Last updated: October 2026
By Jack Lv, General Manager — 925 sterling silver and GRA moissanite manufacturing in Foshan since 2011
Most failed bulk orders do not come from a bad sample. They come from an unsealed one. The buyer and the factory should each sign and retain a sealed gold sample, lock every production parameter in writing before the run starts, and approve a first article before the balance of the order is finished and packed.
Because the sample is usually made by the most experienced setter on the floor from the best-matched stones, while the bulk run is made by the line. The drift shows up as a stone that reads warmer or greyer, a different rhodium tone, prongs sitting higher or lower, or a lighter clasp.
The most expensive failures in jewelry sourcing rarely start with a sample that looks wrong. They start with a sample that looks right and a bulk shipment that does not match it. The sample is usually made by the most experienced setter on the floor, using the best-matched stones in the loose-stone tray and the freshest casting available. The bulk run spreads the same job across several shifts, several casting lots and several polishers, and every handoff adds a small variation that compounds.
Drift shows up in ways buyers notice in the first minute of a showroom review: a stone that reads slightly warmer or slightly greyer than the approved piece, a rhodium tone that behaves differently under retail lighting, prongs sitting higher or lower on the girdle, a lighter clasp, or a chain that hangs with a different drape. None of these is a defect on its own. Together they are a different product from the one the buyer approved and priced.
The correction is procedural rather than technical. A factory can only hold a target that has been defined, sealed and written down. That means three instruments: a sealed physical reference held by both sides, a locked parameter sheet, and a checkpoint before the full run is completed.
A gold sample is one physical unit that buyer and factory both sign, seal and retain; where production and the gold sample disagree, the gold sample wins. A sales sample is marketing material that gets handled and mailed, and it should never be the reference in a dispute.
A gold sample, also called a sealed sample, a golden sample or an approval sample, is one physical unit that both the buyer and the factory sign, seal and retain. It is the contractual reference: where production and the gold sample disagree, the gold sample wins. The word sealed is doing real work here. A sample left loose in a drawer is not a gold sample, because it can be swapped, adjusted, polished or confused with a near-identical piece.
A gold sample is not a sales sample. A sales sample is marketing material and is handled, photographed and mailed around. A gold sample is a quality reference with a paper trail, and it should never leave the custody of the party that sealed it.
The gold sample should capture the whole piece, not just the centre stone. It fixes the metal alloy, the stone grade, the plating colour and finish, the prong profile, the clasp mechanism, the engraving or hallmark position, and the packing method. Anything the buyer cares about in the final shipment should be visible on the gold sample.
Produce duplicates — one retained by each side — sealed separately, labelled identically and photographed from the same angles, so a later dispute can be settled by comparing photographs rather than memories. Sealing means tamper evidence: a signature or tape across the closure so any opening tears it, plus a label carrying the order reference, date and stone grade.
Produce duplicate gold samples. One copy stays with the buyer and one copy stays with the factory. Both are sealed separately, labelled identically, and photographed from the same angles so that a later dispute can be resolved by comparing photographs rather than memories.
Sealing means tamper evidence, not a plastic bag. Typical practice is to place the piece with its label in a sealed bag or box, then run a signature or a tamper-evident tape across the seal so that any opening tears the tape. The label should carry the order reference, the date, the stone grade, the plating colour and the ring size or chain length.
Retention matters as much as sealing. Both copies should be kept for the full production cycle plus the claim window agreed in the contract. The factory copy should stay closed unless a dispute is raised, and it should only be opened with both parties present or on camera. The buyer copy should not be worn as a display piece, because wear changes the polish and the plating and destroys its value as a reference.
Metal, stones, finish, construction, dimensions and plating thickness at minimum, converted from the physical sample into instructions a line can follow. Two clauses deserve extra attention: bulk stones must come from the same grade and matched lot as the sample, and any plating or stone substitution needs written approval.
A sealed sample without a written parameter sheet leads to arguments about tolerance. The sheet converts a physical object into instructions a production line can follow, and it gives the quality team something to measure against. The minimum useful set of parameters covers metal, stones, finish, construction, packing and documents.
Two clauses are worth extra attention. First, stone lot: specify that the stones for the bulk order must come from the same grade and the same matched lot as the sample, and that any substitution requires written approval. Second, plating: state the colour and the finish, and state that the plating thickness is measured in microns and controlled to the agreed specification, without quoting a figure in the contract summary that the workshop cannot verify on the bench.
| Parameter group | What to write down | Why it prevents drift |
|---|---|---|
| Metal | S925 sterling silver, 92.5 percent silver with the balance typically copper, plus any non-silver findings | Stops a lower alloy or a different spring-hardness wire entering the run |
| Stones | Moissanite grade, cut, colour band, clarity band and the approved matched lot | Stops a warmer or greyer parcel being pulled from stock mid-run |
| Finish | Plating colour and finish, polished or satin areas, tarnish-prevention method | Stops tone shifts under retail lighting |
| Construction | Prong profile and seat height, stone pattern, clasp type, hinge and bail geometry | Stops the piece reading as a different product in hand |
| Measurement | Ring size system, chain length, calibrated stone positions | Stops sizing surprises across a size run |
| Packing | Poly bag or box, tag wording, carton marks, polybag quantity per carton | Stops the shipment arriving in an unusable format |
| Documents | Test reports, stone reports, certificate copies attached to the order | Gives customs and retail compliance teams the paperwork they expect |
The factory submits the first finished unit or small batch for comparison against the gold sample before the rest of the run is completed and packed. It is the cheapest gate in the order, because a problem found there costs a correction rather than a shipment.
