AQL for Bags: The Sample-Size Table, Defect Atlas & Inspection Plan Buyers Actually Use

If you’ve ever approved a beautiful sample and then opened the bulk cartons to find crooked logos, jammed zippers, or straps you wouldn’t trust with a laptop — this page is the fix. AQL is the rule that lets you reject that bulk before it ships, using a small random sample instead of opening all 5,000 pieces.

We’ve run this process for 17 years on custom bag orders. Our factory keeps a separate in-house inspection company that re-checks finished goods 100%, so the numbers and definitions below aren’t theory — they’re what our inspectors use on the floor every week. Read the table, copy the checklist into your PO, and you’ll catch problems early enough to still hit your ship date.

What AQL actually means

AQL stands for Acceptable Quality Limit. It’s the worst level of quality you’re willing to tolerate in a shipment before you reject the whole lot. Instead of inspecting every bag, you and the factory agree on a random sample size and a maximum number of defects allowed in that sample.

Those three numbers — 0 / 2.5 / 4.0 — are not percentages of “how many bags are broken.” They’re severity thresholds:

• 0 (critical): zero tolerance. One critical defect and the lot fails. Think anything that could injure the user or break the law.
• 2.5 (major): the limit for defects that hurt function or are obvious to a customer — a zipper that jams, a strap that pulls loose, a logo printed off the panel.
• 4.0 (minor): the loosest limit, for small cosmetic stuff most buyers will still accept — a stray thread, a faint scuff in a hidden spot.

Lower number = stricter. A “major 1.5” plan allows fewer major defects than “major 2.5,” which means a bigger sample and a higher chance the lot fails. That’s the trade-off: tighter quality costs you inspection time and rejection risk.

How to pick an AQL that fits your bag and your risk

There’s no single AQL for all bags, and anyone who hands you one number without asking what you’re making is guessing. Match the limit to value, risk, and how well you know the factory.

Go stricter (drop major to 1.5, or use Level III) when the bag is high value — full-grain leather, custom hardware you tooled yourself — or when it carries weight that has to hold: laptop sleeves, load-bearing straps, child-carrier panels. Also tighten for any first order with a new supplier, because you don’t yet have a track record to trust.

Stay at the standard 0 / 2.5 / 4.0 for mainstream totes, cosmetic pouches, and backpacks with familiar builds — especially repeat orders from a factory with a clean history.

The AQL table for bags

Here’s the part most “AQL for bags” articles leave out: the actual numbers. This is the standard ANSI/ASQ Z1.4 (ISO 2859-1), single sampling, normal severity, General Inspection Level II plan — the one ~98% of buyers default to. Find your order quantity in the left column, read across.

Lot size (units)	Code letter	Sample size (n)	Critical Ac/Re (0)	Major Ac/Re (2.5)	Minor Ac/Re (4.0)
91 – 150	F	20	0 / 1	1 / 2	2 / 3
151 – 280	G	32	0 / 1	2 / 3	3 / 4
281 – 500	H	50	0 / 1	3 / 4	5 / 6
501 – 1,200	J	80	0 / 1	5 / 6	7 / 8
1,201 – 3,200	K	125	0 / 1	7 / 8	10 / 11
3,201 – 10,000	L	200	0 / 1	10 / 11	14 / 15
10,001 – 35,000	M	315	0 / 1	14 / 15	21 / 22

How to read it: “Ac/Re” = accept / reject. Say you order 2,000 backpacks. That’s code letter K, so the inspector pulls 125 random units across cartons. The lot passes if there are 0 critical, ≤7 major, and ≤10 minor defects. Find an 8th major, or an 11th minor, and the lot fails — you don’t ship until it’s reworked.

A few honest notes on this table. Critical is always 0 — any critical defect rejects the lot regardless of count. The minor column is where buyers argue most; if a deadline is brutal and the issues are purely cosmetic, some buyers relax minor by one AQL step. And if your retailer (Walmart, Target, Amazon vendor programs) mandates a specific plan, use theirs — this table is the sensible default when nobody’s forcing one on you.

The Bag-Tuned Quick Plan

If your buyer doesn’t mandate Z1.4 and you just want something fast that you can train an agent or an in-house inspector on in five minutes, this simplified plan mirrors the spirit of the standard without the lookup dance. We use it ourselves when we’re auditing several SKUs in one day.

