usdigitalmarketingindex.com SEO
How should you prioritize useful keywords with very low estimated search demand?
Updated November 15, 2026
Rank them by what a visit is worth and what the page costs, and let the volume estimate break ties only. For a keyword with very low estimated demand, the estimate is too coarse to sort by: at this scale, a reported ten and a reported thirty may reflect rounding and modeling as much as demand. Three things can be judged with more confidence, and they should carry the decision: how close the searcher is to something your organization provides, how much work the answer requires, and whether the page completes a subject your site already covers.
Evaluating conversion relevance
Ask what the person searching would do next if the page answered them well. Grade the keyword on that, in words.
| Grade | The searcher is | Example of wording |
|---|---|---|
| Direct | Choosing a provider or product you offer | a service plus a specific constraint you can meet |
| Adjacent | Solving a problem your offer addresses | a symptom or task that precedes a purchase |
| Supportive | Learning about the field | a definition or background question |
| Unrelated | Unlikely ever to need you | a hobbyist or student question outside your offer |
A direct keyword with a tiny estimate can outrank a supportive keyword with a large one, because a small number of the right visitors is worth more than a crowd with no reason to stay. Where you have first-party data, such as which pages precede inquiries, use it to check the grades.
Estimating coverage effort
Effort is the other half of the ratio. Sort each keyword into one of three sizes.
- A paragraph or section added to an existing page
- A new page that reuses research you already have
- A new page that needs original research, review, or assets
A low-demand keyword can be well served by the first size. A section costs little, keeps related answers together, and avoids creating a thin standalone page. Reserve new pages for questions that need room and that no existing page can hold.
Considering strategic authority
A third consideration applies to sites that cover a subject in depth. Google’s self-assessment questions for content include whether it provides “a substantial, complete, or comprehensive description of the topic” (Google’s guidance on helpful, people-first content). A low-demand question can be the missing piece that makes your treatment of a subject complete for a reader who arrived on a broader page.
Judge this in terms of the reader, and avoid treating it as a ranking formula. The test is concrete: would a reader of your main page on the subject reasonably expect this question to be answered somewhere on the site, and is there a natural place to link to it? If yes, the keyword gains priority. If the question is an isolated curiosity, it does not.
A prioritization scorecard
Combine the three judgments in a simple score. The table below is an illustration. Every keyword and every score in it is an assumption made up to show the arithmetic, and none reflects performance data.
Scoring assumptions: relevance is scored 3 for direct, 2 for adjacent, 1 for supportive, 0 for unrelated. Effort is scored 3 for a section, 2 for a page from existing research, 1 for a page needing original work. Completeness is scored 1 if it fills a gap in an existing subject, otherwise 0. The total is the sum.
| Keyword (illustrative) | Relevance | Effort | Completeness | Total |
|---|---|---|---|---|
| service for a specific building type | 3 | 2 | 1 | 6 |
| symptom customers report before calling | 2 | 3 | 1 | 6 |
| niche regulation question | 2 | 1 | 1 | 4 |
| history of the technique | 1 | 2 | 0 | 3 |
| unrelated hobby question | 0 | 3 | 0 | 3 |
Three rules turn the totals into a queue.
- Drop every keyword with a relevance score of 0, whatever its total. The last row scores 3 only because it is cheap.
- Work from the highest total down. Among equal totals, take the one with higher relevance first, then the one with the higher volume estimate.
- Set a cutoff that matches your capacity for the period, and revisit the rest next period.
The weights are a starting point. If your organization values completeness more, score it out of 2. What matters is that the assumptions are written down, so two people scoring the same list reach the same order.
This scorecard is for keywords whose existence is already established. Finding and validating candidates that tools report as zero is the earlier step, covered in the guide to discovering long-tail queries that keyword tools label as zero search volume.
The threshold to remember: relevance decides whether a low-demand keyword is considered at all, effort decides its form, and the estimate is the last thing consulted.
Add your comment to this answer