Make the cost of a defect traceable

A repair ticket records what happened to a garment. A cost of poor quality ledger connects that event to the resources it used and the loss it caused. The link matters. Without it, a factory can count one rejected piece as scrap, lost sales, replacement production and a customer refund. Some of those entries may describe the same loss.

The useful question is specific: what did this failure cost under the basis we have chosen? A second question follows: what part could a better process avoid? The answers may differ. Historical cost, cash spending and available production time each serve a different purpose.

This guide builds an original example for an apparel factory. It follows an internal label incident and a separate customer claim. Every quantity, wage rate and amount is fictional. Amounts use currency units, abbreviated CU. They are not prices or wage rates for any country. The examples show a management record. Finance is responsible for its link to the accounts.

Separate four kinds of quality cost

ASQ's cost-of-quality framework groups costs into prevention, appraisal, internal failure and external failure. Prevention aims to avoid a quality problem. Appraisal checks whether work meets requirements. Internal failure concerns defects found before the customer receives the product. External failure concerns those found afterwards.

For a garment factory, an approved training activity may be prevention. A planned final inspection is appraisal. Extra sorting caused by a known wrong label belongs to the incident's failure record. A customer repair bill belongs to a different record once the affected product has reached that customer.

The task's purpose helps decide its category. Two people may both be checking labels. One performs the normal release check. The other sorts a held batch after a fault. Record the reason for the work before assigning its cost. Keep the same rule across departments and reporting periods.

ApparelDiary's guide to quality assurance and quality control explains the wider roles. Use it for context on planning and checking. The ledger here follows the cost of a defined event.

Choose the boundary of the ledger

Start with one factory, one reporting period and a clear list of included losses. Record whether the ledger covers finished garments alone or also fabric, trims, samples and services. If a report adds new departments during the year, show the effect of that change. Its totals need the same scope for comparison.

Choose the valuation basis for each type of entry. Labour can use recorded hours and an approved rate. Scrap may use a finance-approved carrying amount. Bought-in repairs use the supported charge. A customer price concession uses the amount actually agreed. These values can share a management report, but its total should not be labelled cash paid.

Write down exclusions as well. A first version might omit uncertain lost orders while retaining them in a separate risk log. That is more useful than adding an unsupported estimate of reputational damage. An excluded item still needs an owner if it could affect a decision.

The reporting boundary also sets the currency and conversion rule. Retain the original amount and currency when a foreign supplier sends a bill. Finance can provide the approved conversion basis. Use the approved exchange rate across departments. Keep a record of any change to the earlier value.

Give each incident and cost line its own identity

An incident ID connects the defect evidence, affected stock, disposition and cost. A line ID distinguishes each cost within that incident. The same incident may create a sorting ticket, a repair ticket and a scrap authorisation. Those are related records, not three separate discoveries of the same garments.

Each cost line should show the activity, quantity, unit, rate, amount and supporting record. Add the cost category, date, responsible reviewer and status. A material issue needs its issue reference. Labour needs a time record and rate basis. A supplier credit needs evidence of approval as well as the original claim.

Retain the difference between estimated, supported and settled amounts. An early estimate can help contain a problem. Label it clearly so readers can distinguish it from a later invoice. When evidence changes an entry, keep the earlier value and the reason for the change.

One garment can have several faults. Give it a traceable unit or bundle reference so its final disposition is counted once. The activity record may contain several repairs, provided their time and material are separate. Several defect observations may refer to one garment. Check the link before using them to count scrapped pieces.

Work an internal label incident from quantity to cost

In incident F241, a factory holds 1,000 garments after finding wrong labels. A complete sort finds 120 affected garments and 880 unaffected ones. The team approves repair for 110 affected garments and scrap for the other 10. The scrapped pieces receive no repair work in this example.

