Keep Keyword Research Evidence With an Illustrated Edition

Keep keyword research as an evidence record rather than a final list of phrases alone. Save the research inputs, marketplace context, dates, and the reason each accepted term fits the actual illustrated edition. Link important claims to relevant chapters or image-supported tasks. When artwork or scope changes, that record helps identify which discovery language needs reconsideration instead of rerunning the entire process from memory.

Preserve the question that produced each research batch

Write the reader need you are investigating before collecting keyword candidates. An illustrated guide to identifying common garden structures may involve different questions from a decorative book of garden photographs. Similar vocabulary can hide different intentions. Record the intended audience, use case, and edition context with the research batch so the resulting phrases can be interpreted correctly later.

Save the tool inputs and the date of collection, including any filters that shape the result. A list detached from its context can be difficult to evaluate after the manuscript changes. If the tool provides estimated measures, retain their definitions and limitations where available. The evidence should show what was observed or modeled rather than turn an unexplained score into a permanent fact about reader demand.

Explain why a candidate fits the book

For each serious candidate, write a short relevance note tied to the manuscript. Identify the chapter, example, or visual task that supports it and assess whether the topic is central or incidental. A phrase about diagram-based instruction should correspond to diagrams that actually explain the method. The presence of a few attractive images does not by itself support every term associated with visual learning.

Record reasons for rejection as well as acceptance. A candidate may imply a different level, a missing format feature, or a specialist subject the book only touches briefly. Keeping those reasons prevents the same misleading phrase from being reconsidered repeatedly just because it appears near the top of a later tool report. The record should preserve editorial judgment alongside the raw suggestions.

Separate shared subject terms from edition-specific claims

Mark phrases that describe the underlying subject separately from those that depend on a particular edition's presentation. The subject may remain constant across ebook and hardcover versions while the artwork or page interaction differs. A candidate tied to a specific physical use should be evaluated against that actual product rather than approved for every format through a shared spreadsheet row.

Use the accepted asset list and current manuscript to verify visual feature descriptions. Do not rely on draft artwork counts or a production plan that has not been completed. If the edition changes from one treatment to another, note the affected research decisions. This allows a targeted review of the terms that depended on the old presentation while preserving still-valid subject research.

Keep metadata selection distinct from research storage

Maintain the full candidate record separately from the approved values used in the publishing workflow. A rejected suggestion should not accidentally enter the final fields because it remains in the same undifferentiated list. Label accepted selections with the reviewed edition, date, and applicable current requirements. The author should be able to identify the submission-ready values without interpreting every historical note again.

Check current KDP metadata guidance before applying the selections. A relevant research phrase may still need to be handled differently in a particular field, and the tool's output does not establish compliance. Preserve the final selection rationale and any adjustments made during the platform review. This creates a traceable path from exploratory research to the actual metadata the author chose to use.

Revisit the evidence when the book changes

When a chapter or illustration set is revised, use the relevance notes to identify affected terms. Reassess their fit against the new material and retain the reason for any replacement. A change in one example may leave the central positioning intact, while removal of a major visual method can invalidate a key phrase. The evidence map helps distinguish these cases without treating every revision as a complete research restart.

Continue to separate observed reader behavior from assumptions about what caused it. A phrase's inclusion does not prove it produced a sale, and a tool's estimate does not guarantee future visibility. Use the record to improve accuracy and make future decisions understandable. Its value is the relationship it preserves between the research, the author’s judgment, and the illustrated book readers actually receive.