Keyword Research for a B2B Website: A Practical Guide

Most B2B keyword lists are built backwards. Someone exports a few thousand terms from a tool, sorts by volume, picks the biggest numbers, and hands the list to a writer. Six months later the blog ranks for terms that bring traffic and almost no sales conversations.

The problem is not the tool. It is that B2B search behaves nothing like B2C. Your buyer is a committee, the search volumes are small, and the phrase that closes a deal might get 40 searches a month while a tangential term gets 9,000. If you optimize for the 9,000, you win a vanity metric and lose the pipeline.

This guide walks through keyword research the way a revenue-focused team actually does it: start from the buyer and the offer, map intent, cluster by page, then prioritize with a method that accounts for low volume. By the end you will have a keyword map you can hand to writers and a way to defend why each page exists.

Start with the buyer, not the keyword tool

Before you open Ahrefs or Semrush, write down two things.

First, the jobs your product does. Not your feature list, the outcomes a customer hires you for. A payroll platform is not hired for "automated payroll." It is hired to stop the finance team from spending three days a month reconciling timesheets. Those outcome phrases are where real searches live.

Second, who searches. In B2B a single deal usually involves several people, and they search for different things. A practical shortcut: list the roles on the buying committee and the question each one types when they are stuck.

Same deal, different searches by role (illustrative)
Role	What they worry about	Likely search
Practitioner / end user	Will this make my day easier?	"how to automate timesheet reconciliation"
Manager / champion	Will my team adopt it?	"best payroll software for mid-size companies"
Finance / economic buyer	What does it cost and save?	"payroll software pricing" / "payroll ROI"
IT / security	Is it safe and compliant?	"payroll software SOC 2 integration"

That single table changes your research. You are no longer hunting one keyword. You are mapping a conversation that happens across several people and several months, and your content has to show up at each point.

Build your seed list from real language

Seed keywords are the starting terms you expand from. The mistake is inventing them at your desk. Pull them from places where buyers already use their own words.

Sources that consistently beat brainstorming:

Sales calls and your CRM. Read closed-won and closed-lost notes. The phrases prospects use to describe their problem are gold, and they rarely match your marketing copy.
Support tickets and your help center search. People type their actual confusion here.
Review sites like G2 and Capterra. Read the "what problem were you solving" sections for your category and competitors.
Community threads on Reddit, niche Slack groups, and LinkedIn comments where your buyers complain out loud.
Google autocomplete and "People also ask." Type your category and watch what completes.

Collect 20 to 50 seed terms. Messy is fine. You are about to expand and filter them.

Expand with tools, then read the SERP

Drop your seeds into a keyword tool (Ahrefs, Semrush, Google Keyword Planner, or the free Search Console "Performance" report if your site already ranks for anything). Export the keyword ideas, the related terms, and the questions.

Now do the step most people skip: open the actual search results for your top candidate terms. The SERP tells you what Google thinks the query means, which matters more than any volume number.

Ask three questions as you look:

What format ranks? If page one is all listicles and comparison pages, a single product page will not break in. Google has decided this query wants a roundup.
Is it informational or commercial? Results full of "how to" guides signal early-stage research. Results full of pricing pages, product pages, and "best X" comparisons signal buying intent.
Can you realistically compete? If the top ten are all DR 80+ domains with 2,000-word guides and dozens of links, that term is a long-term play, not a quick win.

Reading the SERP is the difference between a keyword list and a keyword strategy. It is also where you catch terms that look relevant but mean something else entirely.

Sort by intent, because volume lies in B2B

In B2C, search volume is a decent proxy for opportunity. In B2B it is often misleading. A bottom-of-funnel term like "[your category] for healthcare compliance" might show 70 searches a month, but every one of those searchers is a qualified buyer with a budget. A top-of-funnel term might show 8,000 searches that are mostly students, job seekers, and competitors.

Group every keyword by intent. A simple four-bucket model works:

Informational ("what is", "how to", "guide"): the buyer is learning. Map these to blog posts and resources.
Commercial investigation ("best", "vs", "alternatives", "reviews", "for [industry]"): the buyer is comparing. Map these to comparison pages, category pages, and case studies.
Transactional ("pricing", "demo", "free trial", "[brand] login"): the buyer is ready to act. Map these to product, pricing, and demo pages.
Navigational (your brand, competitor brands): people looking for a specific company.

This is the same discipline that makes paid search profitable, where matching keywords to search intent decides whether a click ever becomes a lead. The logic carries straight into SEO: a term is only valuable if the page behind it answers the exact intent.

A useful gut check: for each keyword, finish the sentence "someone searching this is trying to ___." If you cannot connect that intent to something you sell, the term is a traffic play, not a pipeline play. Keep a few of those for brand reach, but do not let them dominate the map.

Cluster keywords into pages

You do not write one page per keyword. Modern search rewards a single strong page that covers a topic and its many phrasings. So the next step is grouping.

Cluster keywords that share the same intent and would be satisfied by the same page. "Crm for small business," "best crm for small business," "small business crm software," and "affordable crm for small teams" are one page, not four. Splitting them creates thin pages that compete with each other (keyword cannibalization).

A reliable way to cluster without guessing: check whether the same URLs rank across the terms. If two keywords share most of their top-ten results, Google treats them as the same topic, so one page can target both. Tools can automate this with "SERP overlap" clustering, or you can spot-check manually for a small site. The principle stays simple: one intent, one page, every relevant phrasing.

