Keyword Research for B2B SEO: Product Lines to Page Map

Keyword research for B2B SEO fails in a predictable way. Someone opens a keyword tool, filters for volume above a thousand, exports a spreadsheet, and hands it to a content team that has no idea which page each term belongs to. Six months later the blog is full of posts about "what is a hoist" and the product pages still rank for nothing. The fix is not a better tool. It is a different starting point: your product lines and your buyers' job titles, then intent, then a page map.

We run this sequence for manufacturers, exporters and B2B SaaS teams, and it ends with a document a writer can start on Monday morning. Here is the whole thing.

Why volume-led keyword research breaks in B2B

B2B search demand is thin and concentrated. A consumer keyword list for "running shoes" might carry fifty terms above ten thousand monthly searches. A list for industrial winches carries almost none, because the number of people worldwide who buy winches in a given month is small. Filter by volume and you delete your entire market.

The terms that matter are usually the ones a tool marks as low volume or "no data." A purchasing manager searching for "explosion proof hoist 5 ton supplier" is not a rounding error. That query has commercial intent written into every word, and the page that answers it well can carry a six-figure order. Volume is a proxy for traffic. In B2B, traffic is a proxy for nothing you actually want.

There is a second problem. B2B buying involves several people, and they search differently. The maintenance engineer wants specifications and failure modes. The procurement officer wants lead times, certifications and payment terms. The operations director wants total cost of ownership and a reason to switch vendors. One keyword list that ignores roles will produce content that speaks to nobody in particular. If you want the wider picture of how these buyers behave before you start, our guide to B2B SEO for manufacturers and suppliers covers the demand side in more depth.

Step 1: Seed from product lines, not from a keyword tool

A seed list is the raw set of phrases you will expand, score and map. Building it from the tool is backwards. Build it from what you already sell and who already buys.

Start with the product lines as your own sales team names them, then add the names customers use. Those are rarely identical. A factory may sell "material handling systems" while buyers search for "overhead crane" and "jib crane." Both vocabularies belong in the seed list, because the internal one tells you how to structure pages and the external one tells you what to put on them.

Then layer in buyer roles. For each product line, write down three to five job titles and what each person needs to decide. A hoist line might produce: crane engineer (load charts, duty class), plant manager (downtime, service intervals), procurement lead (price bands, warranty). Each role generates its own query patterns, and those patterns become your content briefs later.

Our working rule is that a seed list is finished when every product line has at least one term from each of four angles: the product name, the application, the buyer role and the commercial modifier (supplier, manufacturer, price, wholesale, custom). If a line is missing an angle, you have a gap in the map, not just in the list.

FREE Want a first read of your own site against these points? We check crawlability, keyword coverage and competitors. Get a Free SEO Analysis

Step 2: Classify intent before you judge value

Intent classification refers to sorting each keyword by what the searcher is trying to do, so that you can decide which page type should target it. In B2B, four buckets cover almost everything, and the bucket determines the page, not the other way round.

Two things fall out of this table. First, the highest-volume bucket is almost never the one that pays. Second, most B2B sites have plenty of informational content and almost no commercial investigation pages, which is exactly where a buyer sits two weeks before they email you. Google's own guidance in the Search Central documentation is blunt about this: pages should be created for people first, and the intent behind a query is the main signal of what "for people" means.

A quick classification test

If you cannot decide which bucket a keyword belongs to, read it out loud and ask what the searcher would do next if the page answered perfectly. If the answer is "read another page," it is informational. If it is "request a quote," it is transactional. If it is "shortlist two vendors," it is commercial investigation. This takes ten seconds per term and saves hours of arguing later.

Step 3: Treat low-volume, high-value terms as the core

In B2B, a keyword with forty monthly searches and a clear buying signal is worth more than a keyword with four thousand searches from students writing essays. The trick is telling them apart without guessing.

