Industrial B2B Website: Turning a Technical Catalogue Into Qualified Enquiries

An industrial buyer does not browse like a consumer. They arrive with a precise part number, a standard, a diameter, a material — and they type that straight into Google. If they cannot find your part, they find someone else's, and the enquiry ends there.

That is what makes an industrial website unusual: its value comes not from design but from whether the catalogue can be indexed, filtered and understood. Most industrial sites fail on exactly that point, by publishing their catalogue as a PDF.

The PDF catalogue: the most expensive mistake

It is the sector's most common reflex: the catalogue already exists as a PDF for print, so it goes online and the job looks done. Except a 200-page PDF is a single URL. Your 3,000 part numbers have no address of their own, no title, no description — nothing to rank on.

Google can read a PDF, but it indexes it as one document. A buyer searching for "EPDM O-ring 25x3" will never land on page 147 of your catalogue. They will land on a competitor who turned that part number into an HTML page.

Structuring a technical catalogue so it can be found

One page per part number, or at minimum one page per family with filterable variants. Each page must carry the characteristics buyers actually search on, as text rather than as an image.

• The exact designation and the synonyms the trade actually uses
• Dimensional characteristics spelled out, not only shown on a drawing
• Standards and certifications (ISO, EN, ATEX, CE) — these are keywords in their own right
• Materials and their equivalences
• Competitor equivalents, where the trade uses them as a reference point
• Drawings and datasheets for download, alongside the text and never instead of it

Faceted filters improve the experience but raise a technical question: every filter combination creates a URL. Without correct canonicals you generate thousands of near-identical pages. The usual rule is to index the filters that match a real search — material, diameter — and canonicalise the rest.

The real industrial buying journey

Industrial buying cycles are long and involve several people. The engineer identifies the need, the design office validates feasibility, the buyer negotiates. Those three people search for different things and land on different pages.

• The engineer searches a part number or a characteristic. They need the product sheet, the drawing, availability.
• The design office searches for a solution to a problem. They need application cases and sizing data.
• The buyer searches for a reliable supplier. They need your certifications, lead times, client references and production capacity.

Converting without a cart: the quote request

Most industrial sites are not meant to sell online: prices depend on volume, lead time and the commercial relationship. The goal is therefore the quote request, and it deserves the same care as an e-commerce checkout.

A "Request a quote" button on every product page, carrying the part number across automatically, converts markedly better than a generic contact form. A buyer should never have to retype what they were just looking at.

• Quantity field pre-filled with your usual production tiers
• Ability to attach a drawing or specification
• A stated response time, and one you keep to
• A multi-item quote basket for grouped enquiries

Response time is the most underrated variable here. An enquiry answered within four working hours converts far more often than one answered in three days — not because the buyer is impatient, but because they are consulting three suppliers in parallel and will build their specification around whichever one gives them something to work with first. State a deadline you can keep, even if it is 48 hours: a modest promise you honour beats an instant-reply claim you miss half the time.

The long cycle: what the site must carry at each stage

Six to eighteen months commonly pass between an engineer's first search and a purchase order. The site does not intervene once but at every stage, and its job changes each time. Judging it on the number of forms received measures only a fraction of what it produces.

One practical consequence: last-click attribution lies. The form filled in September often traces back to a product page read in March and found again through a brand search. Judge the site on this month's enquiries and you will cut exactly the pages that feed shortlisting — the ones that never convert directly.

Two measures beat a conversion rate: technical pages viewed per session, and the drawing download rate. Past three product pages in a session, the visitor is actively shortlisting. These are the only signals available during the twelve months when, on the surface, nothing happens.

Technical documentation: gate the download or not?

This is the most argued-over trade-off on an industrial site. Sales wants a form in front of every PDF to capture contacts; the design office on the other side gives up at the third request for an email address. Both are right, but not about the same documents.

The rule that holds: gate what took engineering work, open what describes the product. A dimensioned drawing behind a form costs you shortlist places you will never see in your analytics — while the competitor who left theirs open is already written into the specification.

When you do gate, gate lightly: three fields — work email, company, intended application — convert markedly better than an eight-field form asking for turnover. And add an HTML page describing the document, carrying its contents list and headline figures: you keep the search traffic a bare PDF would never bring you.

