Why the volume column is often empty
The keyword research hub explains where query data comes from and that every Search VolumeThe estimated number of times a query is searched in a period, usually per month. figure is an estimate. This guide starts where those estimates run out. In general practice, there are four common reasons the column is blank or shows a token figure:
- The query is genuinely rare. An engineer looking for a replacement seal for a particular pump series, or a procurement manager checking a compliance standard, types a phrase few other people type. Estimates built from samples cannot see it, though the searcher is real.
- The words are new. A category that did not exist last year has no history to model. Buyers may not yet agree on what to call it, so demand is scattered across several phrasings, each too small to register.
- The market is small. A country with a small population, or a language with few speakers online, gives tools thin samples to work from, so even common questions there show little or nothing.
- Variants are grouped or split. Tools may fold close variants into one figure or report only the most common phrasing, so the phrasing your customers use can look empty while a sibling carries the number.
Each reason calls for the same response: stop treating the figure as the decision and find other evidence that people need the answer.
What low volume does and does not mean
A blank or tiny figure supports fewer conclusions than teams tend to draw from it.
| A low or missing figure… | Does mean | Does not mean |
|---|---|---|
| For a specialist product phrase | The tool cannot estimate it reliably | That no buyer searches it |
| For a new category name | The vocabulary is unsettled or recent | That the problem it solves is not searched |
| In a small country | The sample there is thin | That local buyers search less than elsewhere |
| For a long, specific question | Few people phrase it exactly that way | That one page could not serve many such phrasings |
| Next to a large figure for a broad term | The broad term is more common | That the broad term's searchers are your buyers |
The last row matters most in business-to-business markets. A broad term can carry a large number while mixing students, hobbyists, job seekers and buyers from unrelated industries; a Long-Tail KeywordA specific, usually longer query searched rarely on its own but often as a group. with no measurable volume may be typed only by people with a budget and a deadline. The intent classification guide explains how to label what each query wants (Search IntentThe goal behind a search query: to learn, to find a site, to compare or to act.); in a niche market that label, and the type of person behind it, usually outweighs the number.
What low volume does honestly signal is a ceiling: a page aimed at a rare problem will not bring large traffic. That is acceptable when each visit is valuable, and it is the reason to judge these pages by enquiries rather than sessions.
Sources when tools show nothing
When the tools go quiet, demand has to be inferred from places where buyers describe their problems in their own words. Google's SEO Starter Guide encourages site owners to think about the words people would use to look for their content, including the different words used by people who know a topic well and people who are new to it. The sources below are where those words are recorded.
- Sales calls and enquiry emails. The questions prospects ask before they buy, in their phrasing. Read a batch of recent enquiries and note every question, objection and comparison verbatim.
- Support tickets and chat transcripts. What customers struggle with after buying, which often repeats what others search before buying.
- Customer interviews. The only source that explains how someone came to look for help and what they typed first; the customer interviews tutorial sets out how to run them, and the audience research for content guide turns the findings into a content map.
- Communities and forums. Professional groups, trade forums and question sites show the vocabulary of people who have not yet learned the industry's terms.
- Internal site search. What visitors already on the site ask for, including product codes and phrases missing from navigation.
- Adjacent queries with measurable demand. The broader problem, the standard, the material, the application or the competing method often carries a figure where the specific phrase does not, and shows the direction of interest.
- The results pages. Searching each candidate phrase shows how the engine interprets it: whether it returns supplier pages, technical documents, forums or nothing relevant at all. A results page filled with near misses is evidence of a gap, not of absent demand.
- Your own Search Console data. Queries the site already appears for, however few the impressions, are real demand rather than estimates. Filter the performance report to pages close to the topic and read the queries listed.
Sorting these by Buyer PersonaA research-based profile of a typical customer type: goals, problems and buying behaviour. keeps the list honest: a phrase that only the marketing team uses, and no buyer, is a candidate to drop however neatly it fits the product.
Writing for problems rather than exact phrases
When individual phrasings are too rare to measure, the unit of research becomes the problem, not the keyword. A buyer with a valve that leaks under thermal cycling may type a dozen different things; a single page that answers that problem thoroughly can be found for all of them. The keyword clustering guide groups queries by results-page overlap; in low-volume markets, where overlap is hard to measure, grouping by the underlying problem is the practical substitute.
