Almost every sourcing failure in this category has the same shape. The sample was good, the price was agreed, the bulk order arrived, and something was subtly different. Not wrong enough to reject outright, but different enough that the finished product no longer matched the one that was approved.
The cause is rarely dishonesty. It is that a sample is made from a small batch with extra attention, and a bulk order is made to a specification that may not have captured what the buyer actually cared about. Evaluating a supplier is therefore less about judging a sample and more about finding out what the factory does when nobody is watching.
1. Start from what the stone must survive
Before any supplier conversation, define the service condition. A rhinestone on a pair of jeans washed weekly for two years is a different specification from a rhinestone on a bridal headpiece worn once, and both are different from a stone set into a nail design that will last three weeks.
Three variables define the service condition, and every sensible specification question follows from them.
| Variable | What it changes |
|---|---|
| Number of wash or wear cycles | Coating thickness and the protective top layer become decisive; a single-wear stone does not need a wash-rated coating |
| Mechanical stress on the stone | Whether the risk is coating abrasion, adhesion failure or physical breakage |
| Viewing distance | Whether facet precision and colour consistency are visible enough to justify the cost of achieving them |
Writing these down before requesting a quote changes the conversation. A supplier who receives a defined service condition can quote to it. A supplier who receives only a size and a colour will quote the cheapest stone that matches those two data points.
2. What to sample, and what a sample can and cannot tell you
A sample proves that the factory can produce the stone. It does not prove the bulk order will match it. The distinction matters because most first-order evaluation effort goes into judging the sample itself, which is the part already known to be good.
What a sample genuinely tells you:
- Whether the facet geometry and the sparkle are in the right class for the price.
- Whether the colour is acceptable against a reference, under the lighting you actually care about.
- Whether the coating covers the whole stone including the edges, which is where failure usually starts.
What a sample does not tell you:
- Batch-to-batch colour drift, which only appears on the second and third delivery.
- Coating durability, which needs a wash or abrasion test rather than a visual check.
- Whether the count per pack is accurate, which requires opening a production pack rather than a sample pack.
This is why the sample approval step and the bulk order should be treated as two separate decisions, not one. The samples and lead times guide covers the sequence, including what a sample approval should record so that the bulk order has something to be compared against.
3. The defects worth measuring
Not every defect matters equally. Four categories cover most of what goes wrong, and each has a different root cause, which means each points at a different conversation with the supplier.
| Defect | What you see | Root cause to investigate |
|---|---|---|
| Coating peel or flake | Metal film lifting from the stone, usually starting at the rim | Base adhesion layer quality, or surface contamination before coating |
| Discolouration | Backing darkens, yellows or turns greenish with age or washing | Missing or inadequate protective top layer over the metal film |
| Colour drift between batches | The same reference code arrives visibly different the second time | Colour layer thickness control, or no retained reference standard |
| Size and count deviation | Stones measure outside the stated tolerance, or pack count is short | Cutting tolerance not specified, or counting done by weight without a stated conversion |
The first two are coating problems, and the vacuum coating guide explains why the layer structure is the cause. The third is a process control problem. The fourth is a specification problem, and it is the easiest of the four to prevent, because a millimetre figure and a count tolerance written on the order removes most of the ambiguity.
4. Questions that reveal actual capability
Supplier questionnaires tend to ask questions with obvious right answers. These do not.
"What is your in-house inspection routine, and can I see the record?"
A factory that inspects has a routine with named checkpoints. On this site the routine runs through raw material verification, board dimensions, thickness, density, surface condition, edge quality, compression and packaging marking. A factory without a routine answers in generalities, or answers with certifications instead of a process.
"How do you hold a colour reference between orders?"
The correct answer involves a retained physical standard, not a code in a database. Colour codes alone do not survive a change of coating batch. If the factory cannot describe how the reference is physically maintained, batch drift is likely.
"What happens if a bulk order does not match the approved sample?"
This question is about remedy, not about quality. A supplier who can describe a specific rework, replacement or credit process has dealt with the situation before. A supplier who cannot has probably not been held to a standard.
"Which parts of my specification can you not meet, and what would it cost to meet them?"
Asking directly which requirements are difficult produces more useful information than any open question about capability. Every factory has limits, and a supplier who names them is more reliable than one who agrees to everything.
5. Documents and terms to settle early
The commercial side of a first order is where small ambiguities become expensive. These are the items worth writing down before the order rather than after.
- Delivery term with a named location. A short code without a port or place is not a quotation that a forwarder can price. The export documents and shipping guide sets out how the term decides who arranges and pays for each leg.
- Incoterm-consistent document set, including invoice, packing list, bill of lading or air waybill, and a certificate of origin where the destination requires one.
- Payment staging tied to defined milestones, so that the balance payment is released against something verifiable rather than against a shipping date alone.
- MOQ stated per colour and per size, not as a single order total, because the binding constraint is almost always the smallest line. The MOQ planning guide explains how the figure is built.
6. Scaling up: the second order is the real test
The most informative moment in a supplier relationship is the second order, not the first. The first order gets attention because the relationship is new. The second order shows what normal looks like.
Before scaling volume, check four things against the first delivery:
- Does the colour match the approved reference under the same lighting, without adjustment?
- Does a wash or wear test on the new batch produce the same result as the first?
- Is the pack count and pack composition the same?
- Did the lead time hold, and if it moved, was it communicated before the promised date rather than after?
If all four hold, volume can rise. If any one fails, it is worth resolving before increasing the order, because a defect at low volume is a sample problem and the same defect at high volume is a production problem.
A short checklist for the first order
| Stage | What to complete |
|---|---|
| Before quoting | Define service condition: wash cycles, mechanical stress, viewing distance |
| Sample stage | Judge facet, colour against reference, coating coverage to the edge; record the approval with photographs and lighting conditions |
| Before ordering | Write size in millimetres, count tolerance, pack format, colour reference, MOQ per colour, delivery term with location |
| On arrival | Open a production pack, not the sample; check count, colour against the retained reference, compartment integrity, coating at the rim |
| Before scaling | Repeat the colour and durability checks on the second delivery; confirm lead time behaviour |
None of this is complicated. It is simply the difference between buying a stone and buying a repeatable supply of a stone. For buyers who also need custom colour development, the colour matching process and the OEM bulk order checklist cover the two areas where this distinction costs the most to get wrong.
