Amazon KDP gives you seven keyword slots, two category choices, and one blurb. Those are the only pieces of metadata that decide whether readers ever find your book. Here is the workflow we recommend before you list.
Step 1: find categories, not keywords, first
A category is a room; keywords are the door signs. Start on Amazon’s Kindle store, drill into the sub-sub-sub-category that matches your book, and note the top ten bestsellers. Their ranks tell you how much competition you’re walking into.
Step 2: mine the top-ten titles and subtitles
Amazon indexes title, subtitle, series name, author name, and the seven keyword slots. Everything after that (blurb, book text) is discovery-adjacent. So the words in the top-ten titles are the words Amazon’s search algorithm already associates with sales in your niche. Copy them into a spreadsheet.
Step 3: layer in long-tail phrases
Type your genre into Amazon’s search bar and stop — do not press enter. The autocomplete suggestions are queries with real search volume. “Cozy mystery small town bakery,” “enemies to lovers hockey romance,” “clean regency romance dukes.” These long-tail phrases are the ones that convert.
Step 4: assign one phrase per slot
Do not repeat words across the seven slots — Amazon deduplicates. Each slot is a chance to add a new discovery term. Prioritise phrases that appear in at least three top-ten titles.
Step 5: check every 90 days
Reader search behaviour drifts. Set a calendar reminder to revisit your keywords quarterly. A ten-minute refresh can move rankings more than a whole new book launch.