What Is an Acceptable Defect Rate for Sanitary Napkins?

“What is your defect rate?” lands in our inbox about twice a week. It is a fair question, and no factory can answer it with one honest number. A defect rate is not a property of a factory. It is the output of three decisions the buyer and the factory have to make together first: what counts as a defect, how many pieces get checked, and how many failures a lot can carry before it fails.

Hand the same pallet to two inspectors working from two definitions and you get two rates. One counts only visible stains on the topsheet. The other also counts wing-fold misalignment and print drift. Both are correct inside their own definition, and neither tells you whether to release the container. This is written from the production side, for the buyer who has to sign the release. Terminology first, because it changes the words on the paperwork: in the UK, Commonwealth, African and most Asian markets the product is a sanitary napkin; in the US and Canada it is usually a sanitary pad. Same product, same inspection logic.

Start with the defect definition, not the percentage

The most expensive disputes we see do not start with a bad carton. They start with a contract that says “defect rate below 1%” and nothing else. On loading day the buyer’s inspector opens a 500-piece sample and finds 9 pads with print misalignment. That is 1.8%. The factory says print drift is a minor defect, the minor limit is 4.0, and the lot passes. The buyer says the contract promised 1%. Two people are reading the same sentence and reaching opposite conclusions.

The fix is unglamorous: define the defect classes before you define the number. Once the classes exist, each class gets its own limit, and the argument above disappears, because print drift is no longer competing against a blanket percentage.

The three defect classes, and what belongs in each

Three classes cover nearly everything a napkin line produces: critical, major and minor. What matters is not the label but the list underneath it, signed by both sides before production starts.

Those AQL figures are the range we see most often in this product category. They are not a rule, and they are not a commitment from us. Your retail channel, your return costs and your own tolerance set the right value. Write the value you choose into a specification annex; do not leave it to inspection day.

Two placement decisions cause more arguments than anything else. The first is core shift: is it major at 2 mm or at 5 mm? The second is wrap seal integrity: is a seal that holds but creases a minor, or a critical because it can open in transit? Decide both in writing. On our own line, the defect that re-screening catches most often is not the core at all. It is the individual wrap seal, because a seal that looks closed on the packing table can peel open after a week inside a hot container.

The same three-step logic applies to disposable menstrual pants, which your buyers may call period pants in the UK, Australia and New Zealand, period underwear in the US and Canada, or sanitary pants across much of anglophone Africa. The classes carry over almost unchanged; the extra items to classify are leg elastic attachment and waistband weld integrity.

How AQL turns a percentage into a pass or fail

AQL (Acceptable Quality Limit) is often explained as “the defect rate you accept”. That is close enough for a conversation and wrong enough to cause trouble. AQL is one parameter of a sampling scheme defined in ISO 2859-1 and its US equivalent ANSI Z1.4. It does not describe the whole lot. It describes the risk you accept that a lot at this quality level still gets through.

Three things set the sample: lot size, inspection level, and the class AQL. General Inspection Level II is the default in most consumer-goods programmes; Level I gives a smaller sample, Level III a larger one.

The table then converts the pair (sample size, AQL) into an acceptance number (Ac) and a rejection number (Re). Read across: a lot of 120,000 pads at Level II gives a 500-piece sample, and at AQL 2.5 the standard table gives Ac 21 / Re 22. Twenty-one major-class defects or fewer and the lot is accepted; twenty-two or more and it is rejected. Check the pair against your own copy of the standard. Editions differ, and the arrow rule can move you to a different sample size.

Two things buyers routinely miss. A passed lot is not a defect-free lot: sampling accepts a defined risk, which is why the retained-sample clause in your contract matters more than the AQL value. And a single critical defect overrides the arithmetic. No Ac/Re pair rescues a lot with foreign matter inside a wrap.

Where defects come from on a napkin line

Knowing the failure sources helps you write a classification list a factory can actually work to. Four stations produce almost everything:

• Incoming material. Topsheet grammage drift, variation between SAP deliveries, fluff pulp moisture. A grammage tolerance of plus or minus 5% is a common starting point for nonwoven; tighter is achievable and costs more.
• Core formation. Uneven fluff distribution and uneven SAP spread show up as thin zones, core shift, and absorbency readings below the agreed figure.
• Converting. Wing fold placement, release-paper alignment, embossing, back-sheet adhesive position, individual wrap sealing. Seal defects are born here.
• Packing and handling. Wrong count per bag, carton crush, pallet overhang, humidity exposure in storage.

A report that collapses all four into one “total defects” line is useless for improvement. Ask for counts by station. A factory that can name the station that produced a defect can usually stop it repeating; one that cannot will ship you the same problem next quarter.

What to read before you release the shipment

Ask for these seven items in the pre-shipment report. Most established factories produce them without being asked. The list is here so you can put two suppliers side by side on the same page:

• Lot identification that traces back to production date, line and shift
• Inspection level, sample size code letter, actual sample size, and how the cartons were selected
• The limit applied per defect class, not one blended number
• Defect count per class, referencing the agreed classification list
• Photographs of each defect type found, with a reference piece or scale
• Retained samples: how many, where, for how long. Six months is a common ask.
• The measurement method behind any numeric check, including the test standard used

When a lot hits its rejection number you have three options, and the cost of each should be agreed in advance: reject and re-produce, full re-screening with an agreed cost allocation, or a documented concession with a commercial adjustment. The third is legitimate and common. What makes it dangerous is agreeing it over a messaging app with no record.

What to put in the RFQ and the contract

A specification annex of one or two pages does more for your defect rate than any inspection. Include:

• The classification list as an annex, with the two placement tolerances (core shift and seal integrity) written in millimetres or against a photo standard
• A limit per class, the inspection level, and the standard edition (ISO 2859-1 or ANSI Z1.4)
• Acceptance criteria for measurable properties: grammage, absorbency, rewet, peel strength, each with its test method
• Retained-sample quantity and period, plus who pays for re-testing if a claim is filed
• Re-inspection cost allocation when a lot is rejected, and again when it passes on second submission
• A quality-system requirement. ISO 9001 is the realistic baseline for this product category, and GMP is the other term you will meet in hygiene manufacturing. For destination-market registration the responsibility normally sits with the importer or brand owner; the factory’s job is to supply the technical file, test reports and labelling details that process needs.

The numbers that belong in this annex are numbers you can test. “Good adhesion” is not testable. “Peel strength measured by method X, minimum value Y” is. The same rule applies to the defect rate you ask a factory for: state the definition, the sampling scheme and the limit, and you will get back a number you can act on.

Common questions about sanitary napkin defect rates

Is there a normal defect rate for sanitary napkins?
Not as a single figure. A rate only means something next to a classification list, a sampling plan and an accept or reject rule. Ask two factories for “the defect rate” and you will get two numbers that cannot be compared.

Is AQL 2.5 the same as accepting 2.5% defects?
No. AQL 2.5 is a parameter of the sampling scheme, not a promise about the lot. Lots with a quality level at or better than the AQL have a high probability of acceptance, and lots worse than it have a high probability of rejection. Neither is certainty, which is why retained samples matter.

Can I ask for zero defects?
Zero is realistic for the critical class and unrealistic everywhere else. Demanding zero across all classes usually pushes cost up without improving the product, because the factory prices in full re-screening and rejects on cosmetic grounds. Put zero where safety and hygiene integrity are at stake, and set real limits for the rest.

Who pays when a lot is rejected?
Whoever your contract says pays, which is why the clause has to exist before the first order. Cover three cases: rejection and re-production, re-screening cost, and freight or line-down costs when a rejected lot delays your launch.