Custom Bag Quality and Bag Certification Standards: The AQL, Testing and Compliance Guide for B2B Buyers
An inspection and compliance reference for brand owners, importers and retail buyers: AQL 2.5 sample sizes and acceptance numbers, critical, major and minor defect classes, the physical and chemical test matrix, and how to verify a supplier's reports.
Custom bag quality is controlled by two systems that buyers routinely confuse, and bag certification standards sit inside the second one. The first is statistical inspection under ISO 2859-1 (AQL sampling), which answers whether this lot matches the approved sample. The second is laboratory testing against named standards, which answers whether the material performs and is legal to sell. The default is AQL 2.5 with Critical 0, Major 2.5 and Minor 4.0: on a 501-1,200 unit lot the level II sample is 80 pieces, five or fewer major defects pass and six fail.
Beneath that sits the compliance layer. REACH (EC 1907/2006) governs the EU, CPSIA and California Proposition 65 govern the United States, EN 71-3 and ASTM F2923 apply to children's products, and OEKO-TEX Standard 100 or GRS covers content claims. Testing is invoiced at cost and takes 5-15 working days. Inspection and testing sit in the 2-4 per cent quality bucket of unit price, while a rejected lot costs 35-50 days of production plus freight. Indicative FOB Xiamen prices at a 500-unit MOQ run USD 1.80-4.50 for a simple tote and USD 7.00-14.00 for a laptop backpack.
- Custom bag quality is judged at AQL 2.5 under ISO 2859-1 with Critical 0, Major 2.5 and Minor 4.0, and AQL 1.5 is available for launch orders.
- At general inspection level II a 501-1,200 unit lot is judged on 80 pieces, 1,201-3,200 on 125, 3,201-10,000 on 200 and 10,001-35,000 on 315.
- ASTM D5034 grab tensile, AATCC 127 hydrostatic head, AATCC 8 crocking and ASTM D751 coating adhesion are the four physical tests behind most custom bag quality claims.
- CPSIA sets total lead at 90 ppm and phthalates at 0.1 per cent, while REACH (EC 1907/2006) triggers SVHC communication at 0.1 per cent weight by weight.
- Indicative FOB Xiamen prices at a 500-unit MOQ run USD 1.80-4.50 for a simple tote and USD 7.00-14.00 for a laptop backpack, with inspection and testing in the 2-4 per cent quality bucket.
What Custom Bag Quality and Bag Certification Standards Mean in a B2B Programme
Custom bag quality is not one thing. It is two evidence streams a buyer has to hold separately. The first is conformance: does this production lot match the approved pre-production sample, judged by statistical sampling under ISO 2859-1 (AQL sampling) or its United States equivalent ANSI ASQ Z1.4. The second is compliance: does the material and the finished article meet the regulations of the market it is sold in, judged by laboratory test reports against bag certification standards. A lot that passes inspection can still be illegal to import, and a compliant lot can still fail inspection.
Five buyer profiles read those two streams differently. Amazon and Shopify sellers optimise against return rate, so their defect list is weighted to function: zippers, strap attachment and seam strength. Retail chains and department stores optimise against tender compliance, so they ask for dated audit reports and a full test file before the first purchase order. Corporate and promotional buyers optimise against brand consistency across reorders years apart. Children's brands add CPSIA and EN 71-3 as entry conditions. Importers carry the legal duty regardless of who writes the specification.
The default inspection standard is AQL 2.5 under ISO 2859-1, with Critical 0, Major 2.5 and Minor 4.0, and AQL 1.5 is available for launch orders. Critical defects make a unit unsafe, illegal or unusable. Major defects are what a customer notices and returns. Minor defects are cosmetic and noticed only on close examination. The acronym is widely misread: AQL is the acceptable quality limit, meaning the worst process average still tolerated, not a promise that 2.5 per cent of the shipment will be defective.
Commercial framing matters because quality is priced. Inspection and testing sit in the 2-4 per cent quality control bucket of unit cost, on top of a 500-unit MOQ and 35-50 days of production. A shipment rejected at destination cannot be reworked, only replaced or discounted, and the replacement arrives one full production cycle later. That asymmetry is the reason the written specification, not the inspection, is the real quality lever. Cost mechanics sit on the custom bag MOQ and cost structure page.
Specification Options That Actually Change the Product and the Defect Profile
AQL 2.5 without a written defect list is not a specification. Two inspectors can look at the same bag and record different results, so the defect list, the tolerances and the acceptance numbers belong in the tech pack before sampling starts. Write the list in three columns: the defect, its class, and a measurable threshold. "Loose threads" is not a specification; "loose thread longer than 3 mm on an external seam" is. The same discipline applies to dimensions, where a tolerance of plus or minus 5 mm on a soft panel is normal and a structured panel is held tighter.
Material choice decides which tests you need and which defects you will see. A PU-coated polyester shell needs AATCC 127 hydrostatic head and ASTM D751 coating adhesion. A PVC shell is cheaper but raises REACH phthalate and California Proposition 65 questions that a physical test will not answer. rPET with a recycled claim needs GRS documentation and a transaction certificate per shipment. A leather handbag needs ISO 17075 for hexavalent chromium at 3 mg per kg. Every material choice moves a test into or out of your file.
Hardware and branding move risk in the same way. Stock YKK zippers and Duraflex buckles arrive with published cycle and strength data; a custom moulded pull costs USD 300-2,500 and has no data until it is tested. Plated metal needs a salt spray expectation in writing, commonly 24-48 hours to ASTM B117. Embroidery perforates the panel, which reduces water resistance exactly where the claim matters most, and screen print needs AATCC 8 crocking. Structure adds operations, and every added operation is another first-piece trial where defects originate. Component options are covered on the custom bag materials page.
