Free and even open-source KDP keyword tools exist, but here is the thing nobody selling a $199 tool will tell you: almost all of them, free or paid, wrap the same free public source, Amazon’s own search autocomplete. Once you know that, you do not need MerchantWords’ subscription or Publisher Rocket’s metrics to fill your seven Kindle keyword slots well. This guide covers what KDP keyword research actually requires, the free engine every tool is built on, when the paid options are and are not worth it, and how to extend the method to the Google and AI searches where readers also discover books.
What KDP keyword research actually is
On Kindle Direct Publishing, you get seven keyword slots per book, plus your title, subtitle, and categories. Those keywords decide which Amazon searches your book can appear for, so the job is simple to state: find the phrases readers actually type when looking for a book like yours, and claim the best seven. As guides like Self-Publishing School’s overview of Amazon keyword tools stress, the goal is relevance plus real demand, a phrase enough readers search but not so broad you are buried under bestsellers.
That is the whole task. Everything a keyword tool does for KDP is in service of filling those seven slots and choosing categories with phrases that match how readers search.
The free engine behind every tool: Amazon autocomplete
Here is the open secret. When you start typing in Amazon’s search bar, the suggestions that drop down are real searches real shoppers make, ordered roughly by popularity, terms near the top are searched more than terms near the bottom. That autocomplete feed is the raw material almost every KDP keyword tool repackages. It is free, public, and available to you directly, no subscription required.
Type your genre or topic and watch what Amazon suggests: “cozy mystery”, then “cozy mystery with cats”, “cozy mystery series”, “cozy mystery box set”. Each suggestion is a validated reader search and a candidate for one of your seven slots. Doing this by hand for free gives you the same source a paid tool charges to wrap in a dashboard.
What about the open-source tools specifically?
It is worth addressing the “open-source” part of the search directly. A handful of developers have built and open-sourced free KDP keyword tools, and they are genuinely handy: you can run them at no cost and even inspect or modify the code. But look under the hood and every one is doing the same thing, calling Amazon’s autocomplete endpoint and presenting the results. That is good news, not bad, because it means the value lives in the free data source, not in some proprietary algorithm you are missing by not paying. So if you find a maintained open-source tool you like, use it for the convenience. If you do not, you have lost nothing, since the engine is the same autocomplete you can query yourself in a browser. Open-source or paid, the moat these tools imply does not really exist.
KDP keyword tools compared
It helps to see where each option gets its data and what it costs.
| Tool | Type | Cost | Core data source |
|---|---|---|---|
| Amazon autocomplete | Free, manual | Free | Real Amazon searches |
| Open-source and free apps | Community built | Free | Amazon autocomplete |
| Publisher Rocket | Paid, one-time | $199 | Amazon plus estimates |
| MerchantWords | Paid, subscription | $35 to $149 a month | Amazon clickstream |
| SQSEO | Free core, paid Pro | Free for core | Google, AI search, long tail |
The table makes the point: the top two rows and the bottom one are free, and the free options draw from the same Amazon data the paid tools do, plus, in SQSEO’s case, the Google and AI demand the others ignore. The paid tools add convenience and volume estimates, not a secret source.
Do you really need MerchantWords or Publisher Rocket?
The honest answer is no, not to start. Publisher Rocket is a one-time $199 and MerchantWords runs $35 to $149 a month, and what they mainly add is estimated search volume and competition scores on top of the autocomplete data you can read for free. Those estimates are convenient, but the autocomplete order already gives you the one signal that matters most: relative popularity. A term Amazon suggests first is searched more than one it suggests last, and that ranking is free.
For a first book, or a tight budget, the free method covers the job. The paid tools earn their keep later, when you are publishing at volume and the time saved outweighs the cost, not when you are filling your first seven slots and a precise volume number would not change your choices anyway.
How to do KDP keyword research free, step by step
The free workflow takes about twenty minutes per book. First, brainstorm seed terms: your genre, sub-genre, themes, tropes, and the problems your book solves for a reader. Second, type each into Amazon autocomplete and collect every suggestion, noting the order. Third, widen the net with the “Customers also bought” and category pages of similar books, which reveal the language readers and Amazon use. Fourth, check the same terms in Google autocomplete, since readers research books there too. Fifth, pick the seven phrases that balance clear relevance with real demand, favoring specific multi-word terms over broad ones. The same free, marketplace-native approach works for any storefront; our guide to a free marketplace keyword finder for sellers walks through the identical two-layer method for product listings.
The error to avoid is grabbing the seven biggest terms. The broadest keywords are where you compete with every bestseller; the specific ones are where a new book actually surfaces.
The long tail wins for books too
Specific beats broad in publishing for the same reason it does everywhere: that is where the winnable demand is. Backlinko’s analysis of 306 million keywords found 91.8% of searches are long-tail with a median of 10 searches a month, and Ahrefs’ long-tail study put 92.42% of keywords at ten or fewer searches a month. For a book, “mystery” is unwinnable, but “cozy mystery with cats small town” is a precise reader intent you can own. A widely cited report on product search found around half of shoppers start on Amazon, and book buyers are among the most search-driven of all, so matching their exact long-tail phrasing is how a new title gets found.
