Sampling is not one event. It is four, and each one exists because it catches something the previous one cannot.
1. Counter sample
The question it answers: did the factory read the specification the way you meant it?
Made from the tech pack or the reference sample, in a comparable fabric. You are not judging the final hand-feel yet — you are checking that the construction, the proportions and the closures are what you described. This is the round where most corrections are structural: the neckline is too wide, the crotch snap is in the wrong position, the placket opens the wrong way.
2. Fit sample
The question it answers: does it fit and behave the way you meant?
Made in the intended fabric at the intended weight. A bubble romper pattern that looks right in a light jersey will look wrong in a 300 gsm French terry, because the fabric's own body changes how the volume reads. This is the round where you put it on a child, or on a mannequin of the right size, and decide.
This is also the round to wash it. Shrinkage is a fit variable. A garment that fits on the table and not after the first wash has not been sampled properly.
3. Size set
The question it answers: does the grading hold at the ends of the range?
One garment per size, in the full range. Grading errors hide in the middle of a range and reveal themselves at the extremes. A 0-3M bodysuit that fits perfectly and an 18-24M that gaps at the leg means the grade rule is wrong, not the pattern.
Babywear grades tightly. The increments between sizes are small, so a small error compounds across the range faster than it does in adult clothing.
4. Pre-production sample
The question it answers: is what we are about to make the same as what you approved?
Made with the actual bulk fabric, the actual trims, the actual labels, on the actual production line. This is the last gate. It is also the sample that becomes the reference for the inspection — the one an inspector measures against when the lot comes off the line.
Why the rounds are not optional
Each round is cheaper than the one after it. A correction at counter-sample costs a sample. The same correction at pre-production costs a sample plus a delay. The same correction after bulk has been cut costs the whole lot.
How to make sampling fast
- Consolidate comments. One list per round, from one person. Three emails with one correction each is three rounds of work.
- Mark up the sample. A photograph with the problem circled and a measurement attached removes all ambiguity. "The sleeve is too short" is a conversation; "sleeve should be 24 cm, sample measures 21.5 cm" is an instruction.
- Decide in one pass. If you are waiting to hear from a third party, say so — the factory will hold rather than guess.
- Tell them your deadline at the start. If the pre-production sample has to be approved by a date, that date constrains the whole calendar.
Sample charges are quoted with the garment price so you can see both at once. The charge covers material and the pattern-room time; it is not a profit centre, and it is not usually credited against the bulk order unless that was agreed in writing.