What actually reassures an industrial buyer

The engineer judges the product; the buyer judges the risk. Their problem is not finding the right component, it is not being the person who picked a supplier that delivered late. Anything that lowers that perceived risk is worth publishing, including what limits you.

• Certificates with their validity dates — an ISO 9001 logo without a year reads as expired
• Capacity in numbers — machine count, dimensional range, annual tonnage, size of the design office
• Standard lead times by family, even long ones: twelve weeks stated beats a lead time left unsaid
• Sectors already served — aerospace, rail and nuclear carry requirements a buyer knows how to read
• What you do not do — publishing your dimensional limits and minimum volumes saves wasted enquiries on both sides

Reference anonymity is the usual friction: many customers forbid the use of their name. A case described without naming them — "tier-one automotive supplier, 40,000 parts a year, ±0.02 mm tolerance, delivered just-in-time" — carries most of the proof and needs no permission.

One trivial detail is missed almost everywhere: the name and direct line of the technical contact. A buyer who has to fill in a form to ask a feasibility question calls the competitor whose number is on the page. A generic contact form has never reassured anyone about a supplier's ability to answer.

Languages for export: where to start

A German or Italian buyer who cannot find your catalogue in their language does not conclude that you do not export; they conclude that you are not a supplier for them. But translating 3,000 parts in full is expensive and ages badly. The order in which you translate matters more than the volume translated.

Technically, a language subfolder (/de/, /it/) with reciprocal hreflang tags is enough in the large majority of cases. A country domain earns its place when a local team genuinely runs that market; otherwise it multiplies maintenance cost with no ranking gain.

Budget £3,500 to £8,000 to build a second language into an existing catalogue site, excluding content translation: most of the work is the multilingual model, the templates and the update workflow. Technical translation is then charged per word and needs a reviewer from the trade — a generalist translator will happily turn a functional clearance into a sale.

Connecting the site to the ERP: when and why

The question always comes up. The answer depends on one thing: how often your data changes. A stable catalogue is maintained by hand without pain. A 5,000-part catalogue with moving stock and pricing cannot be maintained manually — the site becomes wrong within weeks, and a wrong catalogue is worse than no catalogue.

The most common integration is one-way: the ERP pushes part numbers, characteristics and availability to the site; the site returns quote requests. That covers the large majority of cases and costs considerably less than two-way synchronisation.

FAQ — Industrial websites

Should an industrial company sell online?

Rarely in the sense of online payment. Prices depend on volume, lead time and the commercial relationship. Publishing the catalogue in HTML with per-part quote requests delivers most of the benefit: being found on technical queries and receiving pre-qualified enquiries.

What does an industrial catalogue site cost?

Between £13,000 and £40,000 for a few thousand parts with filters, quote requests and one-way ERP integration. The main variable is not the number of parts but the quality of the source data: a catalogue already structured in the ERP costs about half as much to publish as one that exists only as PDF and spreadsheets.

How do you rank on highly technical queries?

By writing characteristics as text rather than images or PDFs, using the designations the trade actually uses — including informal ones — and creating a page per part or family. These queries have little volume individually, but they are extremely qualified and lightly contested.

Do you need an English site to export?

If export is a real share of revenue, yes. But translating the catalogue alone rarely suffices: cited standards, units and material equivalences differ by market. A useful English site translates those reference points too, not just the words.

Should an industrial site publish prices?

Rarely the exact price, often a reference point. A buyer with no order of magnitude does not enquire with you: they enquire with whoever gives one. A starting price per family, a quantity break or an hourly machining range filters out irrelevant requests and improves the quality of the enquiries you do receive, without committing you to a figure.

How long before technical queries produce results?

Three to six months for indexing and first positions on long part-number queries; twelve to eighteen on the more contested family-level terms. The limiting factor is rarely SEO, it is the publishing rate. A catalogue released in batches of 200 parts produces measurable results far sooner than a 3,000-page project delivered in one go.

Sources

• Google — Faceted navigation best practices — handling filters and canonicals
• Google — Localised versions and hreflang — structuring a multilingual site for export
• Google — File types indexable by Google — a PDF remains a single indexable URL
• Schema.org — Product — properties for a technical datasheet