What a sample actually proves.

Five stages.

1. 01

   Say what the sample has to prove

   A sample is a test, and a test needs a question. Colour and finish, or moisture and volatile oil, or whether the joint survives a week in a dry room. Write the question down before you ask for material, because it decides how much you need and what you will do with it.

2. 02

   Ask for production material, not a showpiece

   The phrase that works is a sample cut from current production, sent as it comes off the line. What arrives otherwise is the best piece somebody could find, which tells you what the factory is capable of on its best day and nothing about your order.

3. 03

   Seal and retain a duplicate

   Two identical samples, one to you and one sealed and signed at the producer. The retained one is the reference an inspection compares production against, and it turns a later argument about quality from an opinion into a comparison.

4. 04

   Send it to a laboratory where numbers decide

   For anything contracted on a figure, patchouli alcohol, coumarin, iodine number, histamine, the sample is worth testing independently. It costs very little against a container and it establishes what the producer's own analysis should look like.

5. 05

   Order a pre-production sample separately

   For made goods, the sample that matters is the one produced after the order is agreed and before the run starts, made to your specification with your material. Approving it in writing is the last cheap moment in the process.

The retained sample.

This is the part almost every first-time buyer skips and almost every experienced one insists on. Two identical samples are prepared. One travels to you. The other is sealed, signed and dated at the producer, and it stays there.

Months later, when a container arrives and something is wrong, the argument is normally about words. The producer says the goods match the specification. You say they do not. With a retained sample there is a physical object both sides agreed to, and an inspector can hold the production against it. It costs a second sample and a signature.

What a pre-shipment inspection covers →

Four ways it goes wrong.

• The sample that came from a competitor

  A trader with no factory can obtain an excellent sample from somebody who has one. This is the most common reason a first production run looks nothing like the sample: it was never made by the company you are paying.

• Free samples that are not free

  A producer offering unlimited free samples with courier paid is either very large or building the cost into a price you have not seen. Neither is a scandal. It is worth knowing which, because in the second case you are paying for other people's samples too.

• Approving on the wrong attribute

  Buyers approve furniture on finish and then receive furniture that moves, because moisture content was never measured. They approve spice on aroma and receive a lot that fails microbiology. Decide what actually decides acceptance, then evaluate that.

• One sample, one supplier

  Samples are worth most in comparison. Three samples for the same specification from three producers tell you what the specification actually delivers in that market, which no single sample can.

Getting it out of Indonesia.

A small sample usually travels by courier, and the paperwork is lighter than a shipment without being absent. Food, plant material, seeds and essential oils all have their own rules at the destination, and a sample that arrives without them is a sample your customs authority destroys.

Ask what the destination requires before the parcel leaves rather than after it is held. For plant products that usually means a phytosanitary certificate even on a small quantity, and for anything classed as dangerous goods by air it means a declaration and a carrier who accepts it.

Questions buyers ask.