Materials and Construction Decisions That Set the Defect Profile
Shell fabric is judged by force, not by denier. ASTM D5034 (grab tensile) is the reference method: a 25 mm specimen is gripped across its middle and pulled, and the result is reported in newtons for warp and weft. A 600D polyester shell is commonly specified at a 300 N minimum in both directions, and a 1680D build higher. Denier is a useful shorthand, but it does not predict seam performance, coating adhesion or abrasion, so the tech pack should carry the newton figure alongside the denier and the coating weight.
Water resistance is three different tests, and buyers routinely ask for the wrong one. AATCC 127 hydrostatic head measures the water column a fabric holds before it leaks, with shower resistance commonly specified in the 300-1,000 mm range and a higher claim at 1,500 mm or more. AATCC 22 spray rates surface wetting from 0 to 100, with 80 or above the usual floor. AATCC 35 rain test measures penetration onto blotter paper under a spray rig, with the mass gain agreed in grams. Coating adhesion to ASTM D751 is what fails first in the field.
Colour, laundering and trim complete the profile. AATCC 8 covers dry and wet rubbing, with dry grade 4 and wet grade 3-4 the usual floor for dark shades, and AATCC 165 covers transfer onto carpet where a printed or coated face can lift in store. ISO 6330 covers dimensional change where a wash claim is made, commonly a 3 per cent shrinkage limit, and ISO 12945 covers pilling. Seam construction carries stitch density at 8-10 SPI on light fabrics and 6-8 on heavy ones, with bar-tacks at every strap and handle attachment point.
Defect Classification in Custom Bag Quality: Critical, Major and Minor
Every defect found in an inspection is assigned to one of three classes, and the class decides whether the lot passes. A critical defect makes the unit unsafe, illegal or unusable. A major defect is one a customer would notice and return, or one that shortens service life materially. A minor defect is cosmetic and would not cause a return on its own. Once the list is written down, classification stops being a matter of opinion, which is why it belongs in the tech pack rather than in an email sent the week before inspection.
Critical findings in bags are narrower than most buyers expect, and chemistry belongs in this class. Safety items are sharp edges or points on metal trim, a small part that detaches at the specified pull force on a children's product, a strap or handle that separates under rated load, a zipper that fails to close, and any functional failure that makes the product unusable. A laboratory result above a regulatory limit is also critical, even though it is found in a report rather than in the sample, because the consequence is a recall, not a return.
Major defects are the ones that drive returns: colour bleeding below the agreed grey scale grade, skipped or broken stitches on a load-bearing seam, dimensions outside tolerance, a missing bar-tack, print registration outside the agreed allowance, slider action that catches, or panel-to-panel shade variation beyond the agreed Delta E of 1.5. Minor defects are loose threads over 3 mm, removable soiling, light creasing and label skew. Inspectors count defects rather than defective units, so one bag with three minor defects contributes three to the count against the 4.0 allowance.
AQL Sampling Under ISO 2859-1 and ANSI ASQ Z1.4: Sample Sizes and Ac/Re Numbers
The sampling plan converts a lot size into a sample size and a pair of numbers. Under ISO 2859-1 (AQL sampling) and ANSI ASQ Z1.4, the lot size is looked up against a general inspection level to give a code letter, and the code letter gives the sample size. General inspection level II is the default and is what should appear on your inspection report; level I is used where discrimination matters less and level III where it matters more. Samples must be drawn at random across cartons, never from the top layer of the first pallet.
At level II the sample sizes in this category are 80 pieces for a 501-1,200 unit lot, 125 for 1,201-3,200, 200 for 3,201-10,000 and 315 for 10,001-35,000. Each class carries its own acceptance and rejection numbers, written Ac and Re. At AQL 2.5 for majors, the sample of 80 accepts on 5 and rejects on 6; the sample of 125 accepts on 7 and rejects on 8; 200 accepts on 10 and rejects on 11; 315 accepts on 14 and rejects on 15. Critical defects are treated as zero acceptance: one critical finding fails the lot regardless of how clean the rest is.
Choosing between AQL 1.5 and AQL 4.0 is a commercial decision rather than a technical one. AQL 1.5 tightens the sample of 80 to accept 3 and reject 4, which is worth it on a launch order where the first reviews decide the listing, and on any programme where defect cost lands on your brand. AQL 4.0 loosens the same sample to accept 7 and reject 8, which suits low-value promotional goods where a rework cycle costs more than the defects. Tightening raises rejection risk, so switching rules matter: two of five consecutive lots rejected moves normal inspection to tightened.
Bag Certification Standards: The Physical Test Matrix
The physical test matrix is the part of bag certification standards that most buyers discover too late. Tests are booked against a material, not against an idea, and they take 5-15 working days at a laboratory. That means the matrix has to be agreed during sampling rather than after the pre-production sample is signed, or the schedule slips by two weeks. Tests divide into mechanical performance, water resistance, colour fastness, component endurance and transit. The table below sets out the reference methods used across the category.
Two rules make a test report usable. First, agree the pass criterion before the test, because a laboratory reports a value and not a verdict: a hydrostatic head result of 820 mm is neither a pass nor a fail until the specification states 1,000 mm. Second, test the material family and colour group that will actually ship, because a change in coating weight or dye formulation invalidates the result. A report that names 600D polyester says nothing about the 900D shell substituted in bulk.