For each cluster, name a primary keyword (the head term you optimize the title and H1 around) and a set of secondary keywords (variations and related questions you work into H2s and body copy). That structure also feeds your site architecture: clusters become your main service and resource pages, and the supporting articles link up to them.

Prioritize: a scoring method for small volumes

With clusters in hand, you cannot do everything at once. You need an order. Volume-only sorting fails here, so score each cluster on four factors instead.

Business value (1 to 5): how close is this to a buying decision and your offer? A pricing-intent term outranks a definition.
Intent match (1 to 5): how well can a page you can actually build satisfy the SERP?
Difficulty (reverse, 1 to 5): how hard is it to rank, based on the domains and link profiles on page one? Lower difficulty scores higher.
Volume (1 to 5): real demand, but capped so it cannot dominate the other three.

Multiply or sum the scores and sort descending. The point is not mathematical precision. It is forcing business value and winnability ahead of raw volume, which is exactly the bias B2B needs.

A typical healthy roadmap front-loads commercial-investigation and transactional clusters (smaller volume, higher value, often easier because big publishers ignore them) and uses high-volume informational clusters to build authority over time. The bottom-of-funnel pages tend to pay back first.

Don't forget the gaps competitors leave

Two extra passes sharpen the map.

Competitor gap analysis. Most tools have a "content gap" or "keyword gap" report. Feed in two or three competitors and your own domain. The report shows terms they rank for and you do not. Filter that list by intent and business value the same way you did your own. You will usually find a cluster of high-value comparison and industry terms your competitors quietly own.

Long-tail and question terms. B2B buyers ask specific, narrow questions: "does [category] integrate with Salesforce," "[category] for companies under 50 employees." Each has tiny volume. Together they represent the most qualified traffic on your site, and they are far easier to rank for. Mine "People also ask," the "Questions" filter in your tool, and your own sales FAQ.

Turn the research into a keyword map

The deliverable is a simple spreadsheet, one row per cluster, that your team can act on. Columns that earn their place:

Cluster name and primary keyword
Secondary keywords and key questions
Intent bucket
Target page (existing URL or "new")
Page type (blog, comparison, product, pricing)
Priority score
Status

This map is the bridge between research and results. Keyword research that lives in a tool export changes nothing. The same research turned into a prioritized page plan, with an owner for each page, is what moves rankings and, eventually, pipeline. If you are building this from the ground up, our SEO from scratch walkthrough shows where the keyword map fits in the wider plan, and the broader B2B SEO guide covers how to turn ranked pages into actual leads.

Treat the map as living. Re-run the research quarterly. New competitors enter, language shifts, and Search Console will start showing you queries you never targeted (often your best new ideas, because real users found you for them).

FAQ

How many keywords does a B2B website need to target?

Fewer than you think. Most B2B sites do well with 30 to 80 well-chosen clusters, not thousands of individual keywords. Depth on a focused set beats spreading thin across terms you cannot win.

Should I target high-volume or low-volume keywords?

In B2B, lean toward lower-volume, higher-intent terms first, especially commercial and transactional ones. A keyword with 50 searches a month from qualified buyers is usually worth more than one with 5,000 searches from a mixed audience. Use high-volume informational terms to build authority, but do not let them set your priorities.

What's the difference between keyword research for SEO and for PPC?

The research overlaps heavily, the application differs. For SEO you cluster keywords into pages and play a longer game on authority and content depth. For PPC you bid on exact intent and control spend in real time, so negative keywords and tight match types matter more. Many teams run both off one shared keyword map.

Which keyword research tool is best for B2B?

Ahrefs and Semrush are the common paid choices and both handle clustering and gap analysis well. Google Keyword Planner is free but gives ranges rather than exact volumes. Google Search Console is underrated for any site already getting impressions, because it shows the real queries you appear for. The tool matters less than reading the SERP and sorting by intent.

How do I find keywords with no search volume showing in tools?

Tools undercount B2B and long-tail terms badly. A keyword showing zero volume can still get steady, qualified traffic. Pull these from sales calls, support tickets, "People also ask," autocomplete, and your own Search Console query report. If real buyers use the phrase, it is worth a page, regardless of what the tool claims.

How often should I redo keyword research?

A full refresh once or twice a year, with a lighter quarterly check of Search Console for new queries and ranking shifts. Markets, competitors, and buyer language move, and your live search data is the cheapest source of new keyword ideas you have.

In short

Keyword research for B2B is less about finding the biggest numbers and more about mapping a buying decision made by several people over several months. Done right, it produces a prioritized list of pages, each tied to real intent and real business value.

A quick checklist before you start writing:

Built seeds from sales calls, support tickets, and reviews, not just brainstorming
Mapped keywords to the buying committee, not one persona
Read the SERP for your top terms before committing
Sorted by intent, not volume
Clustered keywords into pages to avoid thin, competing content
Scored clusters on business value and difficulty, not search volume alone
Turned it all into a keyword map with owners and priorities

If your current blog brings traffic but few sales conversations, the keyword map is usually where the disconnect starts. We help B2B teams rebuild that map around revenue, then turn it into pages that rank and convert. If that is the gap you are staring at, get in touch for a short review of your keyword strategy and we will show you where the qualified searches are hiding.