Amazon is a search engine. Most people think of it as a store, but when a reader wants a book on intermittent fasting, or managing remote teams, or writing better emails, they type that into the Amazon search bar and buy from the results that appear. If your book does not rank for the searches your potential readers are making, it will not sell — regardless of how good it is.
Keywords are the mechanism that connects your book to those searches. They appear in your title, subtitle, book description, and the seven backend keyword fields that only Amazon's algorithm sees. Get them right and your book is discoverable. Get them wrong and it exists in a category with no organic traffic, or it ranks for searches that nobody is actually making.
Keyword research is not the most exciting part of KDP publishing. It is, however, one of the most important — and one of the most commonly skipped. Authors who skip it publish into the void and wonder why no one is buying. Authors who do it well find their books ranking within 30–60 days and generating steady organic sales long after launch.
Why Keywords Determine Discovery
Amazon's search algorithm (A9) ranks books based on a combination of relevance and sales velocity. Relevance is determined primarily by whether your book's metadata — title, subtitle, description, backend keywords — matches what the reader searched for. Sales velocity is driven by review count, recent sales, and click-through rate from search results.
For a new book with no sales history, the primary lever available to you is relevance. You cannot manufacture sales history before launch, but you can optimise your metadata to ensure that when a relevant search happens, your book appears as a candidate. Strong keyword targeting in the title and subtitle makes your book relevant in Amazon's eyes before a single sale occurs.
The keyword also influences your category placement, which determines which bestseller lists you can rank on. A book that ranks in the top 20 of a well-chosen subcategory will display an "Amazon Bestseller" badge even with modest sales volume — which dramatically increases click-through rate from search results and drives more sales, creating a self-reinforcing loop.
Where Keywords Go on KDP
The Research Process
Good keyword research for KDP starts on Amazon itself, not in a third-party tool. Third-party tools are useful for validation and deeper data, but Amazon's own search autocomplete is the most direct signal of what buyers are actually typing.
- Start with your core topic. Open Amazon's book search and type the two or three words that most directly describe your book's subject. Note every autocomplete suggestion that appears — these are real searches that real buyers are making frequently enough for Amazon to surface them.
- Expand with variations. Add qualifying words to your core topic and observe the autocomplete again: "for beginners," "for women," "step by step," "guide," "workbook," "2025," and so on. Each variation that autocompletes is a separate keyword opportunity.
- Study the top-ranking books. Search your best candidate keywords and open the top three or four results. Read their titles, subtitles, and descriptions. The keywords they are using — especially if the same phrases appear across multiple top-ranking books — are validated winners for that search.
- Check their BSR. Note the Amazon Best Seller Rank (BSR) of the top results. A book with a BSR below 100,000 in the Books category is selling at least one copy per day. Books with BSRs above 500,000 indicate low demand in that keyword niche.
- Validate with a tool. Use a KDP keyword tool (see below) to confirm estimated monthly search volume and competition level for your shortlisted keywords before finalising your choices.
Evaluating a Keyword
Not every keyword with autocomplete suggestions is worth targeting. You are looking for the combination of four signals:
Are buyers actually searching this term? Autocomplete appearance is a positive signal. Tool-reported search volume above a minimum threshold (varies by tool) confirms it. A keyword with no autocomplete and no tool data has no meaningful search volume.
How many books are competing for this keyword, and how strong are they? Check the number of results for the search and the review counts of the top 10 results. If the top 10 all have 200+ reviews and established publishers, ranking as a new entrant is very difficult.
Are searchers looking to buy a book, or just to find information? "How to manage anxiety" may have high search volume but low book-buying intent. "Anxiety management book" or "workbook for anxiety" signals explicit book purchase intent — these convert better.
Does your book genuinely serve this search? A keyword with excellent demand and low competition is worthless if the readers who find your book discover it does not match what they searched for. Poor relevance produces reads that end early (bad for KU earnings) and negative reviews.
The sweet spot is a keyword with confirmed demand, manageable competition (top results with under 50 reviews is a strong signal), clear buyer intent, and direct relevance to your book's actual content. These keywords exist in almost every niche — finding them is the work.
Tools That Help
The Long-Tail Advantage
New KDP authors consistently make the same mistake: they target the broadest, highest-volume keyword in their category and wonder why they cannot rank. "Weight loss book" has enormous search volume and competition so intense that no new author can break into the top results. "Weight loss for women over 50 with hypothyroidism" has far lower volume and nearly no competition — and a book that targets it precisely will rank within weeks and own that result.
Long-tail keywords — three to six word phrases that are highly specific — are where new KDP books find their initial readership. The individual search volume is lower, but the conversion rate is higher (because the searcher knows exactly what they want and your book is exactly that), the competition is lower, and once you rank for several long-tail terms, your overall sales velocity improves enough that you start ranking for the broader head terms as a secondary effect.
Keyword Mistakes to Avoid
- Repeating title keywords in backend fields. Amazon already indexes your title keywords. Using the same terms in your seven backend fields wastes those slots. Use the backend for keyword variations, long-tail phrases, and related terms that could not fit naturally in your title or description.
- Using competitor book titles as keywords. Amazon's terms of service explicitly prohibit using competitor book titles, author names, or ASINs as keywords. Violating this can result in your book being suppressed or your account being flagged.
- Targeting keywords you cannot rank for at launch. A brand-new book with zero reviews targeting "personal finance" will rank on page 50 where no one will see it. Target keywords where the current top results have 20 reviews or fewer — then use the sales velocity from ranking there to build toward broader terms.
- Setting keywords once and never revisiting. Amazon search behaviour changes. Competitor books come and go. A keyword that was excellent at launch may be saturated six months later, or a new long-tail variation may have emerged. Review your keyword performance quarterly and update accordingly.
- Ignoring the subtitle. Many authors treat the subtitle as an afterthought. It is one of the highest-weight keyword placements you have and the most visible benefit statement in search results. A subtitle that includes a well-researched keyword phrase while clearly communicating the book's benefit is one of the highest-ROI optimisations available at zero cost.
For the pricing decisions that pair with keyword strategy to maximise revenue, see my guide on KDP book pricing strategy. For the launch sequence that activates your keyword rankings fastest, see the guide on how to launch a KDP book.
I help authors identify the right keyword targets before they write — so the book is positioned for discovery from day one, not retrofitted after launch.