The Complete Guide to Amazon KDP Keyword Research

Keywords are the invisible infrastructure of your book’s discoverability on Amazon. Get them right and readers who have never heard of you will find your book when they search for exactly what you have written. Get them wrong — or use the same vague single-word approach that most first-time authors use — and your book will be effectively invisible to everyone except people who already know it exists.

This is the guide to getting them right. It covers the full keyword research process from first principles through to specific tactics, with dedicated sections for fiction and non-fiction. Every topic links to a deeper dive when you are ready to go further. Bookmark it and work through it before you publish your next book.

What KDP Keywords Actually Are

When you publish a book on KDP, you complete a metadata section that includes seven keyword fields. These fields are invisible to readers — they appear nowhere on your book’s product page. They exist purely to tell Amazon’s search algorithm what searches your book should appear in.

Each field accepts up to 50 characters. You can use the full 50 characters as a single long phrase, or as multiple shorter phrases separated by spaces — Amazon reads both approaches correctly. You do not need commas between words in the same field, and you should not use commas, as they may be read as separators and waste character space.

The seven fields together give you 350 characters of search positioning information. Used well, this is substantial. Used poorly — filled with single-word keywords that are already covered by your title and categories — it is almost worthless. For a full breakdown of how each field works, see our article on the 7 KDP backend keyword fields explained.

How Amazon Uses Keywords to Rank Books

Amazon’s search algorithm — now commonly referred to as A10 — uses keywords as one of several signals to determine which books to show in which positions when a reader searches. The algorithm reads your backend keyword fields, your title and subtitle, your book description, and your category placement to build a relevance profile for your book.

When a reader types “enemies to lovers small town romance” into Amazon’s search bar, the algorithm scans its catalogue for books whose metadata matches that search query. A book whose metadata — across all fields — accurately and specifically matches that phrase will rank higher than a book that only partially matches it.

Crucially, the algorithm also factors in conversion rate: how often readers who see your book in search results actually purchase it. A book that ranks for a keyword but does not convert signals poor relevance. Over time, high conversion rate is a stronger ranking signal than keyword presence alone. This is why your keywords, your description, and your cover must work together — keywords bring readers to your page; everything else converts them.

The Seven Backend Keyword Fields

KDP provides seven separate keyword input fields, each accepting up to 50 characters. Amazon treats each field as a distinct input and indexes all of them. The most effective approach is to treat each field as a slot for a different keyword phrase or cluster of related phrases.

Best practice for filling the seven fields:

• Use all seven fields — leaving any blank wastes indexing potential
• Do not repeat keywords across fields — each field should add new terms not covered elsewhere
• Do not repeat words already in your title, subtitle, or categories — Amazon already indexes those; repeating them in keywords wastes character space
• Use multi-word phrases, not single words — “cosy mystery English village” is infinitely more targeted than “mystery”
• Fill each field as close to 50 characters as possible without going over — unused characters are unused opportunity

For the complete field-by-field strategy, see our dedicated article on how to choose your 7 KDP backend keywords.

Why Long-Tail Keywords Outperform Broad Ones

A long-tail keyword is a specific multi-word phrase — “amateur sleuth cosy mystery retired teacher English village” — as opposed to a broad single term like “mystery.” The logic behind prioritising long-tail keywords is straightforward:

Less competition. Thousands of books are indexed for “mystery.” A handful are indexed for “amateur sleuth cosy mystery retired teacher English village.” Ranking in the top ten for the long-tail phrase is achievable. Ranking in the top ten for “mystery” against thousands of established titles is not.

Higher purchase intent. A reader who searches for “amateur sleuth cosy mystery retired teacher English village” knows exactly what they want. The reader who searches for “mystery” is browsing. The long-tail reader converts at a significantly higher rate because the specificity of their search reflects the specificity of their preference.

Better reader-book matching. Long-tail keywords bring readers who will love your book, not just readers who might be vaguely interested in your genre. Better matching produces better reviews, higher read-through, and stronger word-of-mouth.

For a full breakdown of long-tail keyword strategy, see our article on long-tail keywords for KDP: why niche beats broad.

