects per hundred units; “a number of defects, not the defective garments”. https://onlineclothingstudy.com/what-is-dhu-how-to-measure-dhu/
- American Society for Quality, “Cost of Quality (COQ)” — COPQ defined as the costs associated with internal and external failure; PAF categories. Read 9 November 2026. https://asq.org/quality-resources/cost-of-quality
- Juran, “The Office: How Poor Quality May Be Hurting Your Bottom Line” — COPQ as the difference between the costs that would disappear if everything were done perfectly the first time and the actual costs of errors. Read 9 November 2026. https://www.juran.com/blog/the-office-poor-quality/
- Juran’s Quality Handbook, 7th edition (ed. Joseph A. DeFeo), McGraw-Hill Education, 2016 — COPQ definition (“all the costs that would disappear if there were no failures”). https://asq.org/quality-press/display-item?item=P1620
- DeFeo, Joseph A., “The Tip of the Iceberg”, Quality Progress, May 2001, pp. 29–37 — COPQ as appraisal and inspection costs plus internal and external failure costs. Abstract read 9 November 2026 (full text requires ASQ membership). https://asq.org/quality-progress/articles/the-tip-of-the-iceberg?id=c5adc775888f4945b84fe2cc8815ea01
- Antonaras, A., Memtsa, C. & Iacovidou, M., “The Challenge of Measuring the Cost of Quality” — the prevention–appraisal–failure categorisation is “attributed to a seminar paper of Feigenbaum in the mid 1950’s”; Feigenbaum is “attributed with having developed the concept of quality cost measurement while working in the General Electric Company in the 1950’s”. Read 9 November 2026. https://pure.unic.ac.cy/ws/files/9354507/The_Challenge_of_Measuring_Cost_of_Quality.pdf
- Bheda, R., “Cost of Quality in the Indian Apparel Industry” — survey of 61 Indian apparel manufacturers (National Institute of Fashion Technology, supported by the Ministry of Commerce, Government of India); figures as reported in Bheda’s 2009 report for GTZ, “Profiting through Quality — Improving Right First Time” (average cost of quality 14.05% of sales; prevention 0.26%, appraisal 3.31%, internal failure 9.86%). The study’s own publication year is given as 2005 in one secondary source and cited as 2015 in another; no journal volume or issue could be verified, and it is not a JTATM article. Read 9 November 2026. https://rajeshbheda.com/wp-content/uploads/2024/07/Profiting-through-Quality-Improving-Right-First-Time.pdf
- JAMA Software, “Cost of Poor Quality (COPQ): Formula and How to Reduce It” — ASQ definition of COPQ and the PAF model (prevention, appraisal, internal failure, external failure); COPQ = internal + external failure; COQ adds prevention and appraisal. Read 6 November 2026. https://www.jamasoftware.com/blog/cost-of-poor-quality/