Report validity is a scope question rather than a date question. A physical test on a fabric platform is commonly accepted for one to two years, but it expires immediately on a material, coating or colour change, and every new component needs its own report. Metal components need a report per plating type rather than per part number. The report itself must carry the laboratory name, an accreditation mark, the standard including its edition year, the sample description with article codes, the test date and a report number you can verify with the issuer.
Specification tables
| Lot size (pieces) | Sample size | Critical Ac/Re | AQL 1.5 Ac/Re | AQL 2.5 Ac/Re (Major) | AQL 4.0 Ac/Re (Minor) |
|---|---|---|---|---|---|
| 501-1,200 | 80 | 0 / 1 | 3 / 4 | 5 / 6 | 7 / 8 |
| 1,201-3,200 | 125 | 0 / 1 | 5 / 6 | 7 / 8 | 10 / 11 |
| 3,201-10,000 | 200 | 0 / 1 | 6 / 7 | 10 / 11 | 14 / 15 |
| 10,001-35,000 | 315 | 0 / 1 | 8 / 9 | 14 / 15 | 21 / 22 |
| 35,001-150,000 | 500 | 0 / 1 | 10 / 11 | 21 / 22 | Not used at this level |
| Reading the result | Defects found at or below Ac pass the lot | One critical fails the lot | Tighter; for launch orders | Default on most programmes | Looser; for low-value goods |
| Class | Definition | AQL applied | Typical bag examples | Disposition |
|---|---|---|---|---|
| Critical | Unsafe, illegal or unusable; zero tolerance | 0 (Ac 0, Re 1) | Sharp edge or point on metal trim, small part detaching under pull force, strap or handle separation under load, zipper that will not close, lead or phthalate result above the legal limit | Reject the lot; containment and root cause before remake |
| Major | A customer would notice and return it, or service life is materially shortened | 2.5 (launch orders 1.5) | Colour bleeding below grey scale grade, skipped or broken stitch on a load-bearing seam, dimension outside tolerance, missing bar-tack, print registration outside allowance, slider action that catches, shade variation beyond Delta E 1.5 | Count against Ac/Re; rework and re-inspect, or accept with a price adjustment |
| Minor | Cosmetic; would not cause a return on its own | 4.0 | Loose thread over 3 mm, removable soiling, light creasing, label skew, slight carton scuffing | Count against Ac/Re; usually accepted with a note on the report |
| Counting rule | Inspectors count defects, not defective units | Each class judged separately | One bag with three minor defects contributes three to the minor count | The lot passes only when all three classes pass |
| Test | Standard | What it measures | Pass criterion to specify | What failure looks like |
|---|---|---|---|---|
| Grab tensile, shell fabric | ASTM D5034 | Force to break a 25 mm grabbed specimen, warp and weft | 300 N minimum in both directions for a 600D shell | Panel tears at the seam or under rated load |
| Coating adhesion | ASTM D751 | Bond strength between coating and base fabric | No delamination at the stated load and no cracking after flex | Coating peels or flakes after folding |
| Tongue tear, coated fabric | ASTM D751 | Force required to propagate a cut | Minimum newtons set by denier | A small cut runs across a panel |
| Hydrostatic head | AATCC 127 | Water column held before leakage | 300-1,000 mm for shower resistance; 1,500 mm or more for a higher claim | Leaks below the stated column |
| Spray test | AATCC 22 | Surface wetting after a fixed spray, rated 0-100 | 80 or above after any agreed wash cycles | Water soaks into the face fabric |
| Rain test | AATCC 35 | Water penetration onto blotter paper under a spray rig | Blotter mass gain at or below the agreed value, commonly 1 g | Water reaches the interior |
| Colour fastness to rubbing | AATCC 8 | Dry and wet rub transfer, rated 1-5 | Dry grade 4, wet grade 3-4 minimum for dark shades | Colour transfers to hands or clothing |
| Colour fastness to carpet | AATCC 165 | Transfer from a printed or coated face onto carpet | Grade 4 minimum where a hard floor claim is made | Print lifts onto surfaces in store |
| Domestic laundering | ISO 6330 | Dimensional change and appearance after washing | Shrinkage 3 per cent or less, with no seam damage | Bag distorts or seams open after washing |
| Pilling and abrasion | ISO 12945 | Surface fuzzing and pilling after abrasion, rated 1-5 | Grade 3-4 or better after the agreed cycles | Face fabric looks worn early |
| Zipper reciprocating | QB/T and YKK internal specification | Open and close cycles before failure | 500-1,000 cycles with no slider or tooth failure | Slider jams or the chain separates |
| Buckle and clip fatigue | Brand protocol, commonly Duraflex | Open and close cycles, then tensile after cycling | 500-1,000 cycles with no crack and no loss of strength | Clip cracks at the webbing slot |
| Webbing tensile | ASTM D5035 or ISO 13934-1 | Break force of webbing in newtons | 900 N minimum for 25 mm polyester webbing | Webbing fails at rated load |
| Salt spray corrosion | ASTM B117 | Corrosion resistance of plated metal parts | 24-48 hours with no base metal corrosion, or as agreed | Plating blisters or rusts in transit |
| Drop test, packed product | Brand protocol, often referencing MIL-STD-810 | Drop of a loaded unit from a stated height | Agreed height, commonly 1 m loaded to rated capacity | Handle or strap attachment fails |
| Transit simulation | ISTA 3A | Vibration, drop and conditioning for parcel shipments | Carton and product pass with no functional damage | Cartons crush and units arrive deformed |
| Regulation or standard | Market | What it restricts | Key limit | Evidence you need |
|---|---|---|---|---|
| REACH (EC 1907/2006) SVHC | EU | Substances of very high concern on the candidate list | 0.1 per cent weight by weight, with communication and SCIP duties | Supplier declaration plus a laboratory screen per material family |
