B2B SEO
Technical SEO for B2B That Gets Your Whole Catalogue Indexed
B2B websites tend to break in the same places: catalogues exported from an ERP, datasheets that outrank the pages they belong to, configurators search engines cannot render and country sites that compete with each other. I find which of these is costing you buyers and fix them in order.
- years in SEO
- 6+
- client websites worked on
- 60+
- person doing the work
- 1
Technical SEO for B2B, and why catalogue sites break in predictable places
Technical SEO for B2B is the work of making a catalogue-led site crawlable, renderable and indexable page by page, so the products, families and application pages buyers search for can actually rank. A typical manufacturer or distributor website makes that hard because it has grown in layers: the original brochure site, a catalogue bolted on when the ERP could export products, a resource library of PDFs, a third-party configurator, a trade portal on a subdomain, and perhaps a separate site for each country. Nobody designed it as a whole, and search engines see the seams.
The symptoms are familiar: Search Console reports large numbers of pages crawled but not indexed, important products barely receive organic visits, a datasheet ranks while its product page does not, and the US site ranks in the UK. Each has a technical cause, and most are fixable without a rebuild.
Intent is part of the diagnosis. Navigational searches for part numbers often land on PDFs or stale URLs. Transactional searches depend on product and RFQ pages being indexed and fast. Commercial investigation queries need application pages that are linked well enough to be crawled often. Technical work that ignores which intent a page serves tends to fix the wrong things first.
Key takeaways
- Catalogue sprawl from ERP and PIM exports is the most common B2B technical problem: thin SKU pages, parameter URLs and discontinued products dilute crawling and ranking signals.
- PDF datasheets often outrank the product page; the fix is HTML specifications on the page plus header-level canonicals or noindex on the PDFs where appropriate.
- Gated content, logins and JavaScript configurators hide the most useful B2B material from crawlers unless a crawlable version exists.
- Exporters need deliberate hreflang, one URL per market and language, and a plan for distributors republishing your product copy.
- Replatforming a B2B catalogue is the highest-risk SEO event most firms face; a URL-level redirect map and pre-launch crawl prevent most losses.
What's included in my technical SEO for B2B
The audit covers seven areas, each ending in developer-ready tickets ranked by impact. I work through them in crawl, render, index, understand order, because a page that is never crawled cannot benefit from better schema.
Crawl and index coverage measured against your product database
A full crawl compared with your product database, XML sitemaps and, where available, server logs showing what Googlebot actually requests. Index coverage is reviewed by template - families, products, applications, resources - against the list of pages that should rank, not against the total URL count.
- Crawl versus product database and sitemaps
- Server log review where logs are available
- Indexation by template, with exclusion reasons
- Staging, portal and test copies found and removed
Filter, configurator and parameter URLs on distributor and manufacturer catalogues
Filter combinations such as material, size and sort order can create thousands of duplicate URLs. Filters that match real search demand become static, linked, indexable pages; everything else stays usable for buyers but out of the index. Internal site search results get the same treatment, because they create infinite URL spaces.
- An indexable-filter list based on demand
- Canonical or blocking rules for the rest
- Internal links kept on clean URLs
- Site search noindexed and out of sitemaps
Catalogue architecture that keeps priority B2B products a few clicks deep
Products grouped into families that match how buyers search, with variants on the family or product page. Application pages linked from the main navigation rather than buried in a resource section. Readable, stable URL slugs that survive the next ERP upgrade, and category and related-product modules that surface high-value products buried in deep pagination.
- Variant consolidation onto family or product pages
- Application pages in the main navigation
- Stable slugs independent of ERP IDs
- Click depth checked for priority products
A rank, route or noindex decision for every industrial datasheet and PDF
Every PDF, datasheet and CAD file is inventoried and given a decision by document type: canonical header to the product page, X-Robots-Tag noindex, or indexable with a proper title and a link back into the site. CAD and 3D viewers are checked for page weight and loaded on demand where they slow pages down.
- A full document inventory by type
- HTTP header canonicals or noindex where they fit
- Gated content reviewed for crawlable summaries
- Viewers loaded on demand, not on page load
Rendering, Core Web Vitals and structured data checked template by template
JavaScript spec panels, tabbed content and configurators are checked to confirm specs, prices and links appear in rendered HTML. Core Web Vitals come from field data by template, where heavy images, viewers, chat widgets and years of tag manager scripts usually cause the trouble. Structured data generated from the catalogue is validated per template, with checks after each release.
- Rendered HTML compared with the source for key templates
- Field Core Web Vitals by template
- Product, ProductGroup and breadcrumb validation
- Post-release checks so schema bugs surface early
Hreflang and canonicals for exporters selling to UK, US and EU buyers
Country and language versions get reciprocal hreflang with x-default on every indexable template, consistent canonicals, and language and market in the URL rather than forced IP redirects. Where UK and US terminology, units or standards differ, separate en-GB and en-US pages are justified; elsewhere, one page is often enough.
