The short answer
Collect buyer language from sales calls, enquiry forms, site search and Search Console. Treat a search as real if you can find evidence of it: impressions, autocomplete predictions, your own customers using the words. Group terms that share an intent, decide which existing or new page should answer each group, and score the groups on buyer fit, intent and effort. Measure groups of pages over months, not single keywords.
Why keyword tools show zero for B2B searches
Each source of search data has a floor below which it stops reporting, and many B2B searches sit under it.
- Keyword Planner. Google's own planner says that "keywords with very low search volumes, or keywords that are considered sensitive aren't discoverable or forecastable"1. Its average monthly searches combine a keyword with its close variants, for the location and network settings you choose2, so a niche UK term can disappear into a broader one or out of the data altogether.
- Google Trends. Trends shows data for popular terms only, so "search terms with low volume appear as 0"3. Its figures are also relative, scaled from 0 to 100, not counts of searches3.
- Search Console. Even your own data hides the rarest searches. Anonymised queries, "those that aren't issued by more than a few dozen users over a two-to-three month period", are left out of the query table, though they are counted in chart totals4.
- Third-party SEO tools. These build their own estimates from their own data and methods. For rare terms, treat their figures as rough and compare tools with care.
For a company selling into a narrow market, a search made by a few dozen people a month can still be made by most of the buyers who matter. That is why our guide to B2B website SEO says you cannot judge B2B keywords by volume alone.
Step 1: collect the words buyers actually use
Start from evidence about your buyers rather than from a tool's suggestions. Each source below shows the language people use before they know your product names.
| Source | What to take from it |
|---|---|
| Sales and discovery call notes or recordings | How prospects describe the problem in their first few sentences, and the alternatives they mention |
| Enquiry and demo form free-text fields | The exact words people typed when asking for help |
| CRM answers to 'how did you find us?' | Searches people remember making |
| Tender documents and requirements lists | Formal category names and the capabilities buyers check for |
| Support tickets and onboarding questions | Problems customers had before they bought |
| Your site search | What visitors could not find; GA4 enhanced measurement records a view_search_results event when a search results page loads5 |
| Search Console queries for your existing pages | Searches you already appear for, including ones you never targeted |
| Google autocomplete and related searches | Phrasings people really type; Google says autocomplete predictions "reflect real searches that have been done on Google"6 |
Write terms down exactly as people said or typed them. Google's SEO Starter Guide makes the point that people who know a topic well use different words from people who are new to it7. In B2B, the people searching are often in the second group: an operations manager with a problem, not a specialist who knows your category name.
Step 2: check the search exists, without a volume figure
For each term, look for at least one piece of evidence that real people search for it:
- Impressions in Search Console for the term or a close variant, over the longest date range available. Even a handful is evidence.
- An autocomplete prediction when you start typing it, in a private browser window with the location set to the UK.
- A results page with relevant businesses on it. If competitors, directories or publications have written pages for the term, someone thinks it is worth answering.
- Use by your own buyers in calls, forms or tenders. This is the strongest evidence of all, because it comes from the people you sell to.
A term with no evidence at all is not necessarily worthless, but it should wait until the terms with evidence are covered.
Step 3: read the results page to find the intent
Search the term and note what kind of pages Google shows. The results tell you what searchers expect to find:
- Supplier and service pages: people want a provider. A service or solution page is the right answer.
- Guides and explanations: people are still learning. A guide that explains the problem and links to the relevant service page fits better.
- Lists, directories and comparison pages: people are shortlisting. A service page may struggle; a clear comparison or 'how to choose' page might fit.
- Unrelated results: the words mean something else to most searchers. Find another phrasing.
Note the words in the titles that rank, too. They are often the clearest guide to how to phrase your own; our guide to rewriting page titles for clicks covers that step.
Step 4: group terms by intent, one page per group
Low-volume keywords add up when you group them. Ten phrasings of the same need, each searched a few times a month, are one topic and belong on one page. Google's Starter Guide says not to worry about anticipating every variation of how someone might search7, and warns that repeating the same words over and over is tiring for users and that keyword stuffing is against Google's spam policies7.
Two terms belong in the same group when the same page would satisfy both searchers and the results pages look alike. If they need different answers, they need different pages. Creating separate pages for close variants tends to leave them competing with each other; our Search Console quick wins guide shows how to spot that.
An illustrative example, for a UK company selling payroll software to care homes (the terms are examples of the method, not researched volumes):
| Group | Example terms | Intent | Page |
|---|---|---|---|
| Category | payroll software for care homes; care home payroll system | Find a supplier | Main solution page |
| Problem | how to pay sleep-in shifts; sleep-in shift pay rules | Understand a rule | Guide, linking to the solution page |
| Integration | rota and payroll integration for care homes | Check a capability | Integrations section or page |
| Switching | moving care home payroll provider | Plan a change | Switching or onboarding page |
Step 5: prioritise by fit, not volume
Without reliable volumes, score each group on things you can judge. A simple 1 to 3 score on each line is enough:
- Buyer fit: do your best customers use these words?