| REACH Annex XVII | EU | Restricted substances by application, including azo dyes, nickel, PAHs, cadmium and phthalates | Aromatic amines from azo dyes at 30 mg per kg; nickel release to EN 1811 | Test report naming the restriction entry and the material |
| California Proposition 65 | United States, California | Lead, cadmium and listed phthalates, enforced by private litigation | No safe-harbour number in the regulation; brands commonly set 90 ppm for lead | Declaration, laboratory screen and a warning assessment |
| CPSIA | United States | Total lead and phthalates in children's products | Total lead 90 ppm; phthalates 0.1 per cent | Children's Product Certificate with third-party laboratory results |
| EN 71-3 | EU, toys and children's articles | Migration of 19 elements from accessible materials | Category-specific migration limits in mg per kg | EN 71-3 test report per material and colour |
| OEKO-TEX Standard 100 | Global, fabric level | Harmful substances in textile articles by product class | Class I for babies, Class II for skin contact | Certificate number, scope and validity date |
| ISO 17075 | Global, leather | Hexavalent chromium in leather | 3 mg per kg | Test report per leather article and colour |
| EN 1811 | EU | Nickel release from metal parts in prolonged skin contact | 0.2 micrograms per square centimetre per week | Test report per metal component and plating type |
| ASTM F2923 | United States | Children's jewellery and metal trim safety | Total lead 90 ppm and cadmium limits for accessible metal | Test report per metal component |
| PFAS and C6 versus C8 | United States states and EU | Per- and polyfluoroalkyl substances in water-repellent finishes | No single limit; several states require reporting and EU work is in progress | Finish declaration stating C6 or C8 chemistry, or PFAS-free |
| FDA 21 CFR | United States | Food contact liners for cooler bags | Compliance with the relevant food contact chapter | Supplier statement plus a test report where a food contact claim is made |
| Scheme | What it covers | Evidence produced | Typical validity | What to check |
|---|---|---|---|---|
| ISO 9001 | Quality management system | Certificate with site address and scope | 3 years with annual surveillance audits | Scope covers bag design and manufacture, not only trade |
| ISO 14001 | Environmental management system | Certificate with scope | 3 years with annual surveillance | Chemical management and waste treatment are inside the scope |
| BSCI / amfori BSCI | Social compliance and labour standards | Audit report with rating and corrective action plan | Commonly 1-2 years | Report is dated, names the auditor and shows open findings |
| Sedex SMETA | Labour, health and safety, environment, business ethics | SMETA report and Sedex reference number | Commonly 1-2 years | Two-pillar or four-pillar scope is stated on the report |
| GRS (Global Recycled Standard) | Recycled content, chain of custody, social and environmental criteria | Scope certificate plus a transaction certificate per shipment | 1 year | Transaction certificate quantity matches the invoice; 20 per cent minimum content, 50 per cent for the logo |
| FSC | Paper and wood-based packaging | Chain-of-custody certificate | 5 years with annual audits | Claim reads FSC 100 per cent, FSC Mix or FSC Recycled |
| WRAP | Social compliance for sewn goods facilities | Certificate naming the facility | 1 year, with unannounced audits | Certificate level and expiry date |
| Checkpoint | When it happens | What is checked | Sample or coverage | Record produced |
|---|---|---|---|---|
| Goods-in material inspection | On arrival of fabric, webbing and hardware | Shade, coating weight, width, plating, zipper action | First roll of each lot plus a stated percentage | Material inspection record with supplier lot numbers |
| First article inspection (FAI) | Before the run starts | Dimensions, construction sequence, print placement against the tech pack | 1-3 pieces built from bulk materials | FAI report and written sign-off |
| In-process QC (IPQC) | Hourly or per operation in cutting, sewing and packing | Marker errors, skipped stitches, missed bar-tacks, wrong trim | Patrol inspection, commonly 5-20 pieces per hour per line | Patrol record with defect counts by operation |
| Pre-final internal inspection | When packing starts, usually at 80 per cent packed | The same defect list as the final inspection, to catch trends early | Internal sample at the same level II size | Internal report used to trigger rework |
| Final random inspection | At 100 per cent packed, before release | AQL 2.5 under ISO 2859-1, quantity, packing and labelling | 80 pieces at 501-1,200 units; see the sampling table | Final inspection report with photographs |
| Third-party inspection | Booked 3-5 days before the required date | Independent AQL, quantity and packing verification | Same level II sample, drawn by the inspector | Report from SGS, Intertek, BV or TÜV with report number |
| Loading supervision | At container loading | Carton count, condition, stacking and seal number | One full container | Loading report with photographs and seal number |
Chemical and Regulatory Compliance: REACH, Proposition 65, CPSIA and EN 71-3
EU programmes are governed by REACH (EC 1907/2006), and two parts matter. The candidate list of substances of very high concern triggers a communication duty at 0.1 per cent weight by weight in an article, plus SCIP notification. Annex XVII is the restriction list and is where the numeric limits live: aromatic amines released from azo dyes at 30 mg per kg, nickel release measured to EN 1811, and cadmium, PAHs and phthalates restricted by application. OEKO-TEX Standard 100 is not law, but it is the practical fabric-level screen most EU retailers accept in place of a substance-by-substance file.
