Guides · Materials
The fabric delivered is not the approved sample: is it a defect?
A signed sample and a delivered roll never match entirely. The real question is where variation ends and non-conformity begins.
In short. Fabric and leather are not identical to themselves: they shift with the dye lot, the yarn batch, the hide and the direction of the pile. Where normal variation ends and non-conformity begins is not settled by the eye, but by the document that went with the sample: what was approved, by whom, and with what reservation. Anyone without a signed reference sample is arguing about a difference with nothing to compare it against.
What exactly is «the approved sample»?
Within one project the word sample stands for three different things, and confusing them is where half of these disputes begin.
- The swatch book: the mill’s square or card, chosen at design stage. It settles a direction, not a shade: it is a cutting from a roll produced months earlier.
- The confirmation sample: a piece of the batch actually allocated to the project, sent before cutting. It is the only one that represents what will arrive.
- The upholstered prototype: a complete example of the model in the chosen fabric. It shows what a square cannot — how the material behaves on the shape of the piece.
A useful approval states which of the three it refers to, carries a date and exists in two copies, one per party. The piece kept in the workshop is the reference sample, and its purpose is to be looked at again months later, when nobody remembers the exact shade. In a hotel all these approvals become one thing in the mock-up room.
Why two rolls of the same fabric are not identical
The commonest cause is the dye lot. Fabric is dyed in batches, and two batches of the same article come out in slightly different shades: that is how dyeing works, not a failing on the mill’s part. This is why large supplies ask for a single lot, and why a reorder months later — one more armchair, a replacement — never matches with the confidence of the first cut.
Three further causes sit alongside colour. The yarn batch alters the hand and the way the surface reads. The direction of the pile, in velvets and chenilles, makes the same fabric read as two shades depending on how the light catches it: two seats cut with opposing pile look like two colours, and are not. And a pattern has a repeat: on a small piece only part of the motif shows, and the perception of the ground changes with it.
With leather the matter is plainer: a hide is a natural surface, and variation between hides, and within one hide, is a property of the material. How much of it shows depends on the tannage and the finish: we have set it out in the guide on which leather to choose.
Non-conformity or lot variation: where is the line?
The line is not a universal threshold, which is why these discussions end in negotiation: the order documents, almost always, say nothing that can be measured against. So the first task is to establish what was promised. Three questions, in this order.
- Is it the right article? Name, colour code and mill of the roll, set against the order. If they do not match there is nothing to assess by eye: a different material was delivered.
- Was there a confirmation sample? If the batch was approved before cutting, the reference exists. If only the swatch was approved, a difference in lot falls within what a swatch book never guaranteed.
- Is the difference within the supply, or between supply and sample? A shade that varies between the delivered pieces is the more serious case, because it shows in the room and no document accounts for it.
All of this is decided once, and early: it belongs in the request for supply. How to word a requirement that can be verified is in the guide to the specification clause.
How to look at a colour difference without a quarrel
A difference in shade is assessed in repeatable conditions, or else each party sees what their own light shows them. The method needs no instruments.
- Place sample and supply touching and on the same plane: compared at a distance the exercise means nothing.
- Orient them the same way for pile and pattern.
- Look in diffuse daylight, then in the light of the place the furniture is going to. Two materials that agree by the window can part company under a warm lamp, and for a hotel that is what counts: the room has that light and no other.
- Look with more than one person, at separate moments, and record the outcome with a photograph of the two pieces side by side.
That record is what later supports a decision. If the check happens on site, at delivery, do it before signing: the order of signatures is in the guide on the handover report.
And if the material is already cut?
This is where the confirmation sample earns its place, because it changes the cost of the remedy. If the non-conformity is found on the roll, before cutting, work stops and the mill is asked to replace it. If it is found afterwards, the remedy is to remake the pieces, and cut material does not go back.
That is why the check sits upstream. A workshop that upholsters in house has an option a long chain does not: the roll comes in, it is looked at, and if it is not the right one it is not cut. Holding a fabric in the store costs days; holding twenty upholstered armchairs costs the supply. What follows is in the guide on changing the fabric mid-production.
Why it weighs more in a hotel than in a house
For two reasons that have nothing to do with taste. The first is quantity: in a house one sofa stands alone, in a hotel fifty armchairs stand in the same lounge, and a lot difference that would pass unnoticed on one piece becomes plain along a row. The second is that the remedy touches operation: re-covering a certified upholstered item is not neutral, and we have written why in the guide on re-covering an upholstered item. In an open property it means rooms out of service.
And a dispute about colour stops the supply, while a stopped supply moves the opening: in a hotel project the value that weighs most is time.
How we work
We upholster in house, as we work wood, marble and iron, and that puts the check on the material inside our own workshop. On projects we ask for a sample of the batch allocated to the work before cutting, we keep the reference piece with the date of approval, and we keep the cut of any one item type to a single lot. Matching across time is managed, not promised: when a reorder comes months later we say so in advance.
The materials are in Collections, the course of a project in the Contract section, and contacts lead to the people who follow it.
Frequently asked questions
Does the swatch from the book count as the approved sample?
It holds for the article and for the choice, not for the shade of the batch: it is a cutting from a roll dyed at another time. Where shade matters — and in a lounge with many identical pieces it does — the reference to approve is a sample of the batch allocated to the project, asked for before cutting.
Can a whole supply be refused over a difference in shade?
It depends on what was agreed, which is why it is worth agreeing in writing beforehand. In practice the useful route is another: establish whether the material is the article ordered, whether a confirmation sample existed, and whether the difference shows in the condition of use. The three answers point to a proportionate remedy.
Do the same criteria apply to leather?
The method of comparison, yes; the premise, no. Variation between hides is a property of the material, and a hide finished to look uniform is a product choice, not the natural condition. On a leather supply what is approved is a set of hides, not a square.
Talk to us The contract division
This page describes a supply practice and cites no standards, classes or acceptance values. No prices, no response times, no company case histories. The reference that binds a supply is the approved sample with its date.

