A counter-sample is the sealed production reference your factory, inspector, and brand compare against before bulk release. If you skip it, a 300-piece run can pass AQL and still arrive tighter, darker, twistier, or rougher than the shirt you thought you approved.
Most first-time founders approve a nice sample and assume the factory will remember the rest. That is not how bulk apparel works. The cheap protection is to lock one physical shirt as the standard for measurement, wash shade, print feel, and label details before cartons start moving.
What a counter-sample actually is
A counter-sample is the buyer-approved shirt kept as the physical truth after pre-production approval. QIMA's inspection conditions guide says the supplier should provide a reference sample approved by the buyer for onsite comparison. That is the practical point: AQL tells the inspector how many units to check. The counter-sample tells the inspector what "correct" looks like.
In blank apparel, that reference should come from bulk-intended fabric, wash recipe, labels, trims, and decoration. A proto sample proves concept. A fit sample proves silhouette. The sealed counter-sample proves the exact production standard you want repeated. That is the same logic behind our guidance on sample-to-bulk quality drop and sample approval before a blank order.
| Sample type | What it proves | What it still misses |
|---|---|---|
| Proto sample | Concept and rough construction | Final fabric, wash, trim, and bulk consistency |
| Fit sample | Measurement points and silhouette | Bulk wash outcome and final labeling |
| PP sample | Bulk-intended materials and process | Line-start drift after production begins |
| Counter-sample | The sealed physical standard everyone compares against | Later lot drift if nobody checks against it |
| TOP sample | What the line produced at the start | Whether the rest of the lot stayed equal |
What you need to lock before bulk starts
The counter-sample should not be a pretty shirt with vague approval by WhatsApp. It should carry the numbers and decisions that matter. ISO's ISO 6330 home-laundering standard exists because wash behavior changes the garment. AATCC's Test Method 150 for dimensional changes in home laundering and Test Method 61 for colorfastness to laundering exist for the same reason.
For a vintage-washed blank, we usually want five things written beside the sample: chest and body length tolerance, wash shade band, post-wash shrinkage target, print or embellishment condition if decoration is involved, and label content. If those are not fixed, the factory will improvise. Then the inspector can only grade defects, not intent.
| Counter-sample checkpoint | Practical target | Why it matters |
|---|---|---|
| Chest, length, shoulder, sleeve | State tolerance by point, often plus or minus 1.0 cm | A shirt can look off before it fails AQL if core measurements drift |
| Wash shade and handfeel | Approved physical sample plus acceptable shade band | Phone photos flatten the difference between intentional wash character and a bad lot |
| Shrinkage after laundering | Usually 3% to 5% max by spec, depending on body and wash | Post-wash size complaints usually start here |
| Decoration or strike-off condition | Same blank color, same wash state, same artwork thickness | Print behavior on a white test tee does not predict a washed blank |
| Label and packing details | Fiber, origin, RN or name, carton marks, size breakdown | A correct body can still fail because the label or carton standard changed |
This is the same discipline behind building a first-drop QC file and writing clear AQL language. The counter-sample is where those words become a real shirt.
Why AQL does not save you if the sample was wrong
QIMA's AQL explanation and the NIST/SEMATECH sampling-plan reference both make the same point: acceptance sampling is statistical. It helps decide whether to accept or reject a lot. It does not define your product standard. If the factory, inspector, and buyer all compare against a loose memory, the lot can be statistically acceptable and still be commercially wrong.
A simple example: you approve a 260 gsm washed tee with a broken-in handfeel and 77 cm body length in size L. Bulk comes out 248 gsm after wash, 75.5 cm long, and slightly darker. That lot might still pass a Major 2.5 / Minor 4.0 inspection if no defect exceeds the sampling threshold. Your customers will still say the reorder feels different.
The counter-sample turns those arguments into physical evidence. That matters even more when you are selling a washed body from Storiginator's T-shirt collection, where fit, shade, and handfeel are part of the product promise, not decoration around it.
How to run the counter-sample process on a 100 to 500 piece order
Keep it simple. Ask for one PP sample from bulk-intended materials, one sealed counter-sample signed by both sides, and one TOP sample from the first production output. On a 300-piece order, that is still cheap insurance compared with reworking 15 cartons after they are packed.
A suggested sequence is: approve the PP sample, write the tolerance and wash notes on the approval sheet, seal one physical shirt as the counter-sample, send the same notes to the inspector, and require the TOP sample to match before the line runs freely. If the TOP sample drifts, stop there. Fixing one line-start issue is cheaper than debating 300 units after folding.
- Use one signed approval sheet with date, PO number, size, color, and wash recipe.
- Photograph the sample, but never let photos replace the physical reference.
- Record at least 4 measurement points and one post-wash result.
- State whether vintage variation is acceptable and where the boundary becomes a defect.
Common mistakes that make the counter-sample useless
The first mistake is sealing the wrong stage. A fit sample made from substitute fabric is not a production reference. The second is approving a washed blank without a post-laundry check. The third is forgetting labels and packaging because the body looked good. Common risks include all three on small brand orders.
The fourth mistake is not sharing the same sample with the inspector. QIMA's inspection conditions specifically ask for the approved reference sample onsite. If only the merchandiser has it, the inspection team is forced to judge against paperwork and memory. That is how "it looked fine in the office" becomes a freight claim.
Key Takeaways
- AQL tells you how many units to inspect. The counter-sample tells you what the right unit looks like.
- For washed blanks, lock measurement, wash shade, shrinkage, print condition, and label details before bulk starts.
- On a 100 to 500 piece run, one sealed PP reference plus one TOP check prevents expensive line-start drift.
- If the inspector does not have the approved sample onsite, your QC system is weaker than it looks.
Frequently Asked Questions
What is a counter-sample in apparel production?
It is the sealed buyer-approved production reference used during inspection and line checks. On a 300-piece blank order, that one shirt often prevents a full lot from drifting on shade, handfeel, or size by 1 to 2 centimeters.
Is a fit sample enough before bulk production?
No. A fit sample proves shape, not full production conditions. You still need a PP or bulk-intended sample and a sealed reference if you want to control post-wash shrinkage, color drift, and label accuracy.
How many measurement points should be written on the approval sheet?
At minimum, record 4 key points: chest, body length, shoulder, and sleeve. On oversized or vintage-washed tees, additional measurements can include neck opening and sweep because a 1.0 centimeter drift there changes the silhouette fast.
Does AQL 2.5 make a counter-sample unnecessary?
No. AQL 2.5 is only an acceptance threshold for major defects in a sampling plan. It does not define your approved wash shade, fabric handfeel, or exact finished measurement, so you still need a reference standard.
Final Thoughts
The counter-sample is boring. That is why it works. Brand founders who lock one real shirt before bulk starts usually spend less time arguing after production. If you need washed blanks that can be spec'd and checked properly, start with Storiginator.
0 comments