US programmes are governed by CPSIA and California Proposition 65, which work differently from each other. CPSIA sets a total lead limit of 90 ppm in substrates and surface coatings for children's products and a phthalate limit of 0.1 per cent, and requires a Children's Product Certificate supported by third-party testing. Proposition 65 contains no pass number in the regulation itself, because it is enforced through private litigation, so brands set internal limits, commonly 90 ppm for lead, and require a warning assessment where a listed substance is present above the agreed threshold.
Three chemical topics decide most bag tenders. Azo dyes are restricted under REACH Annex XVII at 30 mg per kg of listed aromatic amines and are tested per colour, so a five-colourway programme needs five screens. Leather needs ISO 17075 for hexavalent chromium at 3 mg per kg, because a tanning deviation can appear long after the hide was bought. PFAS is the fastest-moving item: C8 chemistry based on PFOA and PFOS is largely restricted, C6 short-chain finishes remain widely used for durable water repellency, and several US states now require reporting.
Metal parts carry their own rules. EN 1811 sets nickel release at 0.2 micrograms per square centimetre per week for parts in prolonged skin contact, which covers buckles, rings and sliders. ASTM F2923 (children's jewellery) is applied by many brands to metal trim on children's bags, with total lead at 90 ppm and cadmium limits for accessible components. Testing is invoiced at cost, typically in the low hundreds of US dollars per material family, and the evidence set is a test report plus a material declaration plus a declaration of conformity.
Social Compliance and Management Systems: ISO 9001, BSCI, SMETA, GRS, FSC, WRAP
Management-system certificates and social audits are separate evidence, and retailers now ask for both. ISO 9001 covers the quality management system and ISO 14001 the environmental system; both are issued to a named site with a stated scope, run for three years with annual surveillance audits, and are worthless if the scope excludes what you are buying. BSCI / amfori BSCI and Sedex SMETA cover labour and ethics. SMETA is reported as two-pillar or four-pillar, and four-pillar adds environment and business ethics to labour standards and health and safety.
Content claims need chain-of-custody schemes. GRS (Global Recycled Standard) certification starts at 20 per cent recycled content, while the consumer-facing logo claim needs 50 per cent or more; the practical control is the transaction certificate, issued per shipment, whose quantity must reconcile with the commercial invoice. FSC covers paper and wood-based packaging, with the claim stated as FSC 100 per cent, FSC Mix or FSC Recycled. WRAP certifies sewn-goods facilities on social compliance. Each is a dated document with a number, and each can be checked with the issuing body.
Ask for dated documents from named bodies, not for logos. A certificate should be requested with its scope statement and expiry date, and an audit report with its findings and corrective action plan rather than its cover page. Our SGS-verified production base runs 4,950 m2 with 137 people across 7 lines and 149 machines at 200,000 units a month, and audit reports and scope certificates are provided as dated documents on request. Where a claim appears on the product, the certificate behind it has to name that product.
Inspection Execution: FAI, IPQC, Final AQL and Third-Party Inspection
Inspection runs at four points, and the last one is the least useful on its own. Material goods-in checks fabric shade, coating weight, width and plating before anything is cut, where a defect costs a roll rather than a finished unit. First article inspection checks the first pieces built from bulk materials against the tech pack and the pre-production sample. In-process QC patrols cutting, sewing and packing and records defects by operation, which is where trends appear. Final random inspection to AQL 2.5 happens at 100 per cent packed.
Third-party inspection is bought when the relationship is new, the order is large, or the destination retailer requires it. SGS, Intertek, BV and TÜV all run consumer-goods inspections in China on a man-day basis, booked three to five days ahead, with the report issued within 24 hours. Market-indicative cost is USD 250-400 per inspector day, and one inspector day realistically covers 300-500 pieces plus the packing and quantity checks. The inspector draws the sample from randomly selected cartons, which is the point of paying for independence.
Read seven fields on every report: the report number, the inspection date, the style and purchase order, the lot and sample size, the AQL and inspection level used, the defect list with counts by class and photographs, and the stated result tied to an acceptance decision. A report that says pass without stating the sample size and the AQL has told you nothing. If a lot fails, the options are rework with re-inspection at two to three days, a price adjustment, or rejection and remake, which costs 35-50 days of production plus freight.
MOQ, Pricing and Lead Time for a Tested and Certified Bag Programme
Quality does not change the commercial structure, but it does add a calendar and a cost line. The custom bag MOQ is 500 units per style and colour, negotiable to 300 on repeats, and a complete tech pack returns a quotation in 24-48 hours. Sampling runs 6-10 working days, or 12-15 for complex builds, and mass production runs 35-50 days after sample approval, reaching 30 days in the off-season. Testing is booked inside that window, not after it, or the shipment waits on a laboratory rather than on a vessel.
Indicative FOB Xiamen prices at a 500-unit MOQ are USD 1.80-4.50 for a simple tote, USD 4.50-9.00 for a basic backpack, USD 7.00-14.00 for a laptop backpack, USD 5.50-12.00 for an insulated cooler bag and USD 12.00-35.00 for a leather handbag; those figures are indicative only. Inspection and testing sit in the 2-4 per cent quality control bucket of unit price, which is the cheapest insurance in the programme. Removing it does not remove the cost, it moves the cost from the production floor to your warehouse and your return rate.