Lot size (units)	Sample size n	Acceptance (Critical / Major / Minor)
≤ 200	32	0 / 2 / 5
201 – 500	50	0 / 3 / 7
501 – 1,200	80	0 / 5 / 10
1,201 – 3,200	125	0 / 7 / 14
3,201 – 10,000	200	0 / 10 / 21

Tighten to “major 1.5” for premium leather or first-time builds by dropping the major number one step (e.g. 5 → 3). For low-risk promo totes on a tight deadline, you can loosen minor one step. Same logic, less table-hunting.

Bag Defect Atlas

A pass/fail number is useless if your inspector and the factory disagree on what counts as a defect. Half the disputes we see between buyers and suppliers aren’t about quality — they’re about definitions nobody wrote down. Put these in your spec, attach photos, and hand them to whoever inspects. (If you’re not sure how to document this formally, our guide on what goes in a tech pack covers the spec side.)

Critical defects — AQL 0

• Sharp or exposed needle, or broken hardware that can cut or injure the user.
• Toxic odor or chemical residue signalling a non-conforming finish (a REACH problem waiting to happen).
• Wrong branding or label — an IP or regulatory landmine.
• Choking hazards on kids’ items: loose snaps, small ornaments that pull off.

Major defects — function or obvious appearance

• Open or skipped stitches: visible gaps where the seam loses strength.
• Zipper failure: teeth that mis-mesh, a slider that jams under normal pull.
• Asymmetry or panel distortion: left and right differ beyond tolerance, gusset twists.
• Handle/strap pull-out: missing bartack, or thread count below spec — the defect that turns into a return three weeks after the customer buys it.
• Edge-paint cracking or peeling on strap edges after flexing.
• Hardware plating issues: rust spots, blistering, an obvious color mismatch between zipper pull and D-rings.
• Severe stains or color transfer on visible faces.
• Mis-print or mis-embroidery outside tolerance — skewed logo, broken stitch lines.

Minor defects — cosmetic, product still usable

• Loose thread ends not trimmed; light fuzz.
• Small scuffs on non-prominent surfaces.
• Light puckering where the seam allowance stacks thick.
• Logo slightly off-center but inside minor tolerance.

Measurement, tolerances & tools (write these into the spec)

• Dimensions: ±3–5 mm for totes; ±5–8 mm for backpacks and duffels.
• Logo placement: ±2–3 mm.
• Stitch density (SPI): 7–10 per inch for fabric bags; 5–7 for thick leather.
• Tools: steel ruler, soft tape, caliper for hardware, and a pull gauge for straps.

Agree on the tools and the tolerances before production starts. “It looks fine” is not a tolerance.

Physical & function tests: what an AQL count alone won’t catch

This is where most “AQL for bags” guides stop short. Counting cosmetic defects in a sample tells you nothing about whether the bag survives six months of use. A zipper can look perfect and die at 800 cycles. So our inspection pairs the visual AQL count with physical tests — the same ones our clients kept asking us to prove. These are the actual specs we run in-house:

• Colorfastness (rub): dry/wet rubbing on white cotton, 10 strokes — must reach grade 3–4 or better, no obvious color transfer.
• Static load: hang 2× the rated weight for 24 hours — handles and straps must not deform, stretch, or pull a stitch.
• Drop test: loaded bag dropped from 0.8 m, 10 times — structure stays intact, hardware doesn’t pop off.
• Seam / bartack strength: pull force on handles, straps, and key seams of 120–300 N depending on the part, with no thread break or seam opening.
• Zipper fatigue: 3,000–5,000 open/close cycles — runs smooth, teeth hold; slider pull strength ≥ 80 N.
• Hardware durability: salt-spray 24–48 h with no rust or plating loss; high/low temp soak (60 °C / −20 °C, 48 h) with no cracking, warping, or glue release.

You don’t run every test on every lot — destructive tests are usually on the golden sample plus periodic batch checks. But naming the numbers in your PO is what separates a factory that says “durable” from one that can hand you a signed test sheet. If durability is the whole reason your customers buy, that sheet is worth more than any marketing line. Our quality control process goes deeper on how we log these.

Why “the sample was great but the bulk was bad” — and how to kill it

Talk to enough brand owners and you hear the same horror story: the pre-production sample was flawless, they paid the deposit, and the bulk that arrived was a different bag. It’s the number-one reason buyers fire a factory and come looking for a new one.

Here’s the mechanic behind it. Most factories don’t have their own sample room — they send the development sample out to a separate pattern house that does meticulous, slow, one-off work. Then bulk runs through a production line that was never set up to match that pattern house. The gap between the two is your defect rate.

That’s the difference an AQL plan can’t give you on its own. AQL catches a bad bulk run after it’s made; a matched sample-to-bulk system stops the bad run from happening. You want both. (We break the full build sequence down in our bag manufacturing guide.)