The unit balance is 880 + 110 + 10 = 1,000 garments. After successful recheck, 990 pieces can enter the normal release process. The ten-piece shortfall still needs an order decision. This ledger does not assume that replacement garments were made or that the original order shipped in full.

The extra sort takes 30 seconds per garment. Total sorting time is 30,000 seconds, or 500 labour minutes. At 9 CU per hour, its recorded labour cost is 500 ÷ 60 × 9 = 75 CU.

Repair takes six labour minutes for each of the 110 pieces. The repair work therefore uses 660 minutes, or 11 hours. At 12 CU per hour, it costs 132 CU. These are recorded example times, not standard allowances or factory benchmarks.

Incident F241 entry

Quantity and rate

Amount, CU

Extra sorting labour

500 minutes at 9 CU per hour

75.00

Repair labour

660 minutes at 12 CU per hour

132.00

Replacement labels

110 labels at 0.08 CU each

8.80

Extra recheck labour

110 minutes at 9 CU per hour

16.50

Scrapped garments

10 pieces at 9 CU carrying amount

90.00

Gross recorded failure cost

Sum of the five cost entries

322.30

Confirmed scrap proceeds

10 pieces at 1 CU each

−10.00

Net recorded burden

Gross cost less the stated proceeds

312.30

The rates in the table cover only the named labour. No separate machine or overhead charge is added to those activities. The 9 CU garment amount includes normal production cost incurred before the incident. It excludes the extra sort, repair and recheck costs in the other rows.

Check where double counting can enter

The 110 repaired garments remain useful stock. Their full existing value has not been lost merely because they needed work. Adding 110 × 9 CU to the repair bill would treat those garments as if they were scrapped. Record the extra work and any supported reduction in value instead.

The ten scrapped garments are different. Their 90 CU carrying amount includes the production cost already assigned to them. Adding the original fabric and normal sewing labour again would duplicate part of that amount. Finance should confirm the contents of the rate before it enters the incident record.

Scrap proceeds also need one treatment. The table shows 90 CU gross and a separate 10 CU recovery. A second method could show an 80 CU net scrap loss. Either can reconcile, but subtracting the same 10 CU again would understate the burden.

These management entries do not replace inventory accounting. IAS 2, paragraph 16 excludes abnormal wasted materials, labour and other production costs from inventory cost. It recognises those amounts as period expenses. The factory's applicable accounting framework and policy determine the actual entries. The incident ledger supplies their evidence.

Keep customer claims distinct from factory repairs

Incident E077 concerns a separate shipment that reached its customer earlier. Forty garments need an approved repair. The agreed response includes third-party repair, return transport and a price concession. No full refund or replacement garment is included in this case.

The repair bill is 40 × 2.50 CU, or 100 CU. Transport costs 45 CU. Two hours spent handling the claim are valued at 18 CU per hour, giving 36 CU. A separate concession of 1.20 CU per affected garment totals 48 CU. The concession is explicitly in addition to repair, not another description of the same settlement.

Gross recorded cost is 100 + 45 + 36 + 48 = 229 CU. The supplier has approved a 60 CU credit linked to this event. A further 80 CU remains under dispute. Only the approved credit enters the recovery column, so the net recorded burden is 229 − 60 = 169 CU.

An agreed credit is not necessarily cash received. Its record should show whether it will reduce a payable or be refunded. The pending 80 CU remains visible without reducing the confirmed total. If the supplier later approves it, update the same incident with evidence and the date.

ISO 10002:2018 provides guidance on handling and reviewing customer complaints. Its public scope includes analysing complaints to improve products and services. Keep the settlement amount in the claim record. Record what failed and what the team learned as well.

Reconcile the cost view with finance

For these two incidents, gross recorded failure cost is 322.30 + 229 = 551.30 CU. Confirmed recoveries total 70 CU, comprising 10 CU of scrap proceeds and a 60 CU supplier credit. Net recorded burden is 481.30 CU. The disputed 80 CU is outside that total.