Payment is T/T 30/70, with L/C at sight for container orders, and the default term is FOB Xiamen under Incoterms 2020, with EXW, CIF and DDP available. Transit protection is specified separately: ISTA 3A for parcel shipments, and IATA rules apply where samples or launch stock move by air. The mechanics of the wider supply chain sit on the source bags from China page, and the production sequence is set out on the bag manufacturing process page.
Step-by-Step: Building a Custom Bag Quality Programme in Eight Steps
Step 1: write the defect list and fix the AQL. Before sampling, produce a three-column list of every defect you care about, its class, and its measurable threshold, then state the inspection level and AQL. Use general inspection level II and AQL 2.5 with Critical 0, Major 2.5 and Minor 4.0 as the default, and AQL 1.5 for a launch order. This document, not the sample, is what the final inspection is judged against.
Step 2: lock the reference sample and the bill of materials. The pre-production sample must be built from bulk materials on bulk tooling and signed as the reference. Alongside it, lock a bill of materials that names every component by article code and supplier: fabric, coating, lining, webbing, zipper, buckles, thread and print method. A locked BOM is what makes a test report meaningful, because it lets you check that the report covers what actually shipped.
Step 3: book the test matrix against the material family. During sampling, agree which tests apply and what the pass criteria are, then book them against the actual fabric, colour group and hardware. Physical tests to ASTM D5034, ASTM D751, AATCC 127 and AATCC 8, chemical screens to REACH (EC 1907/2006), CPSIA or EN 71-3 as the market requires. Allow 5-15 working days and budget the cost per material family.
Step 4: verify certificates and run goods-in inspection. When fabric, webbing and hardware arrive, check shade against the signed swatch under D65, verify coating weight and width, check plating and zipper action, and record the supplier lot numbers. Where a GRS or OEKO-TEX claim is made, match the transaction certificate quantity to the invoice. A defect found here costs a roll; the same defect at final inspection costs a finished unit.
Step 5: run first article inspection before the line starts. Pull one to three pieces built from bulk materials and measure them against the tech pack and the signed sample: dimensions, stitch density, bar-tack placement, print placement and registration, hardware function. Sign the FAI report before cutting continues. A construction error found at piece one costs an hour; the same error found at packing costs the run.
Step 6: patrol in-process QC and record defects by operation. Inspection during cutting, sewing and packing, recorded by operation rather than by unit, is where trends become visible while they can still be corrected. A rise in skipped stitches on one machine or misregistration on one print position is a maintenance or setup problem, not a defect problem, and it is only visible in patrol records. Count defects per operation and act on the top two.
Step 7: run final AQL inspection at 100 per cent packed. Draw the level II sample at random across cartons, inspect against the written defect list, and record counts by class. On a 501-1,200 unit lot the sample is 80 pieces: 5 or fewer majors passes, 6 or more fails, and any critical fails the lot. Where the order is large, new or retailer-mandated, book SGS, Intertek, BV or TÜV three to five days ahead and use their sample.
Step 8: assemble the document pack and release with batch traceability. Before the balance payment and shipment, collect the test reports, the declaration of conformity, the material declarations, the BOM with supplier list, the inspection report with photographs, and the batch or lot numbers that link finished cartons back to fabric rolls and hardware lots. That pack is the defence when a claim arrives in month eight, and it is what makes a reorder repeatable.
How to Choose the Right Programme Partner
Choose on written evidence rather than on presentation. Six things should be available on request: a defect list with classes and thresholds, a named AQL and inspection level, dated laboratory test reports covering the actual BOM, dated audit reports with findings, batch traceability from carton to material lot, and a retained reference sample with shade card and tooling record. A partner who can produce those six has a quality system; a partner who produces a folder of logos has a marketing system.
The comparison table below sets out the three profiles you will meet when sourcing custom bag programmes. The differences that matter are not the unit price, which is visible, but the inspection standard, the documentation, the traceability and what happens when a lot fails. Those four sit outside the price and decide where the money goes in year two, which is where margin in this category actually comes from.
Four red flags account for most bad outcomes, and all four are document problems rather than product problems. A fabricated or edited report, an expired report, a report that does not match the goods, and a report whose scope does not cover the model. Each is cheap to catch at the file stage and expensive to catch at destination, where a compliance failure costs the goods, the freight, the listing and often the retail relationship. The verification routine is below.
Verification takes about an hour per purchase order and should be a written habit. Check the report number against the issuing laboratory, check the date against the material change history, check the article codes and colour descriptions against the BOM, and check that the certificate scope names the product and the site. Reorder discipline is the last test: ask how colour is controlled across orders a year apart and whether the signed sample is retained. Our production team works to that standard on every programme.