A worked example
Say you wrote a cozy mystery set in a bakery. You open Amazon autocomplete and type “cozy mystery”, collecting “cozy mystery with recipes”, “cozy mystery bakery”, “cozy mystery series for women”, “culinary cozy mystery”. You check similar books’ categories and see “Culinary Cozy Mysteries” as a category to target. You confirm a couple of terms in Google autocomplete. From that, your seven keywords write themselves: specific phrases like “culinary cozy mystery”, “bakery cozy mystery with recipes”, and “small town cozy mystery series”, each a real reader search, none of them the impossible head term “mystery”. Twenty minutes, zero dollars, no MerchantWords.
Beyond Amazon: Google and AI demand
Amazon autocomplete covers on-platform discovery, but it is not the whole picture. Readers also find books through Google, where your book’s landing page, your author site, and reader questions all surface, and increasingly through AI assistants. Capturing that demand means researching how readers phrase things off Amazon too, including the questions they ask before buying. Our guide to finding bottom-of-funnel reader questions covers how to mine those decision-stage queries, which feed your book description, your ads, and any companion content that drives readers to the book.
Common mistakes
The first mistake is paying for a tool before trying the free Amazon autocomplete method that underlies it. The second is choosing the seven broadest keywords, where every bestseller competes, instead of specific winnable ones. The third is ignoring autocomplete order, which is a free popularity signal. The fourth is researching only on Amazon and missing the Google and AI demand around your topic. And the fifth is treating keywords as set-and-forget, when re-checking autocomplete after launch reveals new terms to test. Avoid these and the free method matches what the paid tools deliver for a first book.
The AI-search angle
Readers now ask AI for recommendations too. “What is a good cozy mystery with recipes” goes straight into ChatGPT or Perplexity, and the answer names a handful of books. That is a discovery channel no KDP-specific tool tracks, and it rewards books whose descriptions and surrounding content clearly match those conversational queries. Researching the AI-search phrasing around your genre, not just Amazon’s autocomplete, positions your book to be named when a reader asks an assistant what to read next, which is demand with almost no competition today. For a self-published author without an ad budget, that kind of free visibility is worth more than any volume metric a paid tool prints.
Do it free with SQSEO
For the half of the job Amazon autocomplete does not cover, SQSEO is the free tool that fits: enter your genre or topic and it returns hundreds of intent-grouped long-tail and AI-search queries, free, with no signup wall for core research. Use Amazon autocomplete for your seven on-platform keywords, and SQSEO for the Google and AI-search demand that feeds your book description, your author site, and any content marketing around the book. It does not replace Amazon’s on-platform data, it covers the off-Amazon discovery the KDP-specific tools miss. Core research is free forever, and the Pro tier adds trend history, competitor monitoring, AI-visibility tracking, team workspaces, and an API.
Frequently asked questions
What is the best free open-source KDP keyword research tool?
The best free option is the source every tool is built on: Amazon’s own search autocomplete, which shows real reader searches ordered by popularity, no signup or subscription needed. Community-built free and open-source apps simply wrap that same autocomplete feed. For the demand beyond Amazon, on Google and in AI search, we recommend SQSEO, which returns hundreds of intent-grouped long-tail and AI-search queries from one seed for free. Together they cover both on-platform and off-platform book discovery without paying for MerchantWords or Publisher Rocket.
Do I need to pay for Publisher Rocket or MerchantWords?
Not to start. Publisher Rocket is a one-time $199 and MerchantWords runs $35 to $149 a month, and they mainly add estimated volume and competition scores on top of Amazon autocomplete data you can read for free. The autocomplete order already gives you relative popularity, which is the signal that matters most for choosing your seven keywords. The paid tools earn their value when you publish at volume, not for a first book.
How do I find KDP keywords for free?
Use Amazon’s search autocomplete: type your genre, sub-genre, and themes, and collect the suggestions, which are real reader searches ordered by popularity. Widen the net with the categories and “also bought” sections of similar books, and check Google autocomplete too. Then pick seven specific, relevant phrases over broad ones. A free tool like SQSEO adds the Google and AI-search demand around your topic.
How many keywords does KDP allow?
Amazon KDP gives you seven keyword slots per book, in addition to your title, subtitle, and categories. Those seven phrases help determine which Amazon searches your book can appear in, so each slot should hold a specific, relevant phrase that real readers search, rather than a broad term where you compete with every bestseller.
Can I find KDP keywords in Google and AI search?
Yes, and you should, because readers discover books beyond Amazon. Google autocomplete shows how readers phrase searches there, and AI assistants like ChatGPT increasingly answer “what is a good book about X” by naming titles. Researching that off-Amazon and AI-search phrasing, with a free tool like SQSEO, feeds your book description and marketing and reaches readers the KDP-only tools never see.