- Hreflang audit across every country version
- Canonical consistency between markets
- IP redirects replaced with suggestion banners
- Distributor duplicates of your copy identified
Redirects for discontinued parts and RFQ forms tested end to end
Discontinued products redirect to their successor or closest family, or keep a page with a replacement note where buyers still search the old code. RFQ, sample and quote forms are tested on mobile, including drawing uploads, tracked events and the lead arriving in the CRM with its source.
- A discontinued-product redirect rule set
- Replacement notes for still-searched part numbers
- Form rendering and file upload tested
- Submission events and CRM handover confirmed
Experience
Technical SEO from someone who has done it many times
You work with me from the first call to the last report. No account managers, no hand-offs to a junior team.
Book a free call- years in SEO
- 6+
- client websites worked on
- 60+
- documented case studies
- 35
More B2B SEO
The other disciplines, written for manufacturers, distributors and B2B service firms.
Keyword Research for B2B
Research built around the buying committee: specification, standard, part number and application queries that tools show as zero volume but carry real contract value, mapped to one owning page each.
Read moreOn-Page SEO for B2B
Product, capability and application pages rewritten in buyer vocabulary, with specifications in HTML rather than PDFs, certifications stated precisely and an RFQ or sample request where the intent calls for one.
Read moreContent Strategy for B2B
Application pages, buyer guides, material comparisons and case studies by application, written with your engineers and planned around each committee member, from specifier to procurement to finance.
Read moreLocal SEO for B2B
Business profiles and location pages for plants, warehouses, trade counters and regional offices, plus the industrial and trade directories that buyers and search engines both rely on.
Read moreOff-Page SEO for B2B
Supplier, distributor and trade body links, trade press, certification registers, industry directories and review platforms - the sources buyers and AI assistants check before a supplier makes the shortlist.
Read moreCommon technical issues on B2B sites
The issues I find most often on manufacturer, distributor and B2B service sites, with what I do about them.
| Issue | Symptom | Fix | Impact |
|---|---|---|---|
| PDF datasheets outranking product pages | Search visitors land on a PDF with no navigation, no RFQ and no tracking | HTML spec tables on the page; Link rel=canonical HTTP header pointing PDF to page, or X-Robots-Tag noindex where the PDF adds nothing | High - recovers leads from navigational and spec queries |
| Filter and configurator parameters | Thousands of URLs like ?material=pp&neck=24-410&sort=asc crawled and reported as duplicates | Decide which filter combinations deserve indexable pages; canonicalise or block the rest; keep internal links on clean URLs | High on large catalogues |
| Thin ERP-generated SKU pages | Pages crawled but not indexed; near-identical titles and content | Consolidate variants onto family or product pages; enrich templates from PIM fields; noindex true leftovers | High |
| Discontinued products returning errors | Lost links and rankings, frustrated repeat buyers searching old part numbers | Redirect to the successor product or closest family; keep a page with a replacement note where buyers still search the code | Medium to high |
| Gated or login-only content | Spec libraries, CAD and technical guides invisible to crawlers | Ungate what qualifies buyers; publish crawlable summaries of what stays gated | Medium |
| Distributor duplicates of your product copy | A distributor's page ranks for your own product name | Richer manufacturer pages, distinct copy for feeds, and asking key distributors to link back to the source page | Medium |
| Country sites without hreflang | The wrong market's page ranks; quote forms routed to the wrong team | Reciprocal hreflang with x-default, consistent canonicals, market-specific content where needed | High for exporters |
| Staging, portal or test environments indexed | Duplicate copies of the site appearing in search | Authentication on staging; noindex and removal for anything already indexed | Medium |
Crawl, render, index, understand - applied to a catalogue
I work through these in order, because a page that is never crawled cannot benefit from better schema.
- 1
Crawl
Can search engines reach every product family, product and application page through plain links? I compare a full crawl against your product database and sitemaps, and check server logs where available to see what Googlebot actually requests.
- 2
Render
Do spec tables, prices, stock and links appear in the rendered HTML, or only after a script or user interaction? Configurators and tabbed spec panels built in JavaScript are frequent culprits.
- 3
Index
Which pages are indexed, which are excluded and why? Index coverage is compared against the list of pages that should rank, not against the total URL count.
- 4
Understand
Do titles, headings, internal links and structured data tell search engines what each page is about and how products, families and applications relate? This is where technical work hands over to on-page SEO.
Free review
Get a free Technical SEO review
Send me your site. Before the call I look at your rankings and pages, then tell you what I would fix first and what it is likely worth.