Reference tables
| Item | Market-indicative range | Coverage | Turnaround | Note |
|---|---|---|---|---|
| Inspection man-day | USD 250-400 per inspector day | 300-500 pieces plus packing and quantity checks | Report within 24 hours | Charged by SGS, Intertek, BV and TÜV on the same basis |
| Half-day inspection | Roughly 60 per cent of a man-day | Small lots below 500 pieces | Report within 24 hours | Sample size still follows ISO 2859-1 |
| Loading supervision | Included, or a few tens of USD extra | One container | Same day | Worth adding on the first shipment |
| Chemical laboratory screen | Low hundreds of USD per material family | One material family and one colour group | 5-10 working days | REACH SVHC, CPSIA lead or EN 71-3 |
| Physical laboratory test | Low hundreds of USD per test group | Grab tensile, coating adhesion, colour fastness | 5-10 working days | Agree the pass criterion before booking |
| Full compliance package | Around USD 1,000 or more per style | REACH, CPSIA, EN 71-3 and physical tests combined | 6-10 working days | Book during sampling, not after it |
| Re-inspection after rework | Same man-day rate, charged again | Full or partial lot | 24 hours plus rework time | The reason a failed lot is expensive |
| Document | What it proves | Who issues it | When you need it | Red flag |
|---|---|---|---|---|
| Laboratory test report | The material or product meets the named standard | Accredited laboratory | Before the first shipment and on any material change | No report number, no laboratory name, no date |
| Declaration of conformity (DoC) | Legal statement that the product meets the listed regulations | Brand or importer, supported by test data | Per shipment or per production batch | Cites regulations with no supporting reports |
| Material declaration | Substance content by component against a restricted substance list | Material supplier | Before sampling | Generic statement covering all colours |
| BOM with supplier list | Every component, article code and source | Production team | At pre-production sample sign-off | Article codes do not match the goods |
| Batch or lot number record | Traceability from finished carton back to fabric roll and hardware lot | Production team | Per shipment, retained after delivery | No way to link a carton to a material lot |
| Certificate with scope | System or content claim | Certification body | On request and at renewal | Certificate expired or scope excludes the product |
| Inspection report with photographs | What was found and how the lot was judged | Internal QC or third party | Before the balance payment | Pass result with no sample size or AQL stated |
| Product type | Indicative FOB Xiamen band | Recommended AQL | Test package | QC share of unit price |
|---|---|---|---|---|
| Simple tote | USD 1.80-4.50 | 2.5, or 4.0 for promotional runs | ASTM D5034 grab, AATCC 8 crocking, REACH screen | 2-4 per cent |
| Basic backpack | USD 4.50-9.00 | 2.5 | Grab tensile, AATCC 127, zipper cycles, drop test | 2-4 per cent |
| Laptop backpack | USD 7.00-14.00 | 2.5, or 1.5 on a launch order | Full physical matrix plus ISTA 3A transit | 2-4 per cent |
| Insulated cooler bag | USD 5.50-12.00 | 2.5 | ASTM D751 coating, FDA 21 CFR liner statement | 2-4 per cent |
| Leather handbag | USD 12.00-35.00 | 2.5, or 1.5 on a launch order | ISO 17075 chromium VI, EN 1811 nickel, AATCC 8 | 2-4 per cent |
| Children's bag or pencil case | Within the tote and backpack bands | 1.5 | CPSIA lead 90 ppm, phthalates 0.1 per cent, EN 71-3, ASTM F2923 | 2-4 per cent plus certification cost |
| Basis | 500-unit MOQ, indicative only | General inspection level II | Invoiced at cost per material family | Inspection plus testing plus rework allowance |
| Criterion | Low-cost intermediary | Mid-market workshop | Verified programme partner |
|---|---|---|---|
| Written defect list | None, or a generic list | Internal list, shared on request | Three-column list with classes and thresholds |
| Inspection standard | Visual check or none in writing | Internal, undefined AQL | AQL 2.5 under ISO 2859-1, AQL 1.5 available |
| Test reports | Logos and old PDFs | Some reports, unclear scope | Dated reports covering the actual BOM and colour group |
| Audit documentation | Certificate image | One certificate, no report | Dated audit report with findings and corrective actions |
| Traceability | None beyond the carton | Production records only | Batch numbers linking carton to fabric roll and hardware lot |
| Chemical compliance | Claims compliance verbally | REACH statement without a report | REACH, CPSIA, EN 71-3 reports per material family |
| Behaviour when a lot fails | Disputes the defect list | Reworks without a record | Rework, re-inspection and a written disposition |
| Reorder capability | New price and new sample each time | Approximate colour match | Retained reference sample, shade card and tooling record |
| Red flag | How it looks | How to verify | Cost of missing it |
|---|---|---|---|
| Fabricated or edited report | A PDF with altered dates, cropped headers or no accreditation mark | Ask the issuing laboratory to confirm the report number and the sample description | Recall, listing removal and fines at destination |
| Expired report | Valid format, but the test date is more than the accepted window old | Check the test date against the material change history, not the shipment date | Rejected entry at a retailer or port |
| Report does not match the goods | Report names 600D polyester while bulk is 900D, or names a different colour group | Compare article codes and colour descriptions against the locked BOM | Test expense with no legal value |
| Scope does not cover the model | Certificate covers another site, another product family, or trade only | Read the scope statement and the site address on the certificate | Audit rejection and a failed tender |
Frequently Asked Questions
What does AQL 2.5 mean for custom bag quality?
AQL 2.5 is the acceptable quality limit applied to major defects in a final random inspection under ISO 2859-1 (AQL sampling). It is the worst process average still tolerated, not a promise of 2.5 per cent defects. On a 501-1,200 unit lot at general inspection level II the sample is 80 pieces: 5 or fewer major defects pass, 6 or more fail. Critical defects are zero acceptance and minor defects are judged at AQL 4.0.
Should I use AQL 1.5, 2.5 or 4.0 for my bag order?
Use AQL 2.5 as the default, AQL 1.5 for a launch order or a children's product, and AQL 4.0 only for low-value promotional goods. At a sample of 80 pieces, AQL 1.5 accepts 3 and rejects 4, AQL 2.5 accepts 5 and rejects 6, and AQL 4.0 accepts 7 and rejects 8. Tightening reduces defects but raises rejection risk, which costs rework and re-inspection time.
What is the difference between inspection and testing in bag certification standards?
