Fabric Test Standards and Acceptance Criteria That Hold Up | Motion
"Good quality" is not a spec. How to turn performance requirements — shrinkage, colour fastness, pilling, GSM — into measurable acceptance criteria with named test methods and a written remedy.
"Good quality" is not a specification — it is a wish. A performance requirement becomes enforceable the moment it carries three things: a number, a named test method, and a remedy for missing it. "Shrinkage under 3% after three wash cycles, measured flat per AATCC 135, failures re-made at supplier cost" is a clause; "fabric must not shrink" is a conversation waiting to happen. This article turns the performance requirements buyers care about into criteria that survive contact with a production run — and with a dispute.
The Anatomy of a Testable Criterion
Every criterion that holds up in a dispute has the same four parts:
• Part | Weak version | Testable version
• The property | "good colour" | Colour fastness to washing
• The number | "must be high" | Grade 4 minimum
• The method | "tested properly" | ISO 105-C06 / AATCC 61, named on the report
• The remedy | "supplier will fix it" | Below grade 4 = lot fails; re-process or credit at supplier cost
Leave out any one part and the clause degrades into an opinion. The property without the number is unjudgeable; the number without the method invites testing games — washed colder, measured wetter; the method without the remedy means a failed test changes nothing. The
written-spec logic for non-returnable custom orders
applies here at full force: the spec is the standard bulk is judged against.
The Performance Properties That Matter, With Standard Values
These are the criteria that carry most apparel programmes. The values below are the common trade baselines — tighten or loosen per programme, but name a number either way:
• Property | Common acceptance baseline | Named method (example)
• Shrinkage (length and width, separately) | ≤ 3% after 3 wash cycles | AATCC 135, measured flat
• Colour fastness to washing | Grade 4 minimum | ISO 105-C06 / AATCC 61
• Colour fastness to rubbing | Grade 3–4 minimum, dark shades | ISO 105-X12 / AATCC 8
• Pilling resistance | Grade 4 minimum | ISO 12945-2 (Martindale)
• Fabric weight | Nominal GSM with ±5% tolerance | GSM cut-and-weigh, 3–5 positions across width
• Print adhesion | No visible degradation after 50 domestic wash cycles | Wash test on production-cured sample, photographed
Two fields trip buyers most often. First, tolerances belong to the spec, not the supplier's mood — a GSM band with its tolerance decided at quote time is a different product from "about 180 GSM". Second, shrinkage needs a measurement convention — "measured flat after three cycles at 40 °C" beats "normal wash" every time, because shrinkage measured differently is shrinkage disputed (the full convention is in the
shrinkage spec guide
).Testing: Who Tests, When, and What It Costs
Three testing moments cover a programme without buying a laboratory:
1. Sample-stage verification — the supplier tests the production-intent fabric and passes the report with the sample. Costs nothing extra; it is the proof the spec was heard.
2. Pre-production lot test — the actual bulk fabric lot tested before cutting. The cheapest place to find a failure: the cloth can still be returned or re-dyed.
3. Pre-shipment inspection — the
AQL inspection
checks the criteria that are visible in finished goods (workmanship, measurement, appearance) while the lab criteria rest on the lot test plus the retention sample.Independent third-party testing exists for high-stakes programmes: a laboratory panel for your market's requirements is quoted per scope, and either side can commission it — the PO names who pays when it passes (usually the buyer, as due diligence) and who pays when it fails (always the supplier, whose lot was in question).
Writing the Clause: A Copy-Ready Skeleton
One paragraph carries a whole quality programme. Fill the brackets per programme:
> Bulk conforms to the approved sample and this spec: fabric [180 GSM, ±5%, 65/35 combed cotton-polyester]; shrinkage [≤3% after 3 washes, AATCC 135, measured flat]; colour fastness [grade 4 wash / grade 3–4 rub]; pilling [grade 4]; print [no visible degradation after 50 washes, bag-tested 48 h]. Verification: supplier lot test report before cutting; AQL 2.5 pre-shipment inspection. Failures: re-process, re-make or credit affected units at supplier cost, buyer's choice, reported within [14] days of delivery with photos.
The clause names every number, every method and the remedy in one place — the
durability wash protocol
plugs in for decorated goods, and nothing in it requires goodwill to enforce.When the Number Is Missed: The Remedies That Bite
A criterion without a consequence is decoration. The remedy ladder that works, written before production:
- Marginal miss — inside a band the spec allows the supplier to discount against the invoice, agreed as a percentage stated in the PO.
- Meaningful miss — re-processing or re-making the affected units at supplier cost, on a stated timeline.
- Systematic failure — the lot fails as a lot; credit or replacement per the escalation ladder, and the retention sample decides the argument.
The
change-control division
decides whose wallet a failure comes from: the spec is the boundary between "supplier's problem" and "buyer changed their mind". That is precisely why the numbers get agreed while everyone is friendly.Frequently Asked Questions
Do I need to name international test methods, or is a number enough?
A number without a method invites testing games — the fabric can be tested wet, cold or briefly and land on a different grade. Naming the method closes the loophole and costs one sentence.
What fastness grade should I accept for workwear?
Grade 4 to washing is the common baseline for garments that launder often; rubbing fastness at 3–4 covers dark shades on light skin contact. Programmes with industrial laundering should raise the bar and say so in the spec.
Who pays for third-party lab testing?
By agreement, written in the PO. The common split: the buyer pays as due diligence when the lot passes; the supplier pays when it fails, since the question was their lot. Commissioning rights should belong to both sides.
Is a supplier's own test report enough?
It is a start, not a proof. Supplier reports cover the sample-stage check; for high-stakes lots, a pre-cutting lot test from an independent lab — or at minimum the supplier's report against YOUR named method — is the enforceable version.
What GSM tolerance is normal?
±5% of nominal is the trade standard, measured by cutting and weighing at 3–5 positions across the width. Without the tolerance written, a "180 GSM" fabric arriving at 165 GSM is technically a dispute and practically an argument.
How do I make the criteria survive a dispute?
Four parts, one place: property, number, method, remedy — signed at order time with the approval sample filed beside them. The sample shows what good looks like; the criteria show how it is measured; the remedy says what happens when it misses.
Want your performance requirements turned into a clause?
Send the programme's must-haves — we will return the spec block with named methods and baselines filled in, ready to attach to the quote.