How to Research Keywords

Effective keyword research combines several methods. No single method gives you the complete picture — the best keyword sets come from using multiple research approaches and synthesising what you find.

The primary methods, covered in detail in the sections below, are: Amazon’s own autocomplete search; competitor keyword analysis; keyword research tools; reader language research (reading reviews of comparable books); and category browsing to understand what terms readers use to navigate to your type of book.

Using Amazon Autocomplete

Amazon’s search bar autocomplete is one of the most reliable free keyword research tools available. When you type a partial search query into Amazon’s search bar, the autocomplete suggestions that appear represent real searches made by real readers — Amazon’s autocomplete is populated by actual query data, not guesswork.

The process:

1. Go to Amazon.com (or your primary marketplace) and click into the search bar. Make sure “Books” is selected in the department dropdown.
2. Type the beginning of a keyword phrase relevant to your book. For example: “cosy mystery” — and note every autocomplete suggestion that appears.
3. Add different words after your initial phrase and note the new suggestions. “cosy mystery with” — “cosy mystery set in” — “cosy mystery featuring.”
4. Work through synonyms and variant phrasings. If you write “cosy,” also search “cozy” (the US spelling). If you write romance, also search “love story,” “romantic fiction,” “contemporary romance.”
5. Record every suggestion that accurately describes your book. These are your candidate keywords.

For the full autocomplete research methodology, see our article on how to find keywords using Amazon autocomplete.

Analysing Competitor Keywords

You cannot directly see a competitor’s backend keywords — they are invisible on Amazon product pages. However, you can infer them through several methods:

Title and subtitle analysis: Authors who understand keyword strategy put their most important keywords in their title and subtitle, where Amazon weights them most heavily. Reading the titles and subtitles of the top sellers in your category reveals which keywords the most successful authors are targeting.

Description analysis: As covered in our article on using keywords naturally in your book description, descriptions are indexed by Amazon. The specific language used in competing bestseller descriptions often reflects their keyword strategy.

Search result analysis: Search for the keyword phrases you are considering using and note which books appear. If the same books consistently appear across multiple relevant searches, they are effectively ranking for those terms — and you can learn from what makes them rank by studying their metadata.

For the complete competitor analysis methodology, see our article on how to spy on competitor keywords.

Keyword Research Tools

Several dedicated tools exist to help KDP authors with keyword research, ranging from free browser-based tools to paid software with extensive data. Each has strengths and limitations:

Publisher Rocket (paid, one-time fee) is the most widely used third-party KDP keyword tool. It pulls data directly from Amazon to show estimated search volumes, competition levels, and the keywords your specific competitors are using. For authors publishing regularly, it is the most efficient paid research tool available.

Kindlepreneur’s free tools offer basic keyword suggestions and category research without the cost of a full subscription.

KDP Rank Fuel by Vappingo generates 100 targeted keyword ideas for your specific book from a single input — alongside your book description, category recommendations, and ad campaign keywords. It is the most efficient option for authors who want comprehensive metadata preparation in one place rather than managing multiple tools. Try it at app.vappingo.com.

For a full comparison of available tools, see our article on KDP keyword research tools compared.

Keywords for Fiction

Fiction keyword research focuses on several distinct categories of search terms:

Genre and subgenre terms: The specific vocabulary readers use to find your type of fiction. Not just “romance” but “contemporary romance,” “small town romance,” “second chance romance.” Not just “mystery” but “cosy mystery,” “British detective fiction,” “amateur sleuth mystery.”

Setting and period terms: Readers often search by where and when a story is set. “Regency romance,” “Victorian mystery,” “Scottish Highlands thriller,” “1920s detective fiction” are all meaningful search terms with real audiences.

Trope terms (especially for romance): “Enemies to lovers,” “forced proximity,” “grumpy sunshine,” “fake dating,” “brother’s best friend” — romance readers actively search by trope and your keywords should reflect this.

Character type terms: “Female detective fiction,” “amateur sleuth,” “ex-military thriller,” “single mum romance” — readers often search by the type of protagonist they want to follow.