This management measure has a defined scope. It contains labour values, inventory loss and customer charges. It is not a statement that 481.30 CU left the bank during the month. Some wages may have been payable anyway, and some credits may settle later.

Finance should map each line to an existing account entry or explain the difference. Payroll already records the employee expense. Allocating that time to an incident may need no new journal. A second expense journal for the same cost would duplicate it in the accounts.

The ACCA explanation of relevant costs makes a related distinction for decisions. Redistributing an unchanged fixed cost does not create a cash-flow change. Use the historical ledger to understand the event. Assess future choices on their own evidence.

Value a future change against a clear alternative

Before calling a reduction a saving, describe what would happen without the change. Would repairs use existing paid time? Would the factory authorise overtime, buy an outside service or lose a confirmed order? The answer determines the decision calculation.

ACCA's guidance on relevant cash flows focuses on future amounts that change because of the choice. It also considers relevant opportunities given up. Past spending can help explain a failure. Starting a project does not make that past spending avoidable.

Consider a separate planning scenario for the next batch. Without a proposed change, its schedule would require 11 additional repair hours. The overtime rate is 18 CU per hour. The tested change is expected to avoid that work and requires a new fee of 90 CU. Assume the fee and overtime are the only cash flows that differ between the two options.

The conditional net cash benefit is 11 × 18 − 90 = 108 CU. The 18 CU rate applies only to this future overtime decision. The earlier incident keeps its 12 CU valuation rate. The estimate also depends on actually cancelling the overtime requirement.

The fee needs at least 90 ÷ 18 = 5 avoided overtime hours to break even on this narrow basis. If the change avoids only two hours, the result becomes 2 × 18 − 90 = −54 CU. Showing that range exposes the assumption that needs a trial. It is not evidence that the proposed control will work.

Keep time released visible even when cash does not change

Suppose a team removes repair work performed within ordinary paid hours. Payroll may stay the same while operators gain time for useful work. Record that capacity benefit in hours. The production plan must then show whether the available time can serve a real need.

Eleven released hours may be spread among several people and shifts. The factory cannot assume they all become available at a bottleneck. The right skills, machine access, timing and work must coincide. A ledger can record the released effort without promising extra completed garments.

Waiting time also requires care. A held batch may wait overnight while nobody actively works on it. That elapsed delay matters for delivery. It does not represent paid work by an operator throughout the wait. Record the delay and any supported consequence separately from labour activity.

If two people sort for one hour, they use two labour hours. If one person records sorting and repair during the same hour, overlapping entries need review. Use activity logs that let the reviewer reconstruct the work. A total copied from a shift sheet may hide overlap.

Add prevention and appraisal without hiding failure

A wider quality-cost view can sit beside the incident ledger. In the same illustrative reporting bundle, a planned inspection uses 400 labour minutes at 9 CU per hour. That gives 60 CU of appraisal cost. A planned prevention session uses two hours for each of five people, including its trainer. At 15 CU per person-hour, its recorded value is 150 CU.

Component of this limited reporting bundle

Recorded value, CU

Net failure burden from F241 and E077

481.30

Separate planned appraisal activity

60.00

Separate prevention session

150.00

Total of the included records

691.30

This bundle contains only the named records. It is not the factory's total cost of quality. The planned inspection is separate from F241's incident sort and recheck. The same minutes do not appear twice.

Keep the components visible when judging a change. A lower failure total may require more checking. A new prevention cost may appear before the team can measure its effect. ISO 10014:2021 guides managers who seek financial and economic benefits through quality. Its public scope links performance trends with improvement action. It does not supply a universal quality-cost target.

Compare rates using a matched denominator

For F241, 120 affected garments among 1,000 examined give a 12% affected-unit rate. The scrap rate is 10 ÷ 1,000 × 100 = 1%. The repaired share is 11%. These measures describe different outcomes within the same reconciled population.

