What Does an Acceptable Handbag Defect Look Like at AQL 2.5

An acceptable defect at AQL 2.5 is not a good defect; it is a defect that falls below the reject number for its class in a defined sample. In an 80-unit sample drawn from a 501-1,200 piece lot under ISO 2859-1, up to 5 major defects and up to 7 minor defects can be accepted, while any single critical defect fails the lot outright. The distinction that matters is not severity but location: a 2 mm bead of adhesive inside a lining pocket is minor, and the same 2 mm bead on a front flap edge is major, because one is invisible in use and the other is visible in a product photograph. AQL 2.5 does not describe a bag; it describes a lot, and a passed lot still contains the defects that were counted.

The accept numbers in practice

What Does Acceptable Actually Mean at AQL 2.5?

Acceptable is a statistical statement, not an aesthetic one. AQL 2.5 says that a lot will be accepted if the number of major defects found in a randomly drawn sample is at or below the accept number for that sample size. It says nothing about the individual bags in the sample, and nothing at all about the units that were not drawn. Understanding that distinction prevents the two arguments that eat the most time at the inspection table: the buyer who believes a passed lot means defect-free goods, and the production side that believes a counted minor is somehow forgiven.

The numbers are fixed by the standard rather than chosen. Under ISO 2859-1 at general inspection level II, a lot of 501 to 1,200 pieces draws 80 units; at AQL 2.5 the accept number is 5 and the reject number is 6. Five major defects means the lot ships; six means it does not. Nothing in between exists, which is precisely the point of using a standard rather than a feeling.

Which Defects Does an Inspector Count as Major?

A major defect is one a retail customer would notice and reject on the shelf, in the photograph or in the first week of use. The test is commercial rather than technical: would this stop the sale, or generate a return? Everything below is a working classification for a private label handbag programme, and every line should be written into the specification rather than left to the inspector's judgement.

• Structural. A strap attachment seam that is unpicked, back-tacking missing at a stress point, or a handle rivet that has pulled through the reinforcement.
• Functional. A zipper that jams or separates within 500 cycles, a magnetic closure that does not hold, a slider that will not travel, or a foot or stud that is loose.
• Visible surface. Adhesive visible on a front edge or panel, a scratch longer than 10 mm on a visible face, a scuff on a coated body, or a print that is off-register by more than 2 mm.
• Colour. A unit outside Delta E 2.0 against the approved reference, or two shades inside one carton above the agreed band tolerance.
• Measurement. A body dimension outside plus or minus 3 mm on soft construction, or a strap drop outside plus or minus 5 mm.
• Hardware. Plating showing base metal, a mis-set rivet, or a coating that fails 24 hours of salt spray to ASTM B117.

Critical sits above major and is short to describe: anything that makes the bag unsafe or illegal. A needle or blade fragment left in the lining, a sharp burr on hardware, a drawcord on a child's bag long enough to be a strangulation hazard, or a restricted substance above the legal limit. One instance fails the lot, and no amount of goodwill changes that, which is why critical findings are usually handled by a full re-sort plus a documented containment rather than by negotiation.

What Does an Acceptable Minor Defect Look Like on a Real Bag?

Minor defects are the ones a careful buyer notices on close inspection and accepts anyway. They are counted, they appear in the report, and they are the reason a lot can pass while still containing imperfections. Describing them concretely is more usefully done with a threshold than with an adjective.

The pattern across all seven rows is location and permanence. A defect that a customer cannot see, or that disappears on unpacking, stays minor. The same defect on a visible face, or one that is permanent, becomes major. That is the whole classification logic, and stating it in the specification removes the argument entirely.

Where Exactly Is the Line Between the Two?

The line is not a property of the defect; it is a property of where the defect is and who will see it. Three questions settle almost every borderline case, and they should be asked in this order at the inspection table.

1. Will the customer see it before buying? A mark inside a zipped pocket will not be seen; the same mark on the front panel will be. Visibility at the point of sale moves a defect up a class.
2. Will it get worse? A light crease that relaxes is minor; a loose stitch at a strap attachment will propagate under load and is major. Progression moves a defect up a class.
3. Is it inside the stated tolerance? Where the specification names a number, the number governs, and the classification argument stops. This is why the specification matters more than the inspection.

A practical example settles it. Two bags come off the same line, each with a 2 mm bead of adhesive. On the first it sits inside the lining pocket; on the second it sits along the top edge of the front flap, which is the line a customer sees in a photograph and touches when opening the bag. Same defect, same size, two classes, and both classifications are correct. An inspector who classes them identically has misread the standard, whatever direction the error runs.

Does a Passed Lot Mean Every Bag Is Fine?

No, and this is the most consequential misunderstanding in the whole subject. A passed lot means that the sample drawn contained fewer defects than the reject number. The units not drawn were never looked at, and the units drawn are returned to the lot with their defects counted but not necessarily corrected. On a 1,200-piece order with 5 accepted major defects in an 80-unit sample, the lot can be expected to contain a proportional number elsewhere.

