Internal Cost Audits & Wastage Reduction Financial Tracking

A cost sheet states what a garment should cost. Reality rarely agrees. Fabric prices move, markers waste more than planned, labour efficiency slips, and overhead creeps upward. Without an audit, the factory discovers these gaps only when the bank balance disappoints at month end, by which time the evidence is cold and the responsible teams have moved on. An internal cost audit closes that gap: it compares standard cost against actual cost, element by element, in taka, while the production records are still fresh and the corrective action can still help the next order.

The audit is not an exercise in blame. Its purpose is to find taka that is leaking from the factory and plug the leak. A variance of BDT 11.40 per garment sounds trivial until it is multiplied by 40,000 garments and becomes BDT 456,000 in a single month. Internal audits make small numbers visible at their true scale, and that visibility is what drives action.

Standard cost is the planned cost per garment from the approved cost sheet. It is the benchmark, and it must be frozen for the audit period. Actual cost is what the garment really cost, gathered from stores issues, payroll, and the overhead allocation for the same styles and the same period. Variance is actual cost minus standard cost. A positive variance is unfavourable, because the factory spent more than planned. A negative variance is favourable, because it spent less.

The sign convention causes constant confusion, so every audit pack should label each variance in words — favourable or unfavourable — rather than relying on plus and minus signs alone. A reader who misreads a sign will celebrate a loss or panic over a saving, and either mistake wastes management time.

A total variance without a breakdown is nearly useless. Consider the monthly audit of a basic T-shirt style produced at 40,000 pieces. Standard CM per garment is BDT 185.00 and actual CM is BDT 196.40, giving an unfavourable variance of BDT 11.40 per garment, which is 6.16 percent above standard. That total alone suggests nothing actionable. The breakdown tells the story: fabric plus BDT 6.20, trims plus BDT 1.80, labour plus BDT 2.10, and overhead plus BDT 1.30, which sum exactly to BDT 11.40. On 40,000 pieces, the total variance is BDT 456,000, of which fabric alone accounts for BDT 248,000. The corrective action therefore starts with fabric — marker efficiency and cutting wastage on this style — not with a general exhortation to cut costs.

Each element needs a single owner. Fabric variance belongs to cutting and marker planning. Trims variance belongs to stores and merchandising. Labour variance belongs to industrial engineering and production. Overhead variance belongs to finance and factory administration. When one person owns one element, follow-up is simple: ask that person what changed and by when it will improve.

Cutting wastage deserves its own section in every audit because it is large, visible, and fixable. The method is to measure the physical excess first and convert to taka second. Take the Gazipur woven bottom factory in the factory example. Fabric booked for the quarter is 60,000 metres at BDT 280 per metre. Standard waste is 5 percent, which is 3,000 metres. Actual waste is 7.5 percent, which is 4,500 metres. Excess waste is 1,500 metres, and at BDT 280 per metre the excess cost is BDT 420,000. Nobody can argue with 1,500 metres once cutting and stores agree on the count, and the taka value then follows automatically.

End-bit loss works the same way. On 1,200 kilograms of fabric, a standard end-bit loss of 2 percent is 24 kilograms and an actual loss of 3.5 percent is 42 kilograms, so excess loss is 18 kilograms. At BDT 650 per kilogram, the excess cost is BDT 11,700 for the period. Small on its own, but end-bit loss recurs on every fabric lot, and a factory that tracks it will eventually fix its lay planning and remnant policy.

Rejected panels should also be counted as wastage, but tracked separately from marker waste, because the corrective action is different. Marker waste is fixed by better markers; rejected panels are fixed by better cutting quality and fabric inspection. Mixing the two in one number hides both causes.

The best audit packs are short. One page per style, produced monthly, showing standard cost, actual cost, and variance for fabric, trims, labour, and overhead, plus a wastage section in physical units and taka. The pack for the T-shirt example fits on one page: five rows, five columns, and a short note naming the largest variance and its owner. A factory head can read ten such pages in fifteen minutes and know exactly where money leaked that month.

The pack must compare like with like. Actual costs for one product mix must not be compared against standards for a different mix, because style changes manufacture false variances. When the product mix changes, restate the standard for the styles actually produced, and record the restatement openly. Comparability is the audit principle that matters most after honesty about standards.

An audit finding without a corrective action is only an observation. Every variance element above an agreed threshold — for example, 3 percent of standard cost or BDT 100,000 in a month — must produce a corrective action with three attributes: one named owner, one deadline, and one taka savings target. The Gazipur factory marker target illustrates the format: return cutting wastage from 7.5 percent to 5 percent within two months, owned by the marker planning head, saving an estimated BDT 420,000 per quarter at current volumes.

The follow-up audit next month checks whether the variance shrank. If it did, the action worked and the new level becomes the expectation. If it did not, the investigation reopens, because either the action was not implemented or it addressed a symptom rather than the cause. Variances that repeat for two consecutive months escalate to the factory head. This escalation rule prevents the most common audit failure: findings that are noted, filed, and forgotten.

Audits do not only recover losses; they reveal permanent savings. In the Gazipur example, a 1.5 percent marker efficiency improvement applied to annual fabric usage of 200,000 metres saves 3,000 metres. At BDT 280 per metre, that is BDT 840,000 per year, every year, from a single improvement in marker planning. Marker efficiency gains compound because they apply to every future order in the same fabric types, unlike one-off savings that disappear when the order ends.

This is why audit teams should always ask a second question after finding a variance: is there a structural improvement here, not just a recovery? A structural improvement — a better marker library, a revised lay policy, a new fabric inspection standard — keeps paying long after the audit month is forgotten.

Sometimes an audit reveals that the standard itself was wrong. A fabric that was costed at one consumption actually needs more because of a pattern change, or a labour standard that assumed one machine type no longer matches the machines on the floor. In that case the standard must be revised, but openly: document the evidence, obtain approval from the costing head, record the old and new standards side by side, and restate the variance against the corrected standard. Quietly moving a standard to erase a variance is the fastest way to destroy trust in the entire audit system, because once production suspects the standards are manipulated, no variance will ever be accepted again.

Costing or accounts should perform the audit, because production auditing itself creates a conflict of interest. Production teams must cooperate fully — providing cutting records, stores issues, and attendance data — and must own the corrective actions, but the measurement should be independent. The audit meeting should be short and factual: this is the variance, this is the evidence, this is the owner, this is the deadline. Factories that turn audit meetings into blame sessions soon find that records become harder to obtain, and the audit dies from lack of cooperation.

Internal cost audits are the mechanism by which a factory learns what its garments really cost. Freeze the standard, collect actuals, break the variance into elements, convert waste into taka, assign every variance an owner and a deadline, and re-audit next month. A factory that follows this cycle monthly will steadily convert waste into margin, and its cost sheets will become more accurate with every cycle, which improves every future quotation as well.