How a Luxury Handbag Brand Audits Partner QC Systems: A 40-Point Checklist, a 100-Point Score and What an Audit Cannot Prove

What an Audit Costs, and What It Saves

An audit of a QC system costs about three person-days: two reviewing documents at a desk and one on the bench. Its return is measured in the arguments and the rework it removes from the following season, and in the questions it answers that would otherwise stay open for months.

The traceability exercise deserves its reputation. Asking a team to take one finished carton and walk backwards to the hide batch it was cut from reveals more about a system in ninety minutes than any checklist item, because it cannot be prepared for in advance. Ask also for the forward direction, from that batch to every carton it appears in, since a recall narrows by lot rather than by season.

The Six Sections of a 40-Point Checklist

Forty points across six sections is enough to cover a system without becoming a day of box-ticking, and the weighting matters more than the count.

• Incoming materials, eight points: hides received against a declaration, shade banding, batch separation.
• First-piece and in-process control, seven points: approval records, tolerance checks at the operation, operator reference at the bench.
• Finished inspection, eight points: the plan, the defect library, conditioning, the room and its logger.
• Traceability and records, seven points: lot codes, retention, who owns the file, retrieval time.
• Equipment and calibration, five points: measuring tools dated within twelve months, test pieces controlled.
• Corrective action, five points: root cause recorded, action dated, effectiveness checked at the next lot.

The weighting is deliberate. Traceability and corrective action together carry more weight than equipment, because a system with excellent instruments and no record of what it did with them cannot answer a question eleven months later.

Records: Read Twelve, Then Follow One Lot

Reading twelve records before arrival is what turns the site visit from a collection exercise into a verification exercise, and it is the part most audits skip.

1. The inspection plan for the last completed lot, with its standard, level and sample size named.
2. The inspection report for that lot, with the count by defect class.
3. The defect library in use, with its revision date.
4. The approved sample record, and the seal log for the retained reference pieces.
5. The hide declaration for the batch used, with tannery identification and batch reference.
6. The hardware finish declaration, with plating thickness stated.
7. The needle issue and return log for the period of that lot.
8. The metal detection record, with the sensitivity test piece size recorded.
9. The calibration register for measuring equipment used in inspection.
10. The corrective action log, with root cause entries and closure dates.
11. The packing specification and the carton label format.
12. The retained file index, showing who can produce all of the above within one working day.

Then follow one lot. Pick a carton from the finished store, read its code, and ask for the eleven other records that belong to it. A system that produces them in under an hour is a system that will answer a question in eleven months.

Calibration, Needle Control and Metal Detection

Three controls are audited because their failure is invisible in the finished piece and catastrophic when it reaches a customer.

Measuring equipment calibration

Within the last twelve months, with the register matching the tools actually on the bench

Needle control

Issue and return reconciled per operation, with a broken fragment recorded and accounted for

Metal detection sensitivity

A ferrous test piece around 1.0 mm, checked at the start of each shift and logged

Retained reference pieces

Sealed at approval, signed across the seal, three held per style and colourway

Inspection room condition

Held near 45-55% relative humidity and 18-22 degrees, with the reading logged

File retrieval time

One working day, by a named owner, for any record in the set

Needle control is the one founders find uncomfortable to raise and the one no audit should omit. A fragment left inside a construction is a critical defect by any definition, and the only evidence it did not happen is a reconciliation log rather than an inspection.

Scoring: 100 Points and Three Bands

Scoring converts observation into something a procurement meeting can act on, and three bands are enough to avoid a false precision that invites argument.

The automatic fail line is what makes the score honest. Without it, a system can accumulate points for tidy documents while missing the single control that protects a customer.

Corrective Action: Five Days to Cause, Thirty to Closure

A finding without a root cause and a date is an observation, and observations accumulate for years without changing anything at all.

• Root cause recorded within five working days of the finding being issued, not at closure.
• Action described as a change to a method or a record, not as a reminder to be more careful.
• Closure evidenced by a document produced after the change, not by a letter confirming it was made.
• Effectiveness checked at the next lot rather than at the next audit, which is usually twelve months away.
• Recurrence treated as a new finding on the system, since a fix that did not hold was not a fix.

The third line is the discriminator. A letter saying the change was made is a claim; a record produced after the change showing the new method in use is evidence, and only the second one closes a finding. The fourth line matters for the same reason: effectiveness checked twelve months later, at the next audit, is effectiveness checked too late to protect the season in between.

Who Performs the Audit, and How Often

Three arrangements exist, and the right one depends on whether the brand has anyone internally who has done this before.

Twelve months is the normal interval once a system passes, shortened to six where the score was conditional and to immediate re-verification where critical findings were raised. Intervals should be written into the agreement rather than remembered, since remembered intervals lengthen.

What an Audit Cannot Establish

Four claims are regularly expected of an audit, and none of them is within its reach, which is worth stating before the report is read rather than after.

• That the next lot will be good: an audit examines a system, and a system can be followed on one lot and not on the next.
• That the materials comply: that is a declaration question, answered by documents from the tannery and the finisher.
• That the piece will last: durability is answered by load and cycle testing rather than by any record of method.
• That the price is fair: an audit says nothing about commercial terms, and conflating the two corrupts both.

What an audit does establish is narrower and more useful than any of those: whether the system can produce the evidence that would answer a question, if a question were ever asked.

Turning the Audit Into an Agreement

The output worth keeping is not the score but the six lines both parties sign, because those are what the next lot is actually run against.

The surrounding programme is unchanged: MOQ 500 pcs per style and colourway, sampling in 6-10 working days and 12-15 for complex constructions, bulk in 35-50 days, inspection at AQL 2.5 under ISO 2859-1, and handover on FOB Xiamen terms. Export experience on this programme runs since 2014 across 7 production lines with about 200,000 pcs per month of capacity, delivered through our validated partner factory Quanzhou Yingming Bags.

The arithmetic that justifies the three days is short. Reworking one closure on a piece that has reached a customer costs around USD 4.80 in parts and bench time before freight, and one season of avoided rework returns the audit many times over.

Frequently Asked Questions

People Also Ask