Score each term on three signals instead of one. First, commercial modifier: does the phrase contain supplier, manufacturer, wholesale, price, custom, or a specification? Second, role language: does it sound like a job title or a task (procurement, OEM, retrofit, tender)? Third, product specificity: does it name a model, a standard or a capacity? A term that hits two of the three is a keeper regardless of volume.

This is where our own client data is worth looking at. In one RAGSEO client program (client anonymized), a mining equipment manufacturer selling crushers, screening and conveying equipment built localized landing pages around exactly this kind of specific term in English, Spanish, Arabic and other languages. Monthly impressions grew from approximately 20,000 to 1,450,000, average CTR rose from 1% to 2.1%, and inquiries grew 400% versus pre-optimization, with a clear upward trend from the end of 2023 to early 2024. None of that came from chasing broad head terms. It came from pages that matched narrow queries precisely.

One caution on measurement. Keyword tools estimate B2B volumes poorly, often reporting zero for phrases that do convert. Pull the real numbers from Google Search Console once pages are live, and let your own query data correct the tool. The SEO services we run for manufacturers include that monthly reporting loop for exactly this reason.

Step 4: Map keywords to pages, one primary per URL

A page map is a table that assigns every keyword to a specific URL, with one primary term per page and a small set of secondary terms that share the same intent. Without it, two blog posts compete for the same query and both lose. With it, each page has a job.

Build the map in this order:

1. List your existing URLs first, before adding anything new. You will usually find that half your target terms already have a page that is simply under-optimized.
2. Assign one primary keyword per URL. If two pages want the same primary, merge them or change the intent of one.
3. Attach secondary keywords only where the intent matches. "Hoist price" and "hoist specifications" do not belong on the same page.
4. Mark each URL as keep, rewrite, merge or create. Be honest about merges; thin duplicate pages are the most common B2B site problem we see.
5. Add the buyer role and the intent bucket as columns, so a writer knows who the page is for and what it must do.
6. Flag internal links: every new page needs at least two links from existing pages that already have authority.

Run this as a spreadsheet, not a document. Columns: URL, primary keyword, secondary keywords, intent bucket, buyer role, action, internal link sources. That sheet is the deliverable. Everything before this step was preparation.

How many pages do you actually need?

Fewer than most agencies will tell you. A manufacturer with three product lines and four buyer roles rarely needs more than twenty to thirty well-built pages in year one: one page per product category, one per major application, and a handful of comparison and specification pages. The rest is expansion, and expansion should follow query data from Search Console, not a keyword tool export. If you are planning the first twelve months in detail, our first-year SEO plan for B2B companies walks through the sequencing.

Step 5: Refresh the map every quarter

A page map is a living document. Buyer language shifts, new standards appear, and competitors launch pages that change what Google expects for a query. Set a quarterly review and do four things: pull the top queries from Search Console, check which mapped pages are underperforming, look for new query clusters that have no page, and retire or merge pages that never earned impressions.

One more habit worth building. Before you write anything new, search the target query yourself and read the top three results. If they are all directories and marketplaces, you have an opening. If they are all deep technical guides from competitors, your page needs to be better than good. That check takes five minutes and prevents a lot of wasted writing. When the map is stable and the content engine is running, the remaining work is mostly prioritization, which is what our SEO packages are structured around: a fixed number of articles and keywords per month, aimed at the pages you have already decided matter. Current plans and prices, as of September 2026, are listed at ragseo.ai/price.

What good looks like six months in

You will know the method worked when three things are true. Your product and category pages start appearing for transactional queries you deliberately targeted, not just for your brand name. Your writers stop asking what to write about, because the map tells them. And your Search Console query report starts showing phrases you never put in the original list, which means Google has understood what your pages are for and is testing them against adjacent demand.

That last signal is the one to watch. It means the map did its job: it gave the search engine a clear enough picture of your site that it can now find its own way. If you want help building the first version, or auditing the one you have, get in touch and we will reply within 24 hours.