The Golden Sample: How to Lock a Kids Scooter Sample So Bulk Matches It
Almost every buyer seals the wrong object. The sample is approved in the sample room, signed, sent back to the factory in a bag — and it is a hand-built one-off, assembled by the same engineer who answered your technical questions. Then the bulk order runs on a line, with a different jig, a different operator, and a purchasing department under instructions to hit a cost target. When the second container arrives with a thinner tube wall and a blue that no longer matches, the sealed sample sits in a drawer offering no protection at all, because it never represented production in the first place.
A golden sample is not a souvenir of approval. It is the physical definition of the specification, and it only works if it is pulled off the production line, labelled to the point where it can be argued about, and signed by both sides. The sample order guide covers what a sample costs, how long it takes and what to check when the box lands. This piece covers what happens after you say yes.
The sample you approved is not the sample you should seal
On this range there are two kinds of sample, and they are made in different rooms. A development sample exists so you can judge shape, colour and feature set. It is assembled by hand at the factory's sample station, often from parts pulled off a stores shelf, and it is the right object to approve a design on. A production sample is built on the line, by the operator who will build your order, on the jig and from the material batch that will actually be used.
The gap between those two objects is where first orders go wrong. The one difference buyers notice most is tube wall thickness, because it is the cheapest thing a factory can take out and the hardest thing to see: 1.2 mm down to 1.0 mm removes roughly a sixth of the steel in that tube, the scooter still folds, still rolls, and still passes a static load check. Nothing about it looks different. The second is colour, which drifts when a powder batch is replaced and the factory re-plates to the nearest match rather than telling you. The third is a bought-in part — a grip compound, a bearing, a brake spring — swapped for an equivalent that is equivalent on paper only.
So the first instruction to give before any sealing happens: the piece I seal has to come off the line, not out of the sample room. Ask for the first-article run — the first 30 to 50 pieces built with production tooling, production material and the operator who will run your order — and pull your sealed pieces from that run. Pull them at random from the middle of the stack, not the three on top, because the three on top are the three the supervisor stood and watched being made.
Three pieces and one carton, sealed at one moment
The number that works is three assembled units plus one flat carton, and each copy has a different job. One copy stays at the factory, and it has to live at the end of the line it belongs to — wired inside a clear bag to the QC station, not filed in an office drawer. A reference sample in an office is a reference sample nobody opens.
| Copy | What it is | Where it lives |
|---|---|---|
| Factory copy | One assembled unit from the first-article run | At the QC station on that line, bagged and labelled |
| Buyer copy | One assembled unit, same run | Your office, out of sunlight, with the approval sheet |
| Inspection copy | One assembled unit, same run | With whoever inspects your cargo, or held as the neutral copy |
| Packaging copy | One printed carton, flat, unassembled | Same three places, or photographed and filed with the PO |
The fourth object is the one people forget, and printed packaging drifts faster than product. A carton gets re-plated when a colour goes out of tolerance or a packaging supplier changes, and nobody mentions it, because from the factory's side the product itself did not change. If you are ordering a printed colour box, that is a 500-piece minimum on this range, and it is worth sealing the box at the same time as the scooter.
The moment matters as much as the count. Seal after first-article approval and before the packing line starts. Sealing at the end of the run means the first two thousand pieces were built with no physical reference at all, which is the same as having no sample — and the cost of three units at FOB is a rounding error against a container of 4,600.
What the label has to say
A sealed sample without a label is an object. A sealed sample with a complete label is evidence — something a third party can read in two years and know exactly what was agreed. Print the label, do not handwrite it, and attach one copy to the product and one to the bag.
| Field | Why it has to be there |
|---|---|
| Factory model code | QY-6619, not the marketing name. One marketing name can map to two factory codes, and the wrong one gets made. |
| Colour code and process | A Pantone or RAL reference plus "powder coat" or "paint". "Blue" is not a specification and cannot be argued about. |
| Tube wall, as measured | Write the number you measured with a caliper, not the nominal on the drawing. This is the field that decides most disputes. |
| Frame material and finish | Steel or aluminium, and the coating system — it sets how the colour ages in a hot warehouse. |
| Wheel and bearing spec | Diameter, PU hardness, bearing type. The bearing is a bought-in part and is therefore a substitution risk. |
| Batch or lot number | Ties the sealed unit to a specific production run, which is what makes a later comparison possible. |
| Date and quantity of the run | Ageing is real: grips harden and coatings fade. A sample with no date cannot be retired on time. |
| PO or approval reference | Connects the object to the paperwork, so the clause in the order has something to point at. |
| Revision number | GS-6619-Rev1. When anything changes you issue Rev2 and retire Rev1 — you never amend the old label. |
| Factory QC signature and stamp | One named person, not a department. A stamp without a name is not traceable to anyone. |
| Buyer or agent signature | The second signature is what turns an internal reference into an agreement. |
Both signatures go across the sealing tape, not beside it. If the tape is cut and re-taped the signature breaks, and that is the entire point. A tamper-evident bag with a serial number does the same job for less money; write the serial number into the PO so it is on the record.
