KDP keyword research is a positioning workflow, not a promise that a book will rank, sell, or stay discoverable. The useful job is to make the book's reader, market, category fit, and metadata choices less vague before publishing.
Use this workflow before opening Publisher Rocket, KDSpy, or another author research tool. If the manuscript, reader promise, or book category is still unclear, software research will mostly create more tabs to interpret.
Paepae Stack treats author research tools as under-review publishing workflow aids. Verify current pricing, data sources, feature limits, export options, and KDP category behavior before buying. Do not treat any tool as evidence of future sales, ranking, ads performance, or Amazon visibility.
Step 1: write the reader promise first
Start by writing one plain sentence for the reader promise. This is the reason a reader would care before they see keywords.
Use this shape:
This book helps [reader] understand or change [specific problem] without [specific friction].If that sentence is hard to write, pause keyword research. A tool can show phrases and competing books, but it cannot decide what the book is supposed to be for.
Step 2: list seed phrases from reader language
Write seed phrases before checking software. Pull them from the subtitle, chapter headings, reader questions, review language from comparable books, and the problem the reader is trying to name.
Useful seed phrase sources include:
- the book subtitle;
- chapter or section names;
- reader pain points;
- common phrases in comparable book titles;
- phrases a reader might type before they know the author's brand;
- adjacent categories that describe the same reader problem.
Keep the first list short. Ten to twenty seed phrases are enough to begin. The point is not to collect every possible term. The point is to start with phrases that match the book's actual promise.
Step 3: check market fit, not only volume
When you open a research tool, use it to test whether the phrase describes a real reader path and a plausible shelf, not whether a metric looks exciting.
For each candidate phrase, check:
- Do the top competing books solve a similar reader problem?
- Would this book look natural beside them?
- Is the phrase too broad for the book's promise?
- Is the phrase too narrow or unrelated to the actual manuscript?
- Do category and competitor signals support the positioning, or do they point to a different book?
Use Publisher Rocket when the workflow needs keyword, category, competitor, and positioning research in one author-focused path. Use KDSpy when the research job is more Kindle-category and competing-book inspection. Keep both tools in the research lane; neither should be used to claim guaranteed ranking or sales.
Step 4: separate metadata candidates from ad ideas
Keyword research often creates two useful lists. One list belongs near the book metadata decision. The other belongs in later ad, description, content, or launch testing.
Use this split:
| Research output | Use it for | Do not assume |
|---|---|---|
| Close-match reader phrases | Subtitle, description, backend keyword candidates, category framing | That exact phrase will produce ranking or sales |
| Competing book patterns | Positioning, cover expectations, description angle, pricing context | That copying a competitor will work |
| Adjacent phrases | Blog posts, ads, launch content, newsletter angles | That every adjacent phrase belongs in metadata |
| Category observations | Category fit and reader shelf expectations | That category choice alone creates visibility |
Keep a short note beside each phrase explaining why it belongs. If the reason is only "the tool showed it," the phrase is not ready.
Step 5: choose a small test set
Do not turn research into an endless spreadsheet. Choose a small candidate set for the next publishing decision.
A useful first set can include:
- three to five close reader phrases;
- two to three category or shelf notes;
- three comparable books to inspect manually;
- one description angle to test;
- one list of phrases to save for ads or launch content later.
After launch, revisit the research only when there is a real reason: a new edition, category mismatch, description rewrite, ad test, or reader feedback that changes positioning.
Use this with Paepae Stack
Use this sequence when author research is the next bottleneck:
- Clarify the reader promise before opening a tool.
- Use Publishing Tools to review the current author-tool shelf.
- Review Publisher Rocket when the job is KDP keyword, category, competitor, and positioning research.
- Review KDSpy when the job is Kindle category and competing-book inspection.
- Use ManuscriptReport when the bottleneck is manuscript or reader-review analysis rather than keyword discovery.
The workflow is successful when the research narrows the book's positioning choices. It is not successful just because it produces more keywords.