The sample hierarchy: what each sample proves

At a glance

Numbers match the steps on the drawing. These are working terms: makers use the words differently, so agree the name of each step in writing.

A sample is evidence; a signature makes it a benchmark

People use the word sample for very different things: a swatch to choose from, a finished board to match, a cut-through to show a joint, the first piece off a run. They prove different things, and only one of them is meant to bind production. The drawing puts them in the order they usually come.

The working rule is short. A finish sample or a construction sample becomes a control sample when it is signed and dated, given a reference, and a matching copy is kept by the maker. From then on, production is measured against it, not against anyone’s memory of it. Everything after the control sample, the mock-up, the first article and the production check, is a check against that benchmark.

Words differ between firms. What one workshop calls a pre-production sample another calls a first-off, and a prototype is sometimes called a mock-up. Ask each maker what they call each step, and write the name you both mean into the approval note.

How much of the ladder an item needs

Not every item needs seven steps. Match the effort to what is at stake:

• Simple construction in a finish already seen: a signed control sample, then the production check, is usually enough.
• Natural materials (veneer, solid timber, leather, stone): add a range, not a single point. See the section on natural materials below.
• Complex, repeated or high-value pieces: add a mock-up or prototype, then a first article, and release the rest of the run only after the first article is approved. For repeated hotel rooms this is the model-room idea: approve one room, then roll out.
• Anything where comfort or fit is the question: seating, banquettes and desks are judged on the mock-up, not on the finish sample.

Timing and handling. Samples take workshop time and material, so agree early which will be made, in what order, who signs, and how they are handled alongside the commission. On larger schemes it is common to produce them first. On smaller ones they usually follow once the scope and the quotation are settled.

Approve under stated conditions

A sample approved under one light at a desk can look different on site. The cause is metamerism: two surfaces that match under one light source can differ under another. NCS Colour’s own explainer names three standard sources, daylight, incandescent and fluorescent, and recommends checking colour under at least daylight and fluorescent light. Rooms now use other lamps again, so the practical step is to name the lights you check under and to include the lamp type the room will actually have.

Write the conditions on the approval record, so that the person who inspects the finished work can recreate them:

• Light. Name the source or sources. Where possible view in daylight with wall lights and downlights off, and avoid torches or site lamps, which create a lighting scenario the room will never have. That is the approach Stone Federation GB takes for stone, drawing on NHBC guidance for paint finishes.
• Distance and viewing spot. State them. For stone samples, the standards cited by Stone Federation GB call for final viewing from at least 2 m, in light and orientation like the project’s. We found no equivalent figure for joinery or furniture, so choose a distance that suits how the item is used: a handrail is judged at arm’s length, a wall panel from across the room.
• Orientation. As it will be fitted. A gloss finish reads differently on a wall than on a flat top.
• Size. Large enough to judge. A swatch does not show how a large flat surface will look.
• Material. The real species, cut and board. A finish sample on another board proves little.

Natural materials: approve a range, not a point

Veneer, solid timber, leather and stone vary from leaf to leaf, board to board and hide to hide. One sample cannot show that. If the control sample is a single perfect piece, every real piece looks like a failure. If there is no sample at all, nobody can say what failed. Three habits help:

• Show the range. Mark the lightest, darkest and most figured pieces you will accept, with one unacceptable example if it helps, and label the set as the range.
• Write the accepted features down. Stone Federation GB describes a range report as a project document, often written into the specification, that records the expected appearance of a stone with the features accepted and rejected. It is then the benchmark for the delivered material and the installed work. The idea carries over to veneer and timber.
• Keep the same people. Stone Federation GB recommends that whoever approves the material at design stage also takes part in the final inspection, so that the benchmark is not reinterpreted late.

For veneer, approve the layout drawing and a leaf from the actual flitch. The joinery specification sheet gives the sample references (S-01 to S-05) that fix species, finish, veneer, construction and hardware.

Label, sign, keep and change

A sample that is not labelled and signed is a souvenir. Keep a short register, and give each entry the same reference that appears in the specification and on the FF&E schedule, where the “Sample ref” and “Sample status” columns point back to it.

Once the work is made, the control sample is also the yardstick for inspection: see the sheet on snagging and acceptance criteria.

Each sample is given a reference [S-nn] and labelled with the project, item, material and finish, date, and the drawing and revision it relates to. Approval is recorded by signature and date of [designer] and [maker], and of [client] for finishes. A matching copy is held by each party until [the work is accepted]. Front and back photographs, with a scale, are filed with the approval.

Production shall match the approved control sample [S-nn]. Natural variation within the range shown by [range samples S-nn to S-nn / the range report dated [date]] is accepted; features outside it are not. The control sample is not changed, replaced or re-approved without the written agreement of [designer]. A change of material batch, supplier, process or finish after approval requires a new sample and written approval before use.

Samples are approved by viewing from [distance] m, in [daylight] and under [lamp type of the room], at the orientation they will have when installed [vertical / horizontal]. A second check is made under [second light source]. Conditions are written on the approval record.

Before production of the remaining [number] [items], the maker presents [a full-size mock-up of [element] / the first article] for inspection against drawing [number, revision] and control sample [S-nn]. Production of the remainder starts only after written approval by [designer], who responds within [number] working days of presentation.

Standards and sources

• Stone Federation GB: Inspection and snagging of finished internal works using natural stone (August 2022): read 5 Nov 2026. It cites BS EN 12057, 12058 and 1469 for the 2 m stone viewing distance; not read here.
• NCS Colour: The phenomenon of metamerism: read 5 Nov 2026: names three light sources and recommends checking under at least daylight and fluorescent light.
• Make Bespoke Studio: Sampling: how the studio works with option samples and control samples.

Questions we are asked

What is the difference between a control sample and a finish sample?

A finish sample shows how a surface will look. A control sample is a finish sample or a construction sample that has been approved, signed and dated, with a matching copy kept by the maker, so that production can be judged against it. The signature and the second copy are what make it a benchmark.

What is a first article?

The first finished piece from a production run, built by the production process and inspected against the drawings and the control sample before the rest are made or released. It is the production counterpart to a prototype.

What is the difference between a mock-up and a prototype?

The words overlap. Usually a prototype is a working full-size example of one item, and a mock-up is a full-size or partial representation of an element or a room, built to judge appearance and fit. Ask the maker what they mean and write it into the approval note.

Why does a sample look different on site?

Usually because of metamerism: two surfaces that match under one light source can look different under another. Approve samples under named lights, including the lamp type the room will have, and write the conditions on the approval record.

Who should sign a sample, and who keeps it?

The designer and the maker as a minimum, and the client where they approve finishes. Each party keeps a matching copy, and the maker’s copy stays where the work is made, so that production and inspection use the same benchmark.

Does an approved sample cover natural variation?

Only if the variation is shown. For veneer, timber, leather and stone, approve a range as well as a typical piece, or write the accepted features into a range report, so the inspector knows what is natural and what is a defect.