• The sample is a sample. Eighty units from 1,200 is a statistical draw, not a census of the cartons.
• Accepted is not repaired. Unless re-sorting is agreed, counted units go back into the lot as they were found.
• The defect rate continues into the field. A lot with 5 accepted majors typically produces a handful of returns, which is the cost of the standard rather than a failure of it.
• Tightening is available. Where two consecutive lots fail, ISO 2859-1 provides for tightened inspection, which increases the sample and lowers the accept number.

The honest framing for a brand is this: AQL 2.5 sets a defect rate that is commercially survivable, not one that is zero. Programmes that need a lower rate specify a lower AQL or a tighter class definition, and pay for the additional sorting, rather than expecting the standard to deliver more than it promises. The comparison between AQL 2.5 and AQL 4.0 sets out what each buys.

Who Decides the Classification, and When?

The buyer should propose it, the production side should agree it, and both should sign it before the deposit moves. Classification decided at the inspection table is not a classification; it is a negotiation conducted under time pressure, with the cartons already packed and the vessel booking already made. Decided at brief stage, it is a paragraph in a specification that an inspector applies without needing an opinion.

Buyer

Proposes the class definitions and the thresholds, because the buyer knows the channel and the customer

Production side

Confirms what is achievable on the line and on the material, and names the achievable tolerance

Inspector

Applies the agreed definitions to the drawn sample and counts by class, without inventing new ones

Timing

Agreed before deposit; attached to the purchase order; referenced in the inspection booking

Change control

Any change to a threshold during the run is a specification revision, dated and re-signed

Where a third party such as SGS runs the inspection, the classification is applied by their inspector against the specification supplied. That works well only where the specification exists; handed a sample and no thresholds, an inspector will apply the laboratory's default definitions, which may be stricter or looser than the commercial reality of the channel.

How to Write the Line Down So It Holds

One page, three columns, and every line a measurement rather than a description. The format below is what makes a classification enforceable months later, by somebody who was not in the room when it was agreed.

Add two administrative lines and the page is complete: the standard applied, ISO 2859-1 at AQL 2.5 and general inspection level II, and the reference artefact, the sealed golden sample with its signature date. With those, an inspector can work without a phone call, and a claim six months later can be settled by reading rather than by remembering. The same evidence discipline is described in how to act on inspection findings.

What to Do With a Lot That Passes With Defects

Treat the counted defects as data, not as a verdict. A lot that passes with 5 accepted majors is a lot with a known, characterised problem, and the useful response is to find the common cause and remove it from the next run rather than to celebrate the pass.

1. Read the defect types, not just the count. Three majors of the same type is one cause; three unrelated majors is noise.
2. Add the recurring type to the pre-bulk checklist with a threshold, so the next run is measured against it before the line starts.
3. Consider whether the units can be sorted cheaply now. Where the affected units are identifiable by carton, a targeted sort costs less than the returns.
4. Where the type recurs twice, change the specification rather than the sorting, because the process is telling you the tolerance is wrong.
5. Record the result lot by lot, so the series shows whether the programme is improving or drifting.

Programmes coordinated through our validated partner factory Quanzhou Yingming Bags, with export experience since 2014, apply that loop with inspection at AQL 2.5 under ISO 2859-1 and a written report issued to the buyer. The wider standard sits in the handbag quality control reference.

Repair, Discount or Scrap: Pricing the Three Exits

A bag carrying an accepted minor defect is still a commercial asset, and deciding what to do with it is arithmetic rather than taste. The three exits have very different economics, and the choice usually turns on how many working days remain before the vessel rather than on how bad the fault looks under a lamp. Worth noting too that most of these units never reach a customer at all if the recurring causes are dealt with upstream, as the common defects that fail final inspection catalogue sets out.

What Escapes anyway, and What It Costs Later

An AQL verdict describes a lot at a moment, not a customer over a year. What arrives in the returns box is the residue the plan never drew, and pricing it honestly is what makes tolerate versus prevent a decision rather than a feeling. The handling cost of those returns, including what a second experience does to loyalty, is set out in the returns and service reference.

• The implied rate. Five majors in an eighty-unit draw points to a true rate near six per cent, which on five hundred pieces means roughly thirty units carrying the fault into customers' hands.
• Return cost. Each one costs USD 18-35 once inbound freight, replacement and handling are counted, so thirty units is USD 540-1,050 against roughly USD 250-450 for a pre-shipment re-sort.
• Return rates. Mid-market accessories typically run three to seven per cent returns overall, and defects account for somewhere between a quarter and a third of them.
• Late failures. faults discovered after sixty to one hundred and twenty days of use have no presence at inspection at all, so they only surface through your own returns data.
• The loop. Every returning type is written onto next season's checklist with a threshold, which is how one bad lot buys a better specification.

Frequently Asked Questions

People Also Ask