First article inspection, often shortened to FAI, means the factory submits the first finished unit or the first small batch of the run for comparison against the gold sample before the remainder is completed and packed. It is the cheapest quality gate in the whole order, because a problem found at the first article costs a process change, while the same problem found at the port costs a return shipment.
The first article should be inspected under the lighting the buyer actually sells in, side by side with the gold sample, not alone on a bench. A piece compared against nothing always looks correct.
The output of a first article review should be a written decision, not a verbal one. Approve, approve with a listed correction, or reject and re-submit. A conditional approval that names exactly what must change is far more useful than a friendly verbal okay, because it creates a record if the correction is not made.
Treat it as routine if it is caught before shipment, and fix the process rather than only the units. Ask for the sort record showing how many units were inspected, reworked, downgraded and scrapped — that is what tells you whether the problem was isolated or systemic.
A deviation found before shipment is a routine event, not a crisis. The order goes wrong when the deviation is discovered after delivery, when the correction is negotiated informally, or when only the defective units are replaced without the process being fixed. The following sequence keeps the order and the relationship intact.
Buyers should ask for the sort record rather than only the replacement. The sort record shows how many units were inspected, how many were reworked, how many were downgraded and how many were scrapped, and it tells the buyer whether the problem was isolated or systemic.
IQC checks incoming material before it enters production, IPQC watches the line while work happens, FQC checks finished goods and OQC checks the outbound shipment. In our runs IQC verifies the loose moissanite lot and the sterling silver with XRF and physical comparison, while IPQC is where plating tone, seat height and polish quality start to drift.
Our production runs on a four-stage inspection routine: IQC, IPQC, FQC and OQC. IQC is incoming quality control, which checks what arrives at the factory before it enters production. IPQC is in-process quality control, which watches the line while the work is happening. FQC is final quality control, which inspects the finished piece. OQC is outgoing quality control, which checks the packed carton and the shipping marks before dispatch.
IQC is where the loose moissanite lot and the sterling silver are verified against the approved specification, using an XRF spectrometer for the alloy and physical comparison for the stones. IPQC is where plating tone, seat height and polish quality start to drift first, because those are the steps most sensitive to operator and shift changes. FQC catches setting and finish defects on the assembled piece. OQC is the last chance to catch a labelling or packing error that would otherwise turn into a retail complaint.
For a buyer, the practical value of a four-stage routine is that drift is caught at the step that caused it. A parameter sheet tells the line what to do. A four-stage routine confirms that it was done.
Attach the paperwork before the run starts, not after: the Intertek Verified Supplier certificate 483553466_T, valid 2025-12-31 to 2026-12-30 and issued after an on-site audit, and the RoHS conformity declaration under Directive 2011/65/EU as amended. State plainly that for jewellery CE is not a general safety mark — our document covers only the models named on it.
Attach the paperwork to the order before the run starts, not after. Buyers with retail compliance obligations usually want the Intertek Verified Supplier Certificate 483553466_T, valid from 2025-12-31 to 2026-12-30 and issued after an on-site audit; the Certificate of Conformity for RoHS under certificate HYC2025120053RV1, supported by test report HYC2025120053R01 and covering the 27 models listed on that certificate; and the REACH SVHC Test Report BKC25020516BR, which screened 247 SVHC substances with all results below 0.1 percent. Moissanite is documented with a GRA report issued per stone, so there is no single shared report number to quote.
One compliance point is worth stating plainly, because it is widely misstated in the trade. For jewelry, CE is not a general safety mark. Our document is a RoHS conformity declaration under Directive 2011/65/EU as amended by (EU) 2015/863 and (EU) 2017/2102, and it applies only to the models named on the certificate. Buyers should not describe it as a general CE certification for jewelry in their own listings.
Timing should be written into the same agreement so that the seal, the first article and the shipment hold together. Sampling takes 5 to 7 days. Bulk production takes 15 to 20 days. Quotes are answered within 24 hours, Monday to Saturday, GMT+8. The minimum order for custom work is 10 pieces per style, or 30 for multi-stone settings, with tiered pricing that is unchanged at 10 pieces, reduced by 3 percent at 50 pieces and by 6 percent at 100 pieces. Both DDP and DDU terms are available, and the buyer should confirm which of the two the quote assumes before the sample is sealed.
Nabest is the brand of Foshan Nameituo Jewelry Co., Ltd., a jewelry factory established in 2011, operating a 3,000 square metre facility with 6 production lines and a monthly capacity of 15,000 pieces, staffed by more than 300 people and serving more than 3,000 customers across more than 200 countries and regions. We work on sterling silver with moissanite for wholesale and custom orders; stocked designs can ship from a single piece. The loose moissanite we set is purchased from specialist stone suppliers; we set stones and build finished pieces, and we do not cut stones in house.
Both parties should, in duplicate. The buyer copy is the reference for incoming inspection and for any claim, and the factory copy is the reference the line works against. A single sample held by only one side makes every later comparison an argument about which object is authoritative.
Keep both copies through the full production cycle and for the whole claim or warranty window agreed in the contract. The factory copy should be opened only when a dispute is raised and, ideally, only with both parties present so the seal remains credible.
No. The sample proves what the process can achieve on a good day; the first article proves what this specific run is producing. Skipping it moves the discovery of drift from the workshop, where correction is cheap, to the buyer's warehouse, where correction is not.
Then the tolerance was never defined. Tolerance has to be agreed before the run, in writing, parameter by parameter. If a deviation is visible against the sealed gold sample under the same lighting, it is a real deviation regardless of what the bench calls normal.
It rarely changes the unit price, and it usually protects margin. The sample itself is quoted as part of sampling, which takes 5 to 7 days, and the sealed reference prevents the rework, sorting and return freight that a drifting run would otherwise create.
Send your design, size and quantity — factory quote back within 24 hours. Reference: General enquiry