The five places a scooter quietly gets cheaper
Cost reduction on this product is almost never announced. It arrives as a substitute material, a thinner coil, or a different supplier for a bought-in part — and every one of them looks identical in a photograph. These are the five to check on every reorder, in the order they usually happen.
| Point | What changes | How you detect it |
|---|---|---|
| Tube wall | 1.2 mm to 1.0 mm on the deck tube or the T-bar | Caliper, three marked points, same places on both samples |
| Weld | Narrower bead, more spatter, less penetration at the joint | Lamp under the deck, compare bead width against the sealed unit |
| Wheel and bearing | Softer PU, plain bearing instead of a cartridge bearing | Spin it and listen; measure hub width; roll both units across the floor |
| Grip and brake compound | TPR replaced by PVC | Squeeze it and wet it. TPR is warmer and softer, and it grips when wet |
| Fasteners | Stainless swapped for zinc-plated | A magnet, then a week outdoors on one bolt |
The wall thickness is first on that list for a reason. On a frame whose tube weight is around a kilogram — a fair assumption for this range, not a figure on our specification — taking 1.2 mm to 1.0 mm removes a sixth of the steel in the tube, and the product still passes every check an inspector can run without a caliper.
Weigh the carton, it is the cheapest detector you have
Every carton on the scooter range holds 10 pieces and weighs 20 to 23 kg gross. That 3 kg band is not noise: a three-wheel model with a wider deck and a rear brake genuinely carries more steel than a two-wheel starter model, and the spread across the 24 shells is real. It is also your alarm system. Write the gross weight of each carton in your first shipment onto the approval sheet, and weigh one carton from each later lot on a freight scale.
The arithmetic is worth doing once. If your first carton weighed 22.4 kg and the reorder arrives at 20.1 kg, the difference is 2.3 kg across ten pieces — 230 g a scooter, on a product whose net weight is barely 2 kg once the carton and dunnage are taken out. A tenth of the item cannot disappear without something being thinner, and there are only five things it can be.
One caution, because weight is evidence rather than proof: a heavier carton board, an extra manual or a different dunnage pattern moves the figure by a few hundred grams in the other direction. The scale tells you which cartons to open, not what is wrong inside them. Pair it with the caliper and the answer arrives in ten minutes.
A reorder comparison you can run in forty minutes
- Unseal with both sides present, or on video. Unsealing alone destroys the value of the seal.
- Weigh the new carton and compare it with the figure on the approval sheet, before anything else is unpacked.
- Caliper the three points you marked with a paint pen on the sealed unit — deck tube, T-bar, and one leg — so the measurement is repeatable rather than approximate.
- Photograph five points side by side with the sealed unit: weld, hinge, hub, deck print, grip. A phone at 30 cm with the lamp behind you is enough, as long as both units are in frame and lit the same way.
- Compare colour against the Pantone chip, not against the sealed sample. Coatings yellow, and your memory of a colour from two years ago is worse than the chip.
- Put both units on the floor and push them ten metres. Sound and roll rate tell you more about bearings than any micrometer.
The whole kit is a digital caliper, a freight scale, a chip book and a phone — less than the FOB value of one scooter. The reason to spend forty minutes on it is not the forty minutes. It is that a deviation caught at the reorder stage is a conversation, while the same deviation caught by your customer is a return.
The clause that makes the sample count
None of this survives a purchase order that does not mention it. The golden sample has to be an attachment to the PO, with a clause stating that production conforms to golden sample number X, revision Y, sealed and signed on date Z, and that any deviation requires written approval before shipment. Without that sentence the sample is a nice object in a bag.
The useful part of the clause is the timing. A deviation found before loading is handled before loading: rework it, substitute it, or ship it with an agreed adjustment and the adjustment recorded on the invoice. That is a commercial decision taken while the balance is still unpaid. The same deviation found after arrival is a different and much worse conversation. The inspection that surfaces it is an AQL inspection at ISO 2859-1 General Inspection Level II, AQL 2.5 for major and 4.0 for minor defects, critical zero — but an inspector working from a spec sheet will not find a 0.2 mm difference in wall thickness unless the wall is on the sheet, which is the whole argument for a measured label.
Two housekeeping points that get skipped. Samples have a shelf life: PVC grips harden, powder coat fades, and a sample that has sat in a sunny warehouse for two years is no longer a reference. Replace it at twelve months or after five shipments, whichever comes first. And when the model itself changes, issue a new revision rather than editing the old label — an amended label is how two people end up holding different documents with the same number.
Where this fits
If the sample has not been ordered yet, the sample order page has the three lead-time bands: 5 to 7 days for a stock model in branded packaging, 10 to 15 for neutral, 30 to 40 for tooling or a printed colour box. The twelve-point checklist is what to run at inspection, and the golden sample is what makes a point-by-point comparison possible at all. If you are still deciding how much of the product should carry your brand, the OEM versus ODM guide sets out the four customisation levels and their minimums — 50 pieces for a silkscreen logo, 200 for a sticker colour change, 500 for a printed carton, 1,000 for tooling. Every one of those changes needs its own sealed sample and its own revision number, which is why the branding decision and the sealing decision are made in the same week.
On the tricycle range the same discipline applies to SKD packing, where a sealed component set is the only way to prove that the carton you approved still contains what it contained. And on any order, the number of points you can check is set by the questions you asked before the deposit — a supplier who will not put wall thickness on a label in writing will not put it in a bulk run either.
VIKUNO is a sourcing partner, not a factory. The sealing happens at our partner factories in Pingxiang with our QC present, and the buyer copy travels with the approval sheet and the photographs. If a factory will only let you seal a sample-room piece — never a production piece — that is information about how the bulk order will go, and it is better to have it before the deposit than after the container.
Sealing a model before the bulk run? Send the model code, the order quantity and your planned loading week to sales@vikuno.com. We will come back with the label template we use and the five-point comparison sheet, filled in for your model.