Net incident burden per garment examined is 312.30 ÷ 1,000 = 0.3123 CU. Label that denominator on the report. Dividing by 990 released pieces would answer a different question. Neither result is the complete cost of making a garment.

A factory-wide cost ratio needs a wider, matched population. Two selected incidents do not represent every factory loss. Dividing their cost by all sales cannot give a complete poor-quality percentage. Unrecorded events, excluded departments and unresolved claims would make that label misleading.

Compare product groups as well as totals when the mix changes. A coat and a simple vest may carry very different values and repair effort. A cheaper product mix could lower the combined cost even if failures did not fall. Keep quantities, rates and total amounts available so that readers can see which factor changed.

Keep discovery, production and settlement dates separate

A claim discovered this month may concern goods shipped several months ago. Its settlement may occur later still. Keep the original production or shipment group, discovery date, accounting period and settlement date in separate fields.

The period report answers what cost or adjustment was recorded during the period. A shipment-group review answers what happened to a defined group of goods over time. Both can be useful, but recent shipments have had less time to generate claims. Older shipments have had more time to reveal faults. Comparing raw totals can make recent shipments look better than they are.

Preserve late updates against the original incident. A recovery approved in November should not erase the evidence that the October report contained an unresolved amount. The report can show the opening position, new charges, approved recoveries and closing position. That makes the change understandable without rewriting history.

Rank losses without confusing cost and risk

ASQ describes Pareto charts using measures such as frequency, cost or time. Choose one measure and one period for each ranking. The most frequent defect need not create the largest recorded loss.

For a separate small example, 100 minor repairs at 0.20 CU each cost 20 CU. Four incidents at 15 CU each cost 60 CU. The first group leads by count; the second leads by cost. These fictional totals show why the ranking should name its unit.

ApparelDiary's Pareto guide for garment defects and rework minutes develops that distinction further. A ranking helps choose investigations. It does not prove a cause or allow a serious product risk to wait simply because its recorded cost is small.

Resolve missing evidence before adding precision

A neat total can still rest on a weak input. If repair time comes from one remembered job, record that limitation. It should not look as firm as a complete set of signed time tickets. Ask which missing input could change the decision most, then improve that evidence first.

Keep the version and effective date of each estimated rate. A change in payroll burden or costing policy may change the recorded amount. The physical loss could remain the same. Show the effect of a revised rate separately when comparing periods. Otherwise the report may reward or penalise a production team for an accounting change.

Use consistent rounding. Keep enough precision in the supporting calculation to reproduce the amount, then round the displayed currency total under the agreed policy. F241's sorting time is 500 minutes, which is not exactly 8.33 hours. Using the original minutes gives the exact 75 CU amount at the stated rate. Rounding the hours first would produce a small difference.

Missing evidence is not a zero cost. Give the item a pending status and explain whether it is excluded from the confirmed total. Equally, a large early estimate is not a settled loss. Show confirmed amounts and estimates separately. Management can then see both the known loss and the uncertainty.

Close the record with evidence and an owner

At review, quality confirms the affected population and final disposition. Production confirms extra work and time. Stores confirms material issues, stock movements and scrap quantities. Merchandising confirms the agreed customer response. Finance checks rates, account links and recovery status.

Ask a reviewer to trace one incident both ways. Starting from the defect, can they find all related costs? Can they also start from a repair invoice or credit note and find the incident? Each route should lead to its evidence. The first route can expose omissions; the second can expose unsupported or duplicate entries.

Close the operational response when the authorised action is complete. Keep financial items open if a bill, recovery or dispute remains unresolved. A closed repair ticket need not imply a settled claim. Assign a person and review date to each open item so it does not disappear from the next report.

The ledger earns its place when it changes a decision. It should show which loss deserves investigation, what evidence supports the amount and which part a proposed change could avoid. Consistent units, clear status and a traceable record make that possible without inflating the cost or promising savings that have not occurred.

References