The sample was perfect. The bulk order is not. Tiles come in a slightly different shade, aluminium profiles feel thinner, and the sanitary ware glaze looks duller. Most buyers meet this once, and it usually costs more than the sample ever did.
This guide explains why approved samples and bulk production drift apart, and what to write down before the deposit so you have something to measure against. It applies to construction and building materials, where drift is harder to spot than on a mug or a T-shirt.

Why bulk production differs from the approved sample
Samples and bulk orders are often made in different conditions. Knowing the usual causes tells you what to control.
- Different workshop or line. A hand-finished sample from the development room is not the same process as a production line running thousands of units.
- Different material batch. Raw material, glaze, coating or glass changes between lots, even from the same supplier.
- A vague specification. If only a photo was approved, the factory decides what “the same” means.
- Cost pressure. Small cuts in thickness, coating or packing add up over a large order, especially on repeat orders when attention drops.
- Subcontracting. Part of the order moves to another workshop without telling you.
- No owner for the hand-over. Nobody is responsible for carrying the sample standard into production.
Buyers often report that the factory then calls the difference normal variation. Without a written tolerance, that answer is hard to dispute.
What drifts in construction materials
Generic sampling advice is written for consumer goods. Building materials drift in specific places:
| Material | Where drift shows up |
|---|---|
| Porcelain tiles | Shade between batches, size caliber, flatness |
| Aluminium windows and doors | Profile wall thickness, powder-coat or anodising thickness, hardware |
| Glass | Thickness, tempering, edge finish |
| Steel structures | Steel grade, coating, weld finish |
| Sanitary ware | Glaze finish, dimensional warp, fitment |
Each of these can be measured. That is the point: write the measurable item and the limit into the approval, not a general phrase like “same quality”.
Step 1: label the sample by its role
Not every sample is a reference. Label each one:
- Development sample: proves the idea. Not a standard.
- Pre-production sample: made on the production line with production materials. This is the one you approve.
- First-production sample: pieces taken early in the real run, compared to the approved sample.
- Pre-shipment sample: pieces taken from the finished lot at inspection.
Only the second one should become the reference. A sample the factory cannot reproduce on its line is a warning sign, not a standard. For how to request and evaluate each one, see getting samples from Chinese factories before you commit to production.
Step 2: turn approval into a signed sheet
Approval by email message is not enough. The approval sheet should list:
- Specification and material grade
- Finish, colour code or shade reference
- Measurable tolerances taken from the factory or the applicable standard, not invented
- Packaging method
- Defect classes: what counts as minor, major and critical
- Photographs of the approved sample from several angles
Keep sealed, dated reference copies. The common practice is one copy with the factory, one with the buyer side and one with the inspector, as described in HKTDC’s note on golden samples. Ask whoever coordinates your order to confirm where your reference copies are kept.
Step 3: write a change-control rule into the order
One sentence in the purchase order prevents many disputes: no change of material, subcontractor, factory or packaging without written buyer approval. For tiles, add a batch rule: one shade and one caliber per order line, with batch and shade codes on the carton labels. If the order cannot come from one batch, shade-matched lots should be kept on separate pallets.
Step 4: inspect at three points, not one
A single check right before shipping can only reject finished goods. Three checks catch problems while they are cheap to fix:
- Materials in: raw material, glaze, coating or glass against the approved sheet.
- Early production: the first pieces off the line, side by side with the reference sample.
- Pre-shipment: random pieces from the finished lot, measured and compared with the reference. A sampling plan such as ISO 2859-1 sets how many pieces to pull. Agree the acceptance level in the order, as the standard’s tables are copyrighted and the right level depends on the product.
For tiles, colour disputes are usually settled with an instrumental measurement under ISO 10545-16, which covers small colour differences on plain glazed tiles. See also our pre-shipment inspection guide for construction materials and the quality control checklist for tiles, sanitary ware and doors.
At JEES Global, inspection findings come back as a photo report within 48 hours of the inspection, so a decision can be made while the goods are still in the factory.
What to do when drift is found
Compare the goods to the written tolerance, not to memory. Then pick one outcome and state it in the order before the deposit:
- Within tolerance: accept.
- Rework: factory fixes at its cost, then re-inspect.
- Reject: the lot does not ship.
- Price concession: only if you agree the goods are still usable.
Hold the balance payment until the pre-shipment result is known. See China supplier payment terms for how to structure that. If the drift is only found at destination, the process in making the claim for a damaged, short or wrong shipment applies.
Limits you should know
Sampling inspection supports a decision on the lot. It does not prove every unit is good. A visual check also cannot verify material composition; for that you need a test report from a laboratory. A good process narrows the risk, it does not remove it.
Common mistakes
- Approving a “nearly right” sample on a promise to fix it in bulk.
- Leaving tolerances undefined.
- Approving a sample made by a different workshop from the one making the order.
- Relaxing control after several good orders.
- Choosing the factory on the lowest price alone.
Frequently asked questions
Why does bulk production differ from the approved sample?
Samples are often made by hand or on a different line, and materials change between batches. Without written tolerances, the factory can call the difference normal.
What should a sample approval document contain?
Specification, material grade, finish or shade code, measurable tolerances, packaging, defect classes and photographs. For tiles, add batch and shade codes.
Who keeps the golden sample?
Common practice is three sealed, dated copies: one with the factory, one with the buyer side and one with the inspector.
What can I do if bulk goods fail after production has finished?
Measure against the written tolerance, then choose rework, rejection or a price concession as agreed in the order. Keep the balance payment until the result is known.
When should I inspect?
At three points: materials in, early production and pre-shipment. A single final check can only reject finished goods.
Related: Quality inspection in China
Pingback: How to Import Porcelain Tiles from China Without Shade, Size or Breakage Problems – JEES Global