Quality control for stone orders: agreeing acceptance before cutting

This page answers a trust question by talking about timing and restraint. A first order goes wrong less often because a stone was badly chosen than because nobody wrote down what acceptable meant before cutting began. So the account below starts at the moment criteria are agreed, not at the moment goods arrive. It also explains a deliberate refusal: we do not publish a measured value on these pages unless a source document stands behind it. Read the sections as a description of when decisions are made and what can be checked, rather than as a set of test results.

Why acceptance criteria belong in the brief before cutting, not at delivery

On 2026-11-13 Ningbo Rongcheng Stone recorded the rule that governs this page: acceptance criteria are written into the brief before the first piece is cut, and no figure is published without the document that established it. An acceptance criterion is only useful if it exists before there is something to accept. Once a block has been cut into finished pieces, the range of possible corrections narrows sharply: you can re-make a component, but you cannot turn back the decision that allocated a particular part of the block to it. For that reason we ask that the criteria be written into the brief while the drawing and the reserved material are still being agreed. Tone range, the treatment of visible veins, the surface character, the condition of edges and the way pieces are identified are all things two people can reasonably understand differently until they are put into words. Agreeing them early is not bureaucracy; it is the only point at which agreeing them still changes the outcome. At delivery the same words become a basis for discussion rather than a basis for making.

The check areas: visual consistency, dimensions, finish, edges, numbering

When a set is reviewed before it leaves, the review follows the criteria already written down, and it moves across a small number of recurring areas. The eye looks at tone and at how adjacent pieces will read together. Each component is measured against the outline on the drawing. The surface is examined for the character that was approved on the sample. Edges and returns are checked where they are most exposed to handling. And the identifier on each piece is read back against the schedule so that the set is complete and none of its positions is duplicated. What is being tested here is conformance to an agreed description, not an abstract idea of quality. A piece can be within its outline and still be wrong for the position it carries, which is why the drawing and the mark are part of the check rather than an afterthought.

How photographic records and piece numbering support a shared reference

Records are most valuable when they can be tied to a specific instruction. A visual record of a finished set, read alongside the identifiers on the pieces, lets a buyer who is not standing in the works refer to exactly the component being discussed. This is the same logic that makes numbering useful during fabrication: the mark is a stable handle, and a photograph attached to that mark travels further than a verbal description of colour or sheen. Where a project asks for pre-dispatch records, we agree the form at enquiry stage so the right images are captured while the set is still laid out. We describe these as records prepared against the agreed requirement, not as a substitute for the buyer's own inspection on arrival. Their purpose is to give both sides one object to point at.

Project-specific test documentation requested at enquiry, prepared before dispatch

Some projects need documents that a general order does not: a declaration, a record of a particular test, or a statement tied to a destination's import rules. These differ from the qualitative descriptions on this site because they must come from an actual source and be correct for a specific consignment. The time to say so is at enquiry, not at the loading bay. When a requirement is stated early, it can be built into the plan for the order and the correct record can be prepared while the goods are still in production. When it appears late, the choice is between delaying the shipment and sending something that was assembled in haste. We prefer the first conversation. We also decline to name an authority, a report reference or a measured result on a public page without the document that supports it.

Why this site does not publish approximate figures as if they were specifications

A number on a web page looks authoritative whether or not anyone can defend it. For a material that varies from block to block, printing an approximate measured value and treating it as a property of a whole category would be actively misleading: a buyer might build a specification around a figure that no document for their order supports. We therefore keep these pages to qualitative descriptions of process and behaviour. When a project needs a verified figure, we provide it for that order from the document that established it, and it travels with that consignment rather than becoming a public claim. This is also why each structured fact on the site carries a source identifier and a date: the reader can see that a statement is a described practice, not an invented measurement. A figure we cannot point to the source for stays off the page.

What to put in the enquiry so the right records can be prepared

The enquiry is where the inspection plan is set, so it is worth being specific about three things. Describe the acceptance criteria you intend to apply to tone, surface and edges, or tell us that you would like to agree them on an approved sample before cutting. Say which records the project will need at import or at handover, and name the destination, because that often determines which paperwork is relevant at all. Mention whether pre-dispatch visual records are expected and in what form. We then restate those points in the written reply, so that what will be checked and what will be issued are visible to both sides before a schedule is released. None of this requires a measured value up front; it requires that the requirements be stated while they can still shape the work.

Custody of the approved stone sample: the written record a buyer should require

An approved sample that is returned, mislaid, or quietly swapped removes the one object against which a whole order was meant to be judged. If that reference disappears between sign-off and delivery, each side holds a different memory of the agreed tone, and no inspection can resolve it. A short custody record should state which sample was endorsed, who keeps it, where it is held, and how it is marked against later alternatives. Written at approval and countersigned by both sides, it stays with the order file rather than a photograph. It also says what happens if a second sample appears: the new one is a candidate, not a replacement reference.

Ordering the supplier check, the buyer inspection at the works and the check on arrival

Acceptance of a stone order can involve three separate inspections, and confusion follows when nobody records which one carries the authority to accept the set. The supplier review is carried out first, against the criteria written into the brief, and its outcome is logged against the identifiers on the pieces. A buyer or an appointed inspector may then review the laid-out set at the works before crating, with any comment entered as a written note tied to specific marks rather than as a general impression. The check on arrival is a separate event again, and it is not a substitute for either of the earlier ones. Each stage is signed by the party that performed it, and the sequence is set into the brief before cutting.

Writing a stone non-conformance so it cites an agreed criterion rather than an opinion

A complaint phrased as a dislike gives the supplier nothing to check against, whereas a non-conformance tied to a written criterion can be answered with a record. The buyer should describe the specific identifier, name the criterion from the brief that the piece fails, and state the condition observed in terms two people could see the same way, such as a tone falling outside the range agreed on the approved sample or an edge unlike the endorsed character. The note is then entered against the order file and answered in writing by the supplier, so the exchange becomes a traceable decision rather than a string of impressions. Requiring this form at the outset also discourages the late invention of a standard never agreed before cutting.

Re-approval of a substitute stone block or a re-cut against the original criterion

A re-cut is not acceptable merely because it looks better than the piece it replaced; the baseline is the criterion agreed before the first cut, not a fresh judgment at the bench. Should a substitute block or a re-made component appear, the buyer ought to require written confirmation that it was reviewed against the original description and the retained sample, not approved in isolation. Otherwise the reference quietly shifts, and the rest of the set is left judged by a standard that no longer matches the one used for it. The order brief ought to name who performs this re-approval and how the result is recorded. Accepting a replacement on a verbal assurance dissolves the very criterion the inspection plan was built to protect.