How to run a bag inspection (paste this into your PO)

Before sampling

• Confirm lot size, inspection level (GII or GIII), and AQL on the PO.
• Share the Defect Atlas above, the golden sample, and measurement tolerances.
• Prep tools (ruler, caliper, pull gauge), inspection forms, and a photo board.

During inspection

• Randomly pull n units across cartons; record carton IDs — no cherry-picking from one box.
• Measure: L×W×H, strap drop, pocket fit (e.g. a real 16″ laptop in the sleeve).
• Function test: zipper run under load, handle pull, snap/buckle action, seam rub ×10.
• Appearance: symmetry, logo alignment, SPI, edge-paint, trim finishing.
• Packaging: polybag, hangtag/barcode, carton spec and drop test.

After inspection

• Count critical/major/minor against the acceptance numbers — pass or fail.
• Attach photo evidence: defects, label, carton, dimensions.
• Log a Pareto of defect types and send it to the line for the next gate.

Prevent defects upstream, so you don’t live inside inspections

Inspections don’t fix bad processes. They catch them — ideally early enough to still ship on time. The factories that pass cleanly aren’t lucky; they build quality in at three gates instead of hoping the final check saves them:

• Cutting → in-line → pre-pack: three QC gates by design. Don’t wait for final inspection to discover the build drifted.
• Golden sample with numbers attached: SPI, thread type, edge-paint coats and cure time, zipper brand/size, the exact logo file and placement.
• Process controls on the risky parts: bartack length and stitch count on straps, dry-time between edge-paint coats, a 100-cycle bend test, supplier-graded zippers (YKK/SBS) with a run-in test after stitching.
• Close the loop: hand the defect Pareto to the line supervisor and remove the top two causes before the next batch.

This is the same logic that drives production control and certification on our side — and it’s why our on-time delivery doesn’t depend on heroics at the final gate.

Documentation pack buyers expect on file

• AQL sheet: lot size, level, sample size, acceptance numbers.
• Inspection report with photos (defects, labels, cartons, dimensions).
• Materials and finish evidence for any risk claim (REACH, colorfastness, etc.).
• Corrective Action Report (CAR) if the lot fails or the same defect keeps showing up.

Keeping this pack on file is also what makes your cost per bag predictable — rework and reships are the hidden line items that wreck a budget.

Where Meyzy fits

Meyzy is a custom bag manufacturer with SMETA-audited systems and AQL-driven QC built into the PO, not bolted on at the end. We write the AQL and the Defect Atlas into your order, run three QC gates, re-check finished goods 100% through our in-house inspection company, and share time-stamped photo and video evidence so you ship on time with the receipts to prove quality.

If you’re sourcing a specific category, we run the same QC on backpacks, tote bags, travel bags, and handbags. And if you got burned by a previous factory and need a clean restart, our notes on what B2B buyers should check before a bulk order are worth ten minutes.

FAQ

What AQL should I use for backpacks?

Start at 0 critical / 2.5 major / 4.0 minor. Move major to 1.5 for load-bearing, premium, or first-time models — a strap that fails is a return, not a cosmetic note.

What does AQL 2.5 mean for a bag?

It’s the severity limit for major defects: the maximum number of major defects allowed in your random sample before the whole lot is rejected. The actual count depends on sample size — at 125 units (a 1,201–3,200 lot), AQL 2.5 lets you accept up to 7 majors and reject at 8. See the table above.

How many bags do I inspect for a 1,000-piece order?

At General Level II, a 501–1,200 lot is code letter J, so you inspect 80 randomly selected units — not 100% of the order. Pass at ≤5 major / ≤7 minor, with zero critical.

Is a lower AQL always better?

No. A lower AQL raises sample sizes and rejection rates — great for risk control, hard on timelines and cost. Match it to the bag’s value, its functional risk, and your supplier’s history.

AQL inspection vs. 100% inspection — which do I need?

AQL sampling is the industry default and is enough for most consumer bags. We add a 100% check on top through our in-house inspection company, which matters most for high-value items or when a previous batch failed and you need certainty before reordering.

How do I classify a defect that’s both cosmetic and functional?

Use the highest severity that applies. A zipper that looks fine but jams under load is major, not minor.

Can I mix inspection levels (GII on some SKUs, GIII on others)?

Not within the same PO. Lock the level for the lot, or your pass/fail trend stops meaning anything across batches.

Who pays for the inspection?

It varies by contract. Buyers often cover third-party inspection; factories with in-house QC (like ours) fold a baseline check into the order. Spell it out in the PO so it’s not a surprise.