B2B Keyword Research: Low Volume, High Value

Consider an industrial filtration supplier. “ATEX dust extraction system integrator” is marked as zero searches per month. Its average project is worth £72,000, and that phrase has appeared in four qualified sales calls.

Deleting it from the research would be a category error.

That is the central problem with B2B SEO keyword research. Search-volume tools measure estimated query frequency. They do not measure purchase probability, sales relevance or contract value.

For a narrow B2B market, ten searches from the right people can be worth more than 10,000 loosely related visits.

Why B2B SEO keyword research starts below the volume threshold

B2B searches are structurally difficult for keyword tools to measure.

The buying population may consist of a few hundred companies. Purchases happen every three or five years. Several people describe the same requirement differently. Technical, operational and procurement language also changes by sector.

Most platforms model, group and round their volume data. A displayed zero should therefore be read as “no reliable reportable estimate”, not proof that nobody searches the term.

Three questions matter more:

• Has the language appeared in customer, sales or search data?
• Does the query describe a problem the company can profitably solve?
• Do the search results indicate that Google interprets it in the intended way?

Volume still has a role. It helps compare established terms and estimate the upper boundary of demand. It should not act as an automatic rejection rule.

Keyword selection also cannot settle positioning, technical priorities or the wider organic programme. Those decisions belong in a broader B2B SEO strategy. The research has a narrower job: identify commercially useful search language and document the evidence behind it.

A five-source B2B keyword research method

Third-party keyword platforms are useful expansion tools. They are weak starting points because they know nothing about the company’s sales conversations, margins or customer vocabulary.

Start with evidence closest to revenue, then move outwards.

1. CRM records, sales calls and proposals

Pull 12–24 months of won opportunities and qualified losses. Review:

• Discovery-call transcripts and notes
• Form enquiries and live-chat messages
• Proposal titles and scope descriptions
• Objections, lost reasons and competitor mentions
• Questions asked before a technical or commercial decision

Extract exact phrases rather than rewriting everything into polished marketing language. Record who used each phrase, their sector, the sales stage and the eventual outcome.

Prioritise won and genuinely qualified opportunities. A CRM full of students, recruiters and irrelevant enquiries will otherwise corrupt the research.

Implementation, compatibility, migration, compliance and procurement terms often appear late in evaluation. These phrases matter because answering the underlying concerns can address the questions that shorten B2B sales cycles, even when monthly search volume is negligible.

2. Search Console and on-site search data

Export at least 12 months of Search Console queries. Look beyond clicks.

A query producing 20 impressions from an average position of 34 is evidence that the phrase exists. Its lack of traffic may reflect poor visibility rather than absent demand.

Review queries by page, country and date. Separate genuine wording variants from terms that imply a different need. Singular and plural forms can usually sit together; “consultant”, “software” and “training” rarely should.

Internal site searches, support tickets and form free-text fields provide additional language. Before discarding queries because the site receives no organic traffic, use a technical and on-page SEO audit checklist to check whether indexation, relevance or page quality is suppressing the signal.

3. Customer and market vocabulary

Ask customers what they called the problem before they knew the company’s preferred terminology. Useful prompts include:

• What would you have typed into Google?
• Which alternatives did you compare?
• What internal name did the project have?
• Which requirement made the purchase urgent?
• Which words would your finance, technical and operational colleagues use?

Interview people separately where possible. A group workshop tends to converge on the loudest person’s vocabulary.

Competitor sites, trade directories, review platforms and tender documents can expand the list. Treat their terminology as a lead, not validation. Competitors can target fashionable phrases that generate no qualified demand.

4. Search-result inspection

Search each promising term and its close variants. Record the industries, geographies and result types appearing on the first page.

If “asset management platform” returns mostly consumer investment products, a B2B software supplier faces an interpretation problem. If “food factory asset management software” returns relevant vendors and technical resources, the narrower term has stronger alignment despite lower reported volume.

Autocomplete, related searches and question boxes can reveal alternative phrasing. They confirm that language exists within the search ecosystem, but they do not establish commercial value by themselves.

5. Controlled Google Ads tests

Paid search can expose real query activity faster than waiting for organic rankings.

For a narrow offer, a £300–£750 test over six to eight weeks may be enough to collect impressions and actual search terms. Use tightly controlled exact and phrase match groups, strong negatives and the relevant geography.

Do not judge the test by click-through rate alone. Record whether searches came from the intended company type and whether resulting enquiries were qualified.