This is also what the SEO Starter Guide describes when it advises writing for readers rather than for lists of keywords. In practice it means:
- Name the problem in the buyer's words in the title and opening, using the phrasing from sales and support rather than the internal product name.
- Cover the variants naturally. Include the synonyms, the standards, the part numbers and the application terms a specialist would expect, because each is a way in for a different searcher.
- Answer the questions that follow. What causes it, how to check, what the options are, what each costs in effort and risk, and when to call a supplier. Rare queries often come from people far into a decision.
- Show expertise the reader can check. Specifications, drawings, test conditions and worked examples are what a specialist reader uses to decide whether the supplier knows the subject; the B2B content guide covers proof and expertise in depth.
Validating demand after publishing
In a low-volume market, the page is the experiment. Once it is indexed, Search Console's performance report (GSCGoogle's search performance and indexing tool.) shows the queries it appeared for, with impressions, clicks, click-through rate and average position, filterable by page, query, country and device. That is direct evidence of demand that no tool estimate provides.
Three things make the reading reliable:
- Read by page, not by query. Search Console's documentation explains that very rare queries are anonymised: they are left out of the query table to protect searchers' privacy, though they still count in the chart totals unless you filter by query. A niche page may therefore show more impressions in total than its visible queries add up to. The page total is the better measure.
- Watch the query list for surprises. The phrases the page actually appears for are the market telling you its vocabulary. New phrasings belong in the page's headings or in a new page if they reveal a separate problem.
- Join it to enquiries. Impressions show that people searched; enquiries show that the right people did. Track form submissions, calls and quote requests from organic landings on each page, so a page with few impressions and good enquiries is recognised as a success.
A valve maker builds pages from sales-call questions
Gantryfold Valves, an invented maker of industrial valves, found that keyword tools showed no volume for any of its product terms. Its sales team logged the questions prospects asked over one quarter, and the marketing team grouped them into eight problems, such as seat leakage after thermal cycling and valve selection for abrasive slurries. Each became a page written in the customers' phrasing, with test conditions and drawings. After four months, the eight pages together showed 3,400 impressions in Search Console, of which only 1,150 were attached to visible queries; the rest came from anonymised queries. The slurry selection page led with 1,100 impressions and 9 quote requests; the thermal cycling page had 260 impressions and 6 quote requests; two pages had under 100 impressions and no enquiries after four months and were merged into a broader maintenance page. Across the eight pages, organic landings produced 21 quote requests in the period, against 4 in the same months of the previous year from the old product pages.
Prioritising without volume
Without volume, the order of work has to come from the business and from the evidence gathered above.
| Factor | Question to ask | Where the answer comes from |
|---|---|---|
| Value to the business | What is a customer with this problem worth, and does the business want that work? | Sales, finance, the product team |
| Evidence of need | How often does the question come up in sales, support and communities? | The research log |
| Fit with intent | Does the results page show the searcher wants what the business offers? | Reading the results page |
| Competition on the page | Do the current results answer the problem well? | Reading the results page |
| Effort | Can an expert answer it in a page, or does it need testing, drawings or new data? | The team that would write it |
Score each candidate problem on these factors and work from the top. The SEO strategy framework keeps the choice tied to what the business needs from search; in niche markets that is usually qualified enquiries rather than traffic, and the priority list should say so. Revisit the scores once Search Console data arrives, because the first pages published are the best evidence for ranking the rest.
Common mistakes
- Treating a blank figure as proof of no demand. It is the tool's limit, not the market's.
- Chasing the broad term with volume when its searchers are students, hobbyists or another industry.
- Writing one page per rare phrasing. Many thin pages competing for the same problem, each too small to matter.
- Using the company's vocabulary instead of the buyer's. Product names and internal jargon that nobody outside the business types.
- Reading Search Console by query only. Anonymised queries hide much of a niche page's demand; read the page totals.
- Judging the pages by sessions. A handful of visits that turn into enquiries is the result these pages are for.
- Never revisiting the list. The queries a page actually earns are better evidence than the research that created it.
