Samples, and the distance between one good unit and a good batch

The stages, and what each one is capable of proving

A catalogue or reference sample shows what the plant already makes and settles whether the base article is close enough to be worth pursuing. A development sample tests your modifications and usually arrives hand-built. A pre-production sample should come off the intended tooling, in the intended material, by the intended method, and is the one worth arguing over. A sample drawn from the actual run confirms the run matches what was approved. Retained units held by both parties become the reference for any later dispute. Skipping the pre-production stage is the most common and most expensive economy in the sequence.

Why a hand-made unit flatters the product

Samples are often produced by the most capable person in the building, off the line, from material chosen for the purpose, with defects quietly corrected before packing. Nothing about that predicts what happens when the article is made repeatedly at rate by whoever is on shift. Ask explicitly how the sample was made: on production tooling or a prototype tool, in production material or a substitute, by the line or by the development room. Where the answer reveals a difference, the sample tells you about the design and almost nothing about manufacturability, and the gap will appear in the run.

Approval is an act that should leave a record

Approving a sample verbally, or by replying to a message with a single word, creates nothing you can rely on later. Seal and label the approved unit, keep one and leave an identical one at the plant, and record what was approved, by whom, against which written specification, and what remained outstanding. Note deliberately accepted deviations rather than letting them pass silently, because silence becomes precedent. If a colour, a finish or a fit was accepted grudgingly, say so in writing, since a supplier reasonably assumes anything not raised is acceptable at that level.

Test the sample against the life the product will have

Assess the article the way a customer will meet it, not the way it looks on a desk. Ship samples through the parcel network in their intended packaging and see what arrives. Expose them to heat, cold, moisture, sunlight or vibration where the route or the use demands it. Assemble the product following only the supplied instructions. Where the class carries testing obligations, submit samples to a competent laboratory rather than judging by eye. Faults found at this stage cost a sample round; the same faults found after a run cost the run, the packaging and the reputation.

Rounds consume the calendar, so make each one count

Each iteration costs sample charges, courier time and a slot in the plant's development queue, and the elapsed time is frequently the largest component of a launch schedule. Compress the sequence by giving complete, specific feedback in one document rather than a stream of separate messages, marking up photographs, quantifying nothing you cannot measure and describing what you want instead of what you dislike. Decide beforehand which attributes are non-negotiable and which are preferences. Endless refinement is its own failure mode: at some point the remaining differences cost more in delay than they add in the hand of a customer.

Frequently asked questions

How many rounds of samples should I expect?

Fewer than most people fear if the brief is precise, and endless if it is not. A catalogue article with only labelling changes may need one confirmation round. A modified product typically needs a development round, a correction round and a pre-production unit off real tooling. Anything genuinely new runs longer. The variable you control is the quality of feedback, since vague comments produce a sample that misses in a new direction and consume another cycle of the plant's queue.

Should I be paying for samples?

Usually yes, and paying is often the better position. Free samples come from suppliers treating the exercise as sales cost, which tends to mean less care and no obligation. Paying, including for the courier, establishes a commercial relationship and makes it reasonable to specify how the sample must be produced. Many suppliers credit sample charges against a subsequent order. What matters more than the fee is whether the sample was made under production conditions, because that determines what it proves.

What do I do when the run does not match the approved sample?

Act before the balance is paid and before the goods move, since your bargaining position disappears at both points. Compare the incoming units against your retained sealed sample and the written specification, photograph the difference, and quantify how much of the run is affected. Then decide what you want: rework, replacement, a price adjustment, or acceptance of the deviation. Raising it clearly and early is usually productive; raising it after the container has been unloaded and sold from rarely is.