Agree the standard before production, not at inspection
“Minor colour variation” means different things to different inspectors, and the argument always happens when the goods are already made. Two things to settle in writing at order confirmation:
| What to agree | Why it matters |
|---|---|
| AQL level | AQL 2.5 for major defects and 4.0 for minor is the common commercial standard for drinkware, sampled to ISO 2859-1 Level II. On a 3,000-piece order that is a 125-piece sample, accepting at 7 major defects and rejecting at 8. Without a stated level, “pass” is an opinion. |
| What counts as major vs minor | A dent on the body is usually major; a faint mark inside the lid is usually minor. Print misregistration is the one buyers most often forget to classify, and it is the one they most often reject on. |
| The reference sample | The approved physical sample with its Pantone number, held by both sides. Not the artwork file — screens and print do not match screens. |
The eight things an inspector should photograph
- Carton markings and count — against the packing list, before anything is opened.
- Print alignment on several pieces from different cartons, not all from the top one.
- Colour against the approved sample, in daylight, with the sample in the same frame.
- The inside of the lid — gasket seated, threads clean. This is where assembly problems show.
- A filled leak check — inverted for at least 30 minutes, not a quick upright look.
- Vacuum check on insulated items — fill with water at 95 °C and check the outer wall after 5 minutes — anything warm to the touch means the vacuum has failed.
- Base and rim for dents, which is where transit damage concentrates.
- The label or hang tag, including barcode scan — a barcode that will not scan at retail is a full-shipment problem.
Schedule it before the balance, not after. An inspection report that arrives after payment is a record, not leverage. We book third-party inspections (SGS, Intertek, TUV, BV or your own agent) before the balance is due, and we will hold shipment while a re-check is done if something fails.
Our own line check, for reference: we inspect 8% of every batch against your approved sample, and every lid goes through a leak test. That does not replace your inspection — it means the things your inspector finds should be edge cases, not patterns. If they find a pattern, that is our problem to fix before the container moves.
Why Factory QC and Third-Party Inspection Both Matter
A common misconception: 'If the factory does QC, I don't need third-party inspection.' These two processes serve different purposes.
Factory QC (what Tim does) is designed to catch production defects before they leave the production floor. It's systematic, covers every unit at certain checkpoints, and is optimized for throughput. Its weakness: it's inherently subject to commercial pressure — the factory has an incentive to ship on time.
Third-party inspection (SGS, Bureau Veritas, Intertek) is a random sampling process conducted by an inspector with no commercial relationship with the factory. Its strength: independent accountability. Its weakness: sampling means it can miss small-lot defects.
For your first 3–5 orders with any factory, do both.
The Pre-Production Checklist
Before bulk production begins, verify:
Material certificates: Confirm the steel coils being used for your order match the FDA/LFGB test report specifications. Request incoming material test certificate.
Color approval: A production-run color sample (shot during the painting setup) should be approved against your reference Pantone before full production runs.
Logo print proof: Physical sample with your artwork applied — not just a digital rendering. The logo position, size, color, and edge quality should match your approved reference.
Packaging sample: If using custom boxes, the finished sample should be approved before 1,000 cartons are assembled.
The Pre-Shipment Inspection Checklist
Tim's checklist for final inspection. Request this report from any factory you work with:
Visual inspection (100% of units): Scratches, dents, coating bubbles, overspray, printing misalignment, lid defects.
Vacuum performance (sample): Fill with boiling water, measure after 6 hours. Minimum: 65°C. Average: should be 75–80°C for a properly vacuum-sealed unit.
From our side of the bench: we run that vacuum check before painting, not after — a seal that fails on a bare body is a cheap reject, but one that fails after coating is a finished unit in the bin. Sequencing the test that way is most of why our own pre-shipment defect rate stays under 0.5%.
Lid function (sample): Open/close 10 times, verify threading, no stiffness, no leakage when inverted for 30 seconds.
Logo quality: Match against approved reference under consistent lighting. Check adhesion (tape test for printed logos).
Packaging: Correct barcode scannability, correct labeling (Prop 65 if required), correct carton quantities.
Dimensions: Random check of height, diameter, and weight against spec sheet tolerances (±3mm).