Some terms will produce too little activity for a decisive result. That is still evidence. Keep them in a test category rather than inflating bids or broadening targeting until the original question becomes impossible to answer.

How to qualify zero-volume B2B keywords

A zero-volume term earns serious consideration when it has at least two independent signals, preferably including one first-party source.

Strong signals include:

• The phrase appears across several qualified sales conversations.
• Search Console has recorded impressions or clicks.
• Paid-search data contains the term or a close variant.
• Customers independently use the same language.
• Search results consistently match the intended service and buyer.
• The economics remain attractive under conservative assumptions.

Reject or park a term when it is internal jargon, one salesperson’s preferred phrase, tied to an unserved location or associated with a different product category.

Do not create separate research entries for every word order. Preserve one primary formulation alongside observed variants, sources and modifiers.

Consider an ISO 27001 consultancy evaluating “ISO 27001 gap analysis for fintech”. A keyword platform reports zero monthly searches, but the evidence shows:

• Three close mentions across 28 qualified discovery calls
• Nineteen Search Console impressions from an underperforming page
• Fourteen paid-search impressions and one qualified enquiry during a small test

The average engagement is £18,000 with a 55% gross margin. Based on the observed variants, the team models 72 addressable organic visits per year.

Using a 6% enquiry rate, 70% qualification rate and 30% close rate:

72 × 6% × 70% × 30% = 0.91 expected wins

0.91 × £18,000 × 55% = approximately £8,980 expected gross profit

That is not a forecast. It is a sensitivity model. Its purpose is to show that “zero volume” and “zero economic potential” are different claims.

Score keywords by expected commercial value

A practical comparison starts with this formula:

Expected annual gross profit = addressable visits × enquiry rate × qualification rate × close rate × gross profit per win

Addressable visits are not the same as total searches. The estimate should account for geography, likely ranking, click share, search-result competition and the proportion of searches matching the offer.

Using the consultancy example:

The figures are assumptions, so they should be conservative and applied consistently. The comparison matters more than false precision.

Companies without reliable conversion data can use a weighted 100-point score:

• Commercial fit: 30%
• Observed buyer evidence: 25%
• Purchase specificity: 20%
• Search-result alignment and ranking feasibility: 15%
• Reported volume: 10%

Score each factor from one to five, then apply the weighting. Any query with weak commercial fit should be rejected regardless of its total score. Traffic cannot compensate for attracting the wrong buyer.

Build a decision-ready keyword register

The final output should be an evidence register, not a spreadsheet containing 5,000 unfiltered exports.

Each retained entry should include the wording and variants, evidence source, buyer role, sector, relevant offer, geography, tool volume, first-party impressions, commercial assumptions, search-result notes and decision.

Use four decisions:

• Keep: supported by multiple signals and strong commercial fit.
• Test: commercially plausible but lacking enough demand evidence.
• Park: potentially useful after a product, market or terminology change.
• Reject: wrong audience, weak offer fit or mismatched search results.

The register should be portable enough for an in-house marketer, a provider of ongoing SEO expertise or a team delivering a defined SEO project to understand every decision without repeating discovery.

Review it quarterly and whenever the offer, target market or sales language changes. Monthly reinvention usually creates noise; an untouched annual export becomes stale.

FAQ

Are zero-volume keywords worth targeting?

Yes, when first-party evidence, search-result alignment and contract economics support them. A zero from one platform is insufficient evidence either to approve or reject a term.

Does low search volume mean low competition?

No. A term producing ten searches per month can attract every specialist supplier in a high-value category. Inspect the actual search results, ranking strength and paid competition rather than inferring difficulty from volume.

How many keywords should a B2B company research?

For one specialist offer, 30–100 substantiated terms are usually more useful than thousands of exports. The right number depends on the number of products, sectors, buyer roles and geographies—not an arbitrary agency target.

Which keyword research tool is best for B2B?

No single tool has enough context. Combine CRM and call data, Search Console, customer language, search-result inspection, paid-search terms and one reputable third-party platform for expansion.

Can Google Ads validate an SEO keyword?

It can validate that searches occur and reveal the wording, cost and lead quality. It cannot prove that the company can rank organically or that an organic visitor will convert at the same rate.

Summary

• Treat reported volume as an estimate, not a veto.
• Start with sales, customer and first-party search evidence.
• Require multiple signals for zero-volume terms.
• Compare keywords using expected commercial value.
• Deliver a documented keep, test, park or reject decision.

Actualyse builds B2B SEO programmes that compound into a durable source of qualified pipeline. Book a call to talk through where yours stands.