Inspection answers whether the lot matches the approved sample, using statistical sampling under ISO 2859-1. Testing answers whether the material performs and is legal, using laboratory methods such as ASTM D5034 grab tensile, AATCC 127 hydrostatic head, AATCC 8 crocking and REACH (EC 1907/2006) screens. A lot can pass inspection and still fail compliance, so you need both evidence streams.
How many pieces are inspected on a 500-unit custom bag order?
A 500-unit order normally falls into the 501-1,200 lot band, so the general inspection level II sample is 80 pieces. The sample rises to 125 pieces at 1,201-3,200 units, 200 at 3,201-10,000 and 315 at 10,001-35,000. Samples must be drawn at random across cartons. One inspector day realistically covers 300-500 pieces plus the packing and quantity checks.
What counts as a critical defect in a bag inspection?
A critical defect makes the unit unsafe, illegal or unusable, and one critical finding fails the whole lot. Examples are a sharp edge or point on metal trim, a small part detaching at the specified pull force on a children's product, a strap or handle separating under rated load, a zipper that will not close, and any laboratory result above a regulatory limit such as CPSIA lead at 90 ppm.
What is the difference between a major and a minor defect?
A major defect is one a customer would notice and return, or one that materially shortens service life: colour bleeding below the agreed grey scale grade, skipped stitches on a load-bearing seam, dimensions outside tolerance or a missing bar-tack. A minor defect is cosmetic and would not cause a return alone, such as a loose thread over 3 mm, removable soiling or label skew. Majors are judged at AQL 2.5 and minors at AQL 4.0.
Which certifications do custom bags need for the EU and US markets?
EU programmes need REACH (EC 1907/2006), including SVHC at 0.1 per cent weight by weight and Annex XVII restrictions such as azo dyes at 30 mg per kg, and most retailers also accept OEKO-TEX Standard 100. US programmes need CPSIA at 90 ppm total lead and 0.1 per cent phthalates plus California Proposition 65. Children's products add EN 71-3 and ASTM F2923, and recycled claims need GRS.
How much does laboratory testing cost for a custom bag programme?
Testing is invoiced at cost rather than marked up. A single chemical screen or physical test group on one material family and colour group is typically in the low hundreds of US dollars, and a full compliance package covering REACH, CPSIA, EN 71-3 and the physical matrix is around USD 1,000 or more per style. Turnaround is 5-15 working days, so book during sampling.
How long are test reports and certificates valid?
A physical test on a fabric platform is commonly accepted for one to two years, but it expires immediately on a material, coating or colour change, and every new component needs its own report. ISO 9001 and ISO 14001 certificates run three years with annual surveillance. BSCI and Sedex SMETA audit reports are commonly accepted for one to two years, and GRS scope certificates run for one year.
How do I check that a test report is genuine?
Check five things: the report number against the issuing laboratory, the test date against the material change history, the article codes and colour descriptions against the locked bill of materials, the accreditation mark, and the sample description. A report with no laboratory name, no report number or no date is not evidence. Scope matters as much as date, because a certificate can be valid and still exclude your product.
What is the minimum order quantity for a tested and certified bag programme?
The custom bag MOQ is 500 units per style and colour, negotiable to 300 on repeat programmes where patterns, dies and colour references already exist. The number is batch economics rather than policy: mill minimums, dye lots, line changeover and screen amortisation all cost a fixed amount. Testing cost per unit falls as quantity rises, because it is charged per material family rather than per piece.
How much does a custom bag cost per unit?
Indicative FOB Xiamen prices at a 500-unit MOQ are USD 1.80-4.50 for a simple tote, USD 4.50-9.00 for a basic backpack, USD 7.00-14.00 for a laptop backpack, USD 5.50-12.00 for an insulated cooler bag and USD 12.00-35.00 for a leather handbag. Those figures are indicative only and move with fabric, hardware, print colour count and structure. Inspection and testing sit in the 2-4 per cent quality bucket.
How much does a custom bag sample cost and how long does sampling take?
A sample costs typically USD 50-150 per style and is credited back on the first bulk order. Sampling takes 6-10 working days, or 12-15 for complex builds, and courier adds 3-5 days. Tooling and moulds for custom hardware run USD 300-2,500. Book your laboratory tests inside the sampling window so the shipment waits on a vessel rather than on a report.
How long does custom bag production take with inspection and testing?
Mass production runs 35-50 days after sample approval and deposit, reaching 30 days off-season. Add 2-3 days for final inspection and 5-15 working days for laboratory testing, which should run in parallel with production rather than after it. Sampling adds 6-10 working days, or 12-15 for complex builds. Sea freight then runs 25-35 days.
Which artwork file formats do you need for printing and embroidery?
Supply vector artwork in AI, EPS or PDF with every colour called out as a Pantone TCX reference, plus print size in millimetres and placement with a tolerance. For embroidery supply the same vector file plus a stitch file in DST or PES if one exists. Raster files above 300 dpi at final size work for heat transfer and do not work for screen separations. Embroidery perforates the panel, which lowers water resistance there.
What payment terms and shipping terms apply to a custom bag order?
T/T 30/70 is the default: 30 per cent deposit to start material booking and 70 per cent before shipment, with L/C at sight for container orders. The default term is FOB Xiamen under Incoterms 2020, with EXW, CIF and DDP available. Ask for the final inspection report and carton photographs before releasing the balance, because those are the evidence if a claim arises later.
Need a structured plan for your bag programme?
Send your tech pack or a reference photo and we will return a costed specification, a sampling plan and a production schedule.