Amazon KDP Keywords Guide for Authors (2026)
Good KDP keywords help the right readers find your book. Bad ones make your metadata look spammy, confuse search intent, and weaken the product page before anyone even clicks.
Short answer: choose keyword phrases that match the book's real audience, genre, problem, or promise. Keep each field specific, distinct, and reader-facing rather than stuffing broad or misleading terms everywhere.
Amazon KDP's help guidance, checked July 20, 2026, says authors can add up to seven keyword fields and recommends phrases that accurately describe the book. That matters because keywords are only one part of discovery. They work best when the title, subtitle, categories, cover, description, and actual manuscript all support the same reader promise.
If your manuscript is still moving, finish the positioning work inside ShakespeareAI's book writer before you lock in metadata. Cleaner drafts produce cleaner keyword choices.
What KDP keywords actually do
KDP keywords are not magic ranking buttons. They are signals that help Amazon understand what your book is about and which searches are most relevant. The keyword fields matter most when they reinforce the rest of the listing rather than trying to overpower it.
Think of them as search-context labels:
- What kind of book is this?
- Who is it for?
- What reader problem, trope, or subject does it address?
- Which phrase would a real buyer type before seeing your cover?
Important: a keyword field should help Amazon classify the book more clearly. It should not be used to sneak in unrelated search traffic, competitor names, or claims the manuscript cannot support.
How to choose stronger KDP keyword phrases
1. Start from reader intent, not from vanity terms
Many authors choose keywords they personally like instead of the phrases readers actually use. Broad single words like fantasy, romance, or business rarely tell enough story on their own. Phrases work better because intent is clearer.
Instead of chasing maximum breadth, ask:
- What would a reader search if they wanted this exact type of book?
- What problem or emotional payoff are they looking for?
- Would this phrase still make sense if they saw it next to my cover and description?
2. Use phrase-level specificity
Amazon's metadata guidance points authors toward keyword phrases rather than vague standalone words. In practice, that means looking for 2-3 word combinations that reflect a real search pattern.
| Weak keyword | Stronger keyword phrase | Why it works better |
|---|---|---|
| romance | small town romance | Clearer subgenre and reader expectation. |
| fantasy | epic fantasy adventure | Adds tone and scope. |
| productivity | time management for teachers | Targets a specific nonfiction audience. |
| mystery | cozy mystery with cats | Matches a concrete browsing pattern. |
3. Give each keyword field a different job
Do not burn all seven fields on near-identical phrasing. Spread them across meaningful angles:
- Primary genre or subject
- Specific trope, setting, or audience
- Problem-solution phrasing for nonfiction
- Comparable reading intent without naming other authors
- Format or use case when it genuinely matters
This is the same principle covered in AI book metadata generator: distinct fields make review easier and help your listing say more with less repetition.
4. Match the keyword to the manuscript you actually wrote
Misaligned keywords are one of the fastest ways to create a weak product page. If your book is a quiet character drama, do not optimize like it is a high-concept thriller. If your nonfiction book is beginner-level, do not signal an advanced professional manual.
The question is not whether a phrase gets searches. The question is whether the right reader would still feel satisfied after clicking, reading the description, and buying the book.
Sample keyword frameworks by book type
Fiction
Fiction keywords usually work best when they combine genre with a recognizable reader hook.
- Subgenre:
paranormal romance,military sci fi - Trope:
enemies to lovers,chosen one fantasy - Setting:
Victorian mystery,small town romance - Tone or promise:
dark academia thriller,feel good women's fiction
Related reading: how to outline a mystery novel with AI, how to write a romance novel with AI, and AI fantasy book generator workflow.
Nonfiction
Nonfiction readers usually search around outcomes, roles, or problems to solve. That means keyword phrasing often follows a pattern like topic + audience or problem + method.
lead generation for coachesmeal prep for busy parentsmindfulness for teensbudgeting for freelancers
If you are still shaping the offer, start with how to generate a nonfiction book outline with AI so the keywords reflect a clear chapter promise.
What to avoid in KDP keyword fields
Amazon's metadata guidance warns against keywords that mislead or manipulate customers. In practice, avoid:
- Other author names or book titles
- Promotional language like
freeorbestselling - HTML or formatting tricks
- Unrelated categories or trend-jacking phrases
- Overly generic filler repeated across fields
The safest rule is simple: if a phrase would make a reader feel tricked after opening the product page, it does not belong in your metadata.
A practical seven-field workflow
Draft keywords in this order
- Primary genre or subject phrase
- Reader audience phrase
- Trope, setting, or use-case phrase
- Problem-solution phrase
- Emotional or outcome-led phrase
- Adjacent discovery phrase
- One backup phrase you can test later
After drafting the list, compare it against your title, subtitle, categories, and description. If the same concept is already doing heavy lifting elsewhere, you may want a different keyword field to fill a gap rather than repeat it.
How KDP keywords fit into the rest of your metadata stack
Keywords should not be chosen in isolation. They work alongside:
- Your title and subtitle
- Your KDP categories
- Your book description
- Your cover's genre signal
- Your manuscript's actual positioning
That is why this article pairs naturally with how to create KDP metadata for an AI-assisted book, how to write a book description for Amazon, and KDP book formatting with AI. Publishing quality comes from the full system, not one field.
When to update your keywords
Update your KDP keywords when you have a better understanding of the book's audience, positioning, or product page language. Examples include:
- After revising the subtitle
- After learning which reader promise is clearest
- After tightening the description
- After realizing the current fields are too broad or repetitive
Do not update them just to chase random trends. The goal is a more accurate listing, not constant metadata churn.
Build the draft before you optimize the listing
Use ShakespeareAI to shape the manuscript, chapter flow, and export-ready files first. Then turn that cleaner draft into sharper KDP metadata and a more coherent publishing workflow.
Start writing with ShakespeareAIFAQ
Do KDP keywords need to be exact-match search terms?
No. They should reflect real reader intent, but they do not need to read like an awkward exact-match SEO string.
Should fiction authors use trope keywords?
Yes, when the trope is a real part of the book and matches the reading promise. Tropes can help fiction readers self-select quickly.
Can I use the same keyword idea in the title and in the keyword fields?
Sometimes, but avoid wasting multiple fields on the same phrasing. Use the extra space to add context the title does not already communicate.
What matters more: keywords or description?
They do different jobs. Keywords help discovery context; the description helps conversion after the click. You need both.