Luxury Handbag Quality Control: AQL 2.5 Explained in Four Numbers

AQL 2.5 in Four Numbers

Everything that matters about AQL 2.5 can be stated in four numbers. First, the sample size: 50 pieces drawn from a 500-piece lot at general inspection level II. Second, the accept and reject numbers: 3 and 4 for major defects. Third, the defect classes: critical at zero, major at 2.5, minor at 4.0. Fourth, the switching rules: two rejections in five consecutive lots moves the plan to tightened.

Standard

ISO 2859-1, single sampling, normal inspection

Inspection level

General level II unless agreed otherwise

Lot size

500 pcs per style and colourway

Sample size code letter

H at 281-500 pieces

Sample size

50 pieces

Accept number, major

3

Reject number, major

4

Critical defects

Zero acceptance on any plan

Minor defects

AQL 4.0, accept 5 and reject 6 at 50 pieces

Re-work reserve

Held above the run quantity

The four numbers are a contract, not a technicality. Once they are written on a purchase order, a dispute about quality stops being an argument about whether a piece looks acceptable and becomes a question about a count. That is the whole value of the plan, and it is why the numbers belong in the order document rather than in an email.

What AQL Actually Means, and What It Does Not

AQL stands for acceptance quality limit, and it is the worst tolerable process average when a continuing series of lots is submitted. That definition contains two limitations worth understanding. It is a process average rather than a guarantee about an individual lot, and it is expressed as a percentage only loosely: an AQL of 2.5 does not mean 2.5 per cent of the lot may be defective, it means the sampling plan was built around a process average of that order.

What it does not do is describe the pieces that were not drawn. A plan that draws 50 and accepts on 3 says nothing directly about the other 450, and a buyer who reads an AQL as a warranty will be disappointed twice: once by the standard, and once by the argument that follows. The correct reading is that AQL 2.5 sets a consistent, defensible rule for deciding whether a lot is accepted.

Reading the Table: Sample Size by Lot Size

The sample size is set by the lot size and the inspection level, and it grows far more slowly than the lot does. Doubling a lot does not double the sample. A 1,500-piece lot draws 125 pieces against 7 and 8; a 5,000-piece lot draws 200 against 10 and 11. That sub-linear relationship is deliberate, and it is the reason a large run needs process control at the bench rather than a bigger inspection at the end.

Two practical notes on the table. The code letter changes with the inspection level, so a plan quoted at level I draws fewer pieces and is a weaker plan than the same AQL at level II. And the figures above are for a single plan; a double plan, which allows a second draw before a decision, is available and is sometimes worth agreeing on large or expensive lots.

Critical, Major and Minor on a Handbag

Defects are classified before they are counted, and the classification does more work than the count. A critical defect is one that makes the piece unsafe or non-compliant: a broken needle fragment left in a panel, a restricted substance above a legal limit, or a closure that fails and spills contents. A major defect is one a customer would return the piece over. A minor defect is one a critical eye notices but a customer would keep the piece over.

The fourth card is the one that prevents the most arguments. A natural hide is not a uniform substrate, and a specification that treats every mark as a defect produces a run in which nothing passes. The agreement to make at sampling is which panel positions are visible enough to reject marks on, and how large a mark has to be before it counts.

How a Handbag Lot Is Drawn and Opened

A sample is only defensible if it is drawn at random from the whole lot. Cartons are selected first, at roughly the square root of the total carton count rounded up, and pieces are then drawn from across those cartons rather than from the top layer of the first one. Each piece is checked against the written tolerance sheet, and each finding is tagged and photographed, because a finding without a record cannot be argued about.

• Select cartons at about the square root of the carton count, rounded up.
• Draw pieces from across the cartons, not from the top layer of one.
• Check exterior, interior, hardware, edge and base in a fixed order, so nothing is skipped.
• Tag and photograph each finding before any re-work begins.
• Record readings rather than verdicts, so a later drop can be compared.