PDFs, datasheets and CAD files: rank them or route them
Google indexes PDFs and will rank them, which is why a well-linked datasheet so often beats a thin product page. That is not always bad. A datasheet that ranks for a part number at least puts you in front of the buyer. The problem is what happens next: a PDF has no navigation, no RFQ form, no related products and, unless it is set up carefully, no tracking.
My default is to put the information that ranks on the HTML page and treat the PDF as a download. Where a PDF duplicates the page, a canonical in the HTTP header pointing to the product page tells search engines which one to prefer. Where the PDF adds nothing for search, X-Robots-Tag noindex keeps it out of results without blocking downloads. Where a PDF is the only source of something valuable, such as a full technical manual, it stays indexable with a descriptive filename, a proper title in the document properties and a link back to the site inside the document.
Multilingual export sites: common setup vs working setup
Exporters often launch country sites without a plan for how they relate to each other. These are the patterns I see and the ones I recommend.
Common setup
- Language chosen by IP redirect, so crawlers only see one version
- UK and US English sharing one URL with no market signals
- Machine-translated catalogue pages with English titles left in place
- Hreflang on some templates but not on products
- Each country team uploading its own PDFs and certificates
Working setup
- Language and market in the URL, with a suggestion banner instead of forced redirects
- Separate en-GB and en-US pages where terminology or standards differ
- Translation reviewed by a native speaker with technical vocabulary
- Reciprocal hreflang with x-default on every indexable template
- One central resource library with market-specific versions where needed
Migrations: the highest-risk moment for a B2B catalogue
Migrations deserve their own warning. When a B2B firm moves to a new ecommerce platform, PIM or CMS, product URLs, PDF paths and country structures often change all at once. Without a URL-level redirect map, years of part number rankings and inbound links from distributors can disappear in a weekend. I build the map from crawls, analytics, backlink data and Search Console, test redirects on staging and monitor indexing closely after launch.
Where SEO meets CRO
A form that fails to load is a technical SEO problem too
RFQ forms, quote configurators and sample request flows are often third-party scripts that break silently, load slowly or are blocked by consent tools. I test them as part of every technical audit: does the form render on mobile, does the file upload for drawings work, does the submission fire a tracked event, and does the lead arrive in the CRM with its source. Technical fixes that bring buyers to a broken form achieve nothing.
From technical fixes to growth
Technical work clears the path. What travels down it comes from the other disciplines: a keyword map from B2B keyword research that tells you which pages deserve to be indexed, templates rebuilt through on-page SEO, and new application pages from your content strategy.
For exporters and firms with several sites, technical and local SEO for B2B overlap on location data and market targeting. For the full approach, go to B2B SEO, or see my technical SEO service.
Growth from real projects
Learn more about technical seo
Technical SEO in other industries
Technical SEO for B2B FAQs
Should we noindex our PDF datasheets?
Not automatically. If a PDF duplicates information that is now on the product page, a canonical header pointing to the page or a noindex header is usually right. If the PDF is the only source of useful detail, such as a full manual, keep it indexable and make sure it links back to the site. Decide per document type, not across the board.
Our ERP generates thousands of product pages. Is that a problem?
Only if many of them are thin, duplicated or unwanted by searchers. The question is which pages buyers search for and which exist only because the ERP has a row for them. Variants usually belong together, obsolete items need redirects or replacement notes, and the remaining pages need enough data from the PIM to be useful.
Should we ungate our whitepapers and technical guides?
Ungate content that helps buyers decide whether you fit, such as specifications, application guides and compliance information. Gating that material hides it from search and frustrates engineers. Keep gates for things that justify a sales conversation, such as detailed calculators or custom reports, and publish a crawlable summary of what is behind each gate.
Do we need hreflang if we only have UK and US English sites?
If both versions are indexable and target different markets, yes. Without it, search engines may show the US page to UK buyers or treat the two as duplicates. Hreflang tells them which URL suits which market, and it matters more when spelling, units, standards and contact routes differ between the versions.
How do we protect rankings when we move to a new platform?
Plan the migration as an SEO project from the start. Crawl the old site, list every URL with traffic, links or rankings, map each one to its new equivalent, and test the redirects before launch. Keep PDF and image paths in scope, since they often carry links. Then monitor indexing and part number queries closely for several weeks.
Can technical SEO for B2B be delivered through our developers or ecommerce agency?
Yes, that is the normal setup. I write tickets with the issue, evidence, the required change and how to test it, in the language your developers already use. I can join sprint planning or review pull requests and staging builds, and I re-crawl after each release to confirm the fix behaves as intended.
Start technical SEO for B2B with an index check of your catalogue
Send me your domain and roughly how many products you list. On a free 30-minute review I will show you how much of the catalogue search engines actually index, where PDFs or parameters are getting in the way, and what I would fix first.