Before relying on a keyword number, ask what it measures, where it came from, when it was collected, and whether it describes the relevant marketplace. Keep estimates separate from observations and from invented examples. Use the phrase’s fit with the manuscript as a necessary check regardless of how impressive the number appears.
Ask what the number actually represents
A research table can display precise values under headings such as volume, demand, or competition without defining them. Begin by identifying the quantity being measured. Does the number describe a direct observation, a model estimate, a provider’s internal score, or something else? A label that sounds familiar does not establish that two tools use the same definition.
Write the definition beside the figure in your research record. If the provider cannot explain it, mark the value as uninterpretable for the decision at hand. This does not prove that the number is fabricated. It means you lack the information needed to use it responsibly. Avoid ranking candidate phrases by unexplained values simply because a sorted spreadsheet creates the appearance of analytical certainty.
Recover the scope and collection context
Ask for the marketplace, language, period, and collection date associated with the data. Determine whether the figure concerns the exact phrase or a broader grouping. Record any known exclusions. A value about one context may offer little evidence for a different audience or edition, and a copied table can lose the context that originally made its entries meaningful.
Consider two candidate phrases with similar-looking numbers from different sources. One may be an estimate across a broad period, while the other may be a short observation under a narrower definition. Treating them as directly comparable can create a false preference. Preserve their differences and decide whether the evidence supports the comparison you want to make before selecting the apparently larger opportunity.
Distinguish AI-generated suggestions from measured data
An AI tool can propose reader questions and candidate phrases, but the presence of numbers in its response does not establish that it measured them. Ask for the actual data source and method when a numerical claim matters. Follow the evidence to its origin rather than accepting a citation-shaped label or a confident explanation as proof that the table came from a real observation.
If the response uses hypothetical numbers to illustrate a method, keep them clearly labeled and out of the live decision table. If the source cannot be verified, retain the phrase as a possible editorial idea while removing the unsupported quantitative claim. This preserves potentially useful brainstorming without allowing invented precision to determine the book’s audience, title, or scope.
Check relevance independently of the metric
For each phrase under consideration, identify the reader question it represents and the manuscript passages that answer it. Review the implied skill level, subject, and outcome. A large number cannot make an unrelated phrase accurate. Likewise, an uncertain estimate does not automatically make a well-matched phrase unusable as an editorial candidate; it limits what you can claim about its measured popularity.
Suppose a workbook teaches basic planning for a small community event, while a proposed phrase suggests professional event management. Even a well-documented metric would not solve that scope mismatch. Revise the candidate or reconsider the manuscript deliberately. Consult current KDP metadata guidance before using the final wording, and keep the book’s actual content central to the selection.
Record a bounded decision
Summarize what the evidence supports for each candidate. You might retain a phrase because it clearly describes the reader task, while noting that its supplied volume estimate remains unverified. Another phrase may be rejected because its meaning does not fit the book, regardless of the metric. These decisions are more useful than a simple score that hides both relevance and uncertainty.
Revisit important data when the context or manuscript changes. Keep the original source details so the next reviewer can tell whether a new figure uses the same method. Do not translate a research score into a guaranteed ranking or sales prediction. A careful keyword record helps the author explain why the metadata is relevant and what evidence informed it, while remaining honest about what the numbers cannot establish.
Frequently asked questions
Clear answers for this publishing decision.
Does a very precise number mean the estimate is reliable?
No. Precision in presentation does not establish measurement quality. Examine the definition, source, method, scope, and uncertainty. A rounded value with clear context may be more useful than a detailed figure with no traceable basis.
Can I keep AI-suggested keywords if their metrics are unsupported?
Keep them as candidates and review their meaning against the manuscript. Remove or clearly separate unsupported numerical claims. Brainstorming a relevant phrase and measuring its demand are different tasks that require different evidence.
Should I choose only the phrase with the highest score?
Only compare scores whose definitions and context make the comparison meaningful, and still check relevance. A score cannot replace the editorial judgment that the phrase accurately represents the book and complies with current metadata requirements.
What should the title, subtitle, description, and keywords look like before I upload?
Check the mechanical limits first, then make sure title, subtitle, categories, and description describe the same book for the same reader; a keyword is useful when it matches how that reader searches, not because it is popular.
Draft the listing fields in a book project