The fixed order matters more than it sounds. An inspector who moves freely around a piece will find the obvious faults and miss the ones that require a deliberate look: the lining slack at the base, the mitre gap at the corner, the feeler test at the hardware plate. A written sequence is what makes the inspection repeatable between inspectors.

Switching Rules: Normal, Tightened and Reduced

ISO 2859-1 is not a single plan but a set of three with rules for moving between them. Normal is where a programme starts. Tightened follows two rejections within five consecutive lots, and it draws a larger sample against the same accept number, which is what makes it stricter. Reduced is only available after a sustained run of accepted lots and with the agreement of the responsible party, and it should be treated with caution on a luxury line.

For a designer programme, the useful way to read the switching rules is as a signal rather than a procedure. A move to tightened is telling you that the bench has drifted, and the correct response is to find the station that drifted rather than to inspect harder. A luxury line rarely wants reduced inspection at all, because the cost of a finding at the destination exceeds the cost of the inspection by a wide margin.

What AQL 2.5 Does Not Cover

Four things fall outside a sampling plan and have to be specified separately. Compliance: restricted substances under REACH Annex XVII, California Prop 65 and CPSIA where they apply, none of which can be found by looking at a piece. Material performance: tensile strength under ISO 3376, finish adhesion under ISO 11644 and colour fastness under ISO 105, all of which require a laboratory.

Traceability is the third: hide provenance evidenced through Leather Working Group tannery audits, and CITES documentation where a listed shell is specified. Consistency over time is the fourth, and it is the one a plan cannot deliver at all: an AQL tells you whether a lot passes, not whether the next lot will look the same. That is a bench discipline, described further under luxury quality control.

Cost of a Tighter Plan and the Re-Work Reserve

A tighter plan costs inspection time and re-work time, and both are small compared with a finding at the destination. Moving a major tolerance from AQL 4.0 to AQL 2.5, holding the re-work reserve above the run quantity and adding a second inspector on the visible stations adds roughly USD 4.80 per unit on a 500-piece run. Against a return rate on a designer piece, that figure is not a close call.

The reserve is the part most often cut and the part that pays for itself. Because a rejected piece cannot always be replaced quickly, particularly on an exotic or a dyed-to-match shell, the order carries a cutting allowance above the run quantity so that a finding at inspection corrects the shipment rather than delaying it. The allowance is agreed at sampling, alongside the 6-10 working day window and the 35-50 day bulk schedule.

Glossary: Twelve Terms on an Inspection Report

AQL

Acceptance quality limit: the worst tolerable process average

Lot

The quantity submitted for inspection as one unit

Code letter

The key that converts lot size and level into sample size

Sample size

The number of pieces drawn, 50 from a 500-piece lot

Accept number

The largest defect count at which the lot passes, 3

Reject number

The smallest count at which it fails, 4

Critical defect

Unsafe or non-compliant; zero acceptance

Major defect

One a customer would return the piece over; AQL 2.5

Minor defect

One noticed but tolerated; AQL 4.0

Normal inspection

The starting state of the plan

Tightened inspection

Larger sample, same accept number, after repeat rejections

Re-work reserve

Pieces held above the run quantity to replace findings

Twelve terms is enough to read any inspection report issued on a handbag programme, and enough to write the clause that governs one. Anything beyond these is either a commercial term or a material specification, and both belong in separate sections of the order rather than inside the inspection clause.

Writing the Clause Into a Purchase Order

The programme runs through our validated partner factory Quanzhou Yingming Bags, an SGS-validated site with 7 production lines and about 200,000 pcs per month of capacity, holding export experience since 2014. Inspection is drawn at AQL 2.5 under ISO 2859-1 with a re-work reserve held; sampling takes 6-10 working days, or 12-15 on a complex build; bulk occupies 35-50 days; minimums sit at 500 pcs per style and colourway; and shipments leave FOB Xiamen with FOB, CIF and DDP terms available.

Frequently Asked Questions

People Also Ask