Mood and tone terms: “Heartwarming romance,” “dark psychological thriller,” “laugh-out-loud cosy mystery,” “slow burn romance” — tone qualifiers attract readers whose expectation will match the actual reading experience.

For genre-specific keyword guidance, see our articles on KDP keywords for fiction and KDP keywords for non-fiction.

Keywords for Non-Fiction

Non-fiction keyword research focuses on the specific problems readers are trying to solve, the methods they are searching for, and the audience identifiers that describe who they are.

Problem-based keywords: “How to stop procrastinating,” “overcoming anxiety without medication,” “pricing strategy for freelancers” — readers search for solutions to specific problems. Your keywords should reflect the problem your book solves in the exact language your readers use.

Methodology keywords: “Cognitive behavioural therapy anxiety,” “Stoic philosophy daily practice,” “zero-based budgeting guide” — readers who have researched their problem often search by methodology.

Audience identifier keywords: “Productivity for ADHD adults,” “investing for beginners UK,” “weight loss for women over 50” — audience identifiers attract highly targeted readers whose specific situation your book addresses.

Outcome keywords: “How to write a novel in 90 days,” “double your freelance income,” “learn to meditate beginners” — outcome-focused searches capture readers who know what they want to achieve.

Choosing Your Final Seven

After research, you will have a list of candidate keywords. Selecting the best seven involves balancing two factors: relevance (does this keyword accurately describe my book?) and opportunity (is there a realistic chance of ranking for it?).

A useful framework for final selection:

• Two or three high-specificity phrases that describe your book’s exact niche with precision — these are your best chances of ranking quickly
• Two or three mid-competition phrases that are more broadly searched but still specific to your subgenre — these build discoverability over time
• One or two audience or mood qualifiers that capture reader intent rather than genre classification

Avoid: single-word keywords, keywords that repeat your title or categories, keywords that do not accurately describe your book, and prohibited keywords (Amazon’s list includes competitor names, bestseller rank claims, and quality descriptors like “best” or “free”). For the full list, see our article on prohibited KDP keywords: what not to use.

Keywords in Your Description

Your backend keyword fields are not the only place Amazon indexes keywords. Your book description is also read and indexed by the algorithm. This means keywords that appear naturally in your description contribute to your search ranking for those terms — independently of your backend fields.

Write your description to accurately describe your book using the natural language of your genre, and your most important keywords will appear organically. Do not force them in — keyword stuffing in descriptions harms conversion rates and is increasingly identified and discounted by Amazon’s algorithm. For the full strategy, see our article on using keywords naturally in your book description.

Tracking and Updating Keywords

Keywords are not a set-and-forget element of your metadata. Reader search behaviour evolves, new tropes emerge in fiction, new problems become salient in non-fiction, and seasonal trends create temporary keyword opportunities. A keyword set that was optimal at launch may underperform six months later.

Review your keywords every three to six months. The signals that suggest a keyword update is needed: declining organic search visibility, stagnant sales despite consistent advertising, or significant shifts in your category’s bestseller landscape. For the full update methodology, see our article on when to update your KDP keywords.

Common Keyword Mistakes

The most frequently made keyword errors across all KDP publishing experience levels:

• Single-word keywords. “Mystery,” “romance,” “thriller” — these generate enormous competition and zero targeted traffic. Never use single-word keywords unless they are extremely specific proper nouns.
• Repeating title and category keywords. Amazon already indexes your title, subtitle, and categories. Repeating them in your backend fields wastes 50 characters of indexing opportunity.
• Targeting only the most popular searches. The most popular keywords are the most competitive. A book with no sales history cannot rank for “romance” against established bestsellers. Start with specific, achievable terms.
• Never updating keywords after launch. The keyword set you chose before publication is a starting hypothesis. Update it as you learn which searches your book is actually appearing in and converting from.
• Ignoring the description as a keyword vehicle. Many authors treat keywords as exclusively a backend matter. Your description also ranks — write it to be accurate and specific, and the keywords take care of themselves.

For the complete mistake-avoidance guide, see our article on common KDP keyword mistakes and how to fix them.