AI visibility for UK manufacturers: publish the conformity paperwork and the capability data as text
Measured on 15 October 2026 across 204 UK readings of ChatGPT and Gemini: the standard references, the UKCA and CE position and the lead time data a British industrial buyer filters on.
Published by AI Knows Us · Updated 16 October 2026
What puts a UK manufacturer or industrial supplier into an assistant's answer is paperwork the business has already produced, published as text on the website instead of locked in a PDF. The buyer is not searching for a brand. They are searching with a standard reference, a grade, a part code, a process and a tolerance, and they are checking whether a supplier can legally and technically supply the thing. So: the UKCA and CE marking position written in words, the Declaration of Conformity as a readable page, the standard each product is tested to named by its reference, the quality certificate shown with its certifying body and scope, and the process, material and lead time data a buyer filters on. Our 15 October 2026 benchmark asked ChatGPT 102 buying questions and asked Gemini the same 102, which is 204 UK readings, and exactly one of them cited our own domain. ChatGPT produced that one. Gemini produced none.
What the run measured, and the boundary of it
Those questions were buying questions in a single category, AI visibility, asked in three markets. No separate measurement was run for manufacturing, so the dataset contains no industrial citation rate and nobody should quote one. What it gives an engineering business is the source mix behind an answer, which is the part that changes where the next quarter of effort goes.
| Figure | Value |
|---|---|
| Date run | 15 October 2026 |
| Engines | ChatGPT and Gemini |
| Usable readings | 597 of 612 attempted |
| UK readings | 204 |
| UK readings citing our own domain | 1 |
| Citations recorded | 10,652 across 2,708 distinct domains |
| Supplier's or agency's own site | 85.6% |
| Official or platform documentation | 9.7% |
| Review directories | 0.8% |
Our own product reads five surfaces, ChatGPT, Gemini, Perplexity, Claude and Google AI Overviews. This dataset covers two of them on one date, and the method and market comparison are in what AI assistants actually cite across 597 readings of ChatGPT and Gemini in three markets. The 85.6% is the operative figure for this trade: the pages being cited are suppliers' own pages, so the asset to build is your own catalogue in text.
What a British industrial buyer actually puts in the box
- A standard reference. The BS or BS EN reference for the material grade or the product type, often with a condition or class after it. The buyer has taken it straight off a drawing or a specification.
- A part or drawing code. Their own code, an original equipment manufacturer's code, or a competitor's code they are trying to cross reference.
- A process with a constraint. The process, the material, the tolerance to be held, the finish, and the quantity.
- A compliance question. Whether a product can be placed on the market in Great Britain, in Northern Ireland or in the European Union, and what marking it needs for each.
Every one of those is a filter. A specific published fact passes the filter. A vague sentence fails it silently, because a sourcing engineer building a shortlist will never ring to find out.
Publish the conformity paperwork as a page, not an attachment
This is the largest unworked opportunity on a British manufacturer's website. The documents already exist, because they have to, and they are almost always a scanned attachment or a login protected download. A fetcher cannot read either.
| Document | What it proves to a buyer | Where it usually hides |
|---|---|---|
| UKCA and CE marking position | Which market the product may be placed on, and under which rules | One line in a brochure PDF |
| Declaration of Conformity | The product, the legislation applied and the standards used | A scanned signature on a download page |
| Declaration of Performance, for construction products | The declared performance against the harmonised specification | A separate documents portal |
| Quality management certificate | The certifying body, the accreditation behind it, and the scope of what is covered | A logo in the footer |
| Material test certificates and traceability | That the batch is what the paperwork says it is | Issued with the delivery, never described on the site |
| Substance and material compliance statements | That the product can be used in the buyer's application | An email on request |
Turn each row into a page that says, in text, what the document covers, which products it applies to, who issued it, when it was last reviewed, and how to request the signed copy. Keep the signed PDF as well. The page is what gets read and quoted; the PDF is what gets filed.
A certificate without its scope is a badge
Quality claims are where British suppliers lose credibility fastest, and the fix is one sentence. Name the standard, the certifying body, whether that body is accredited by the United Kingdom Accreditation Service, the certificate number if you publish it, and above all the scope. A quality certificate covering machining at one site does not cover fabrication at another, buyers know this, and the supplier who states the scope is the one who looks competent rather than evasive.
Capability data is the catalogue, and the catalogue is the content
Publish the data a buyer filters on, each item as text on a page that can be read on its own.
- Processes by their trade names, with the machine envelope and the capacity per process.
- Materials by grade, with the standard reference, and the forms you hold or can source.
- Tolerances as the range you routinely hold on each process, not as a claim about precision.
- Finishing and treatment done in house against subcontracted, named either way.
- Lead times in working days, from enquiry, from order and from tooling sign off, with the figure reviewed on a stated schedule.
- Minimum order quantity stated honestly per process, including where there is no minimum for a repeat customer.
- Commercial terms in pounds: whether a price is ex works or delivered, the Incoterm where you export, carriage, payment terms, and whether a figure excludes VAT.
The Northern Ireland position deserves its own sentence rather than a footnote, because the marking a product needs there is a genuine buyer question and the supplier who answers it plainly is answering something competitors leave ambiguous.
The platform question, and where enquiries actually come from
Industrial directories and sourcing platforms are useful, and they are not the lever the trade assumes. Review directories took 0.8% of the 10,652 citations recorded on 15 October 2026, while 85.6% went to a supplier's or vendor's own pages. Keep the listings accurate and current, then spend the real effort on your own grade, process and part pages. How the various lead services differ, and which one matches an industrial sales cycle, is in the four things sold under one phrase. The same catalogue argument written for a different market, where the dominant platforms are different, is in part numbers, standards and the platforms that outrank you.
Questions about AI visibility for UK manufacturers and industrial suppliers
Does publishing tolerances and lead times help competitors?
Competitors already know what your machines can do. Buyers do not, and they are the ones using the figure as a filter. The supplier who publishes the range wins the enquiry the vague supplier never hears about.
Should the Declaration of Conformity really be a web page?
Publish a readable page describing it and keep the signed document available. The page answers the buyer's question and can be quoted; the signed copy is what goes in their file.
We sell through distributors. Is this still our job?
Yes, because the specification question is asked before anyone chooses where to buy. Publish the technical pages and name the distributors who stock the product, which makes your page the one with the complete answer.
Is a sourcing platform listing enough on its own?
It is a useful entry and a thin one. On 15 October 2026, 85.6% of the 10,652 citations recorded went to somebody's own website, so the listing supports the work rather than replacing it.
How do we measure whether this changed anything?
Write down the ten enquiries you most want, as a buyer would type them, with no company names in them. Ask each twice on ChatGPT and twice on Gemini, record who is named and what is cited, publish the technical pages, and ask again a month later using the same wording.
The paperwork is already written
Almost everything an assistant needs from a British industrial supplier has already been produced, signed and filed: the declarations, the certificates, the test data, the capability list a sales engineer recites on the phone. It is simply not on the website in text. Converting it is a quarter of patient, unspectacular work with a measurable outcome, because the enquiries it produces arrive with a specification already attached. We sell that conversion, so the fair disclosure is our own result: one citation of our domain in 204 UK readings on 15 October 2026. AI Knows Us publishes that figure because a supplier's trade runs on verifiable claims and so should ours.