What this guide is, and what it is not
There is no B2B or manufacturing client on this site. Nothing below is a case study in your sector, and if you want one you should ask a studio that has one.
What this is instead is a method, and the part of it that comes from work rather than from reading comes from a large consumer catalogue: Beds.ie, 17,000 products, where the job was structure, product-data enrichment and a feed that reports its own failures. Between April and August 2026 that took the site from 5,189 distinct ranking queries to 24,054 — not because five times the content was written, but because the pages already existed and the work made them legible.
A manufacturer with 400 product lines has the same problem at a smaller scale. The checkout is missing and the sales cycle is twenty times longer, but the structural fault and the fix are the same. That is the honest extent of the transfer, and the rest of this page is written as a method rather than as proof.
The shortlist is the real process
A consumer site is read once, by one person, who decides. A B2B site is read several times over weeks, by somebody who is not the budget holder, and then forwarded to people who never visited it at all.
That changes what the page has to do. It has to survive being read carefully by a specifier looking for a reason to rule you out, and then it has to survive being read cold by a finance or operations person who sees only the page they were sent and none of the context.
Most B2B sites fail the second test badly. The page that gets forwarded is usually a product page, and it usually assumes the reader has already read the homepage, already knows what the company does, and already knows why this product rather than the one above it. Written that way it reads, to the person it was forwarded to, as a page about nothing in particular.
The fix is unglamorous: every page that a buyer might reasonably send to a colleague should name what the thing is, what it is for, what it costs or what band it sits in, and who it is not for. A page that does that is also, not coincidentally, a page that ranks.
Turning a catalogue into pages
If your range lives in a PDF, a downloadable price list, or a single long page with a table on it, that is almost certainly the biggest thing wrong with the site, and it is fixable without a redesign.
The work, in the order it is usually worth doing:
- Decide what the unit is. One page per product line, per capability, or per material and process — whichever is the thing a buyer searches for and compares. Getting this wrong by one level is the commonest planning error: a page per category is too coarse to rank and too vague to forward, a page per SKU variant is a thousand near-identical pages that compete with each other.
- Get the specification onto the page as text. Tolerances, dimensions, materials, lead times, certifications, compatibility. A number inside a PDF or an image is invisible to search and unselectable by the person trying to paste it into a comparison.
- Keep the download, but stop making it the gate. A data sheet a buyer can have without filling in a form is worth more than a lead form nobody completes, because the buyer who downloads it is the one who shortlists you.
- Link the range to itself. Alternatives, the next size up, what it replaces, what it is used with. Internal links are how a crawler finds depth and how a buyer navigates a range they do not yet understand.
- Name things the way buyers do, not the way your ERP does. Internal part naming is the most reliable way to build a catalogue nobody can find.
None of that is a design exercise. On Beds.ie the visible design barely changed; the product data and the structure did, and that is where the fivefold rise in ranking queries came from.
Offer a first step smaller than a phone call
The second structural fault is a site whose only route is a contact form, for a purchase with a six-month cycle.
A buyer three weeks into research is not ready to be sold to and knows that a phone call will try. What they want is a specification, a price band, a lead time, a drawing, a compatibility table, or a sample. Every one of those is a smaller commitment than a call, and every one identifies a real buyer better than a form fill does.
Publishing a price band is the one that causes arguments internally, and it is usually worth winning. The objection is that every job is bespoke; the reality is that a buyer qualifying five suppliers will drop the two who would not indicate anything, because they have to put a number in a spreadsheet by Friday. A band with its assumptions stated — "from £X for a run of Y, excluding tooling" — costs you nothing and keeps you in the comparison.
And a long sales cycle is an argument for search, not against it. Six months of research is several searches per buyer; being present at the third one is how you end up on the shortlist rather than in a final bid you were never going to win.
Our own data, and why it is not your market
This is worth being blunt about, because it is the thing most agency pages blur.
Those are the terms a studio selling to you competes for. They are not your market, and anybody who shows you that table as though it were your opportunity is misreading their own research.
What the table is good for is one inference about your market, by analogy. Technical B2B niches are very often like that too: competitors buy the clicks and nobody writes the specification pages. That is a gap you can test cheaply.
What to check on your own site
- Is every product line or capability its own URL, or does the range live in a PDF, a table or a filter state?
- Is the specification on the page as selectable text?
- Pick the page a buyer would most likely forward to a colleague. Read it as somebody who has seen no other page on the site. Does it say what the thing is and who it is for?
- Is there anything a buyer can take away without talking to anybody?
- Is there a price, a band, or a stated reason there is not one?
- Do your product pages link to the alternatives and the adjacent sizes, or only up to the category?
- Are things named as buyers say them, or as your part numbering says them?
- Do you know what your buyers actually search for, measured, or is that an assumption?
The first and the last of those are where the money is. Everything in between is tidying.