- Closeness to an enquiry: is the searcher looking for a supplier, or just learning?
- Evidence: how strong is the proof that the search exists?
- Existing page: do you already have a page that nearly answers it? Improving it is usually quicker than writing a new one.
- Your expertise: can you answer better than the pages already shown?
That last line matters. Google's guidance on helpful content lists, as a warning sign, deciding "to enter some niche topic area without any real expertise, but instead mainly because you thought you'd get search traffic"8. Low competition is a reason to write only if you have something real to say.
Step 6: map each group to a page and write for the buyer
For each group, choose the page that will answer it, and work the main phrasing into the places that matter: the title, the main heading, the opening paragraph and the anchor text of links from related pages. Our guide to internal linking for B2B websites covers the links. Then answer the question fully, in the buyer's words, with the details a buyer would otherwise have to email you about.
Record which group each page targets. When you later review Search Console, you will know which queries a page was meant to win and can see whether it does.
How to measure low-volume keywords
- Measure pages and groups, not single queries. Rare queries are anonymised and hidden from the query table4, so a page's totals in the Pages view are more complete than any query list.
- Use long windows. Weekly numbers on low-volume pages are mostly noise. Compare three months with the same three months a year earlier.
- Count impressions first. On a new or rewritten page, impressions for the target group are the earliest sign that Google sees the page as relevant. Clicks and enquiries follow, or do not.
- Check against the CRM. The point of low-volume B2B keywords is the quality of the enquiries they bring.
Expect many results to be inconclusive at this scale. Our guide to before-and-after SEO testing shows how to compare changed pages with control pages so you do not mistake noise for a result. When the page already exists and has slipped, our guide to refreshing old content covers whether to update, merge or remove it.
Common mistakes
- Dropping every term a tool shows as zero. This removes most of the searches your buyers make.
- Chasing broad, popular terms that bring visitors who will never buy, such as students or job seekers.
- Using your own product names as keywords before buyers know them.
- One page per variant. Thin pages that compete with each other instead of one page that answers the need.
- Judging a page after two weeks. Low-volume pages need months of data.
Where Artiweave fits
The queries Search Console already shows for each page are the low-volume evidence this guide describes, and they are part of what Artiweave reads. From them it proposes specific changes, such as a clearer title or a section that answers those searches, for a marketer to approve, edit or reject. When the numbers are too small to judge, the outcome is recorded as "insufficient traffic".
Frequently asked questions
Is it worth targeting keywords with zero search volume?
Often, yes, in B2B. A tool showing zero usually means the volume is too low for it to estimate, not that nobody searches. If the term comes from your buyers, describes a problem you solve and returns results that match the page you would write, it can be worth more than a popular term your buyers never use.
How do I know a low-volume search really exists?
Look for evidence other than a volume figure: impressions for the term or close variants in Search Console, autocomplete predictions, which Google says reflect real searches, the term in your sales notes or site search, and a results page that shows other businesses answering it.
Should I create a separate page for every keyword variation?
No. Group variations that share an intent and answer them on one page. Separate pages for close variants tend to compete with each other, and Google's SEO Starter Guide says not to worry about anticipating every variation.
How do I measure results for low-volume keywords?
Measure groups of queries and pages over months, not single keywords over days. Use the Pages view in Search Console, because the rarest queries are anonymised and hidden from the query table, and judge the result against enquiries in your CRM.
Sources
- Google Ads Help: Use Keyword Planner (accessed 3 November 2026)
- Google Ads Help: About Keyword Planner forecasts, average monthly searches (accessed 3 November 2026)
- Trends Help: FAQ about Google Trends data (accessed 3 November 2026)
- Google Search Central Blog: A deep dive into Search Console performance data filtering and limits, 19 October 2022 (accessed 3 November 2026)
- Analytics Help: Enhanced measurement events, site search (accessed 3 November 2026)
- Google Search Help: How Google autocomplete predictions work (accessed 3 November 2026)
- Google Search Central: SEO Starter Guide, last updated 10 December 2025 (accessed 3 November 2026)
- Google Search Central: Creating helpful, reliable, people-first content, last updated 5 October 2026 (accessed 3 November 2026)
Turn the searches you already appear for into approved changes.
Artiweave proposes changes from your Search Console evidence, publishes only what your team approves, with automatic rollback, and reports each outcome honestly. It is in private beta.