WEBVTT

00:00.100 --> 00:05.165
A sample that looks right does not prove that the production lot will stay consistent.

00:05.115 --> 00:08.170
First, freeze a measurable baseline.

00:08.170 --> 00:16.353
Put the drawing revision, material, colour, critical dimensions, functions, packaging and labels into one specification pack.

00:16.353 --> 00:19.729
Avoid vague language such as same as sample.

00:19.729 --> 00:22.963
Second, give the sample an identity.

00:22.963 --> 00:29.059
Record its number, production date, tooling or line, inspection results and photographs.

00:29.059 --> 00:34.136
After approval, the buyer and supplier should each retain the same approved version.

00:34.136 --> 00:37.087
Third, approval is not the end.

00:37.087 --> 00:43.930
A change in material, structure, process, source, packaging or labelling needs a documented review.

00:43.930 --> 00:51.688
If safety, regulation or a critical function may change, submit a new sample and repeat the relevant testing and approval.

00:51.688 --> 00:55.708
Fourth, separate sample approval from lot inspection.

00:55.708 --> 00:59.780
The approved sample shows what one agreed reference looks like.

00:59.780 --> 01:09.819
It does not prove that every production unit conforms, and it does not replace market-specific testing, risk analysis, technical documentation or certification.

01:09.819 --> 01:16.481
Before shipment, give the inspector the approved sample, specification pack and defect classification.

01:16.481 --> 01:22.603
Acceptance sampling evaluates the lot; the approved sample supplies the comparison baseline.

01:22.603 --> 01:26.353
Good sample approval is not one word: approved.

01:26.353 --> 01:35.193
It is an evidence chain with controlled revisions, measurable criteria, retained references, traceable changes and a separate production inspection.