KDP Keyword Research: The 7-Slot Playbook (2026)

Most KDP authors treat their 7 backend keyword slots like a sock drawer. They cram in a dozen single words, hit save, and wonder why the book sits at BSR 800,000. This guide is the opposite of that. It is a tactical, hands-on system for researching, scoring, and deploying the keywords that actually move books off the page. By the end you will have a repeatable workflow, a scoring rubric, and a worksheet you can copy for every book you publish.

Why Most KDP Keyword Research Is Wrong

The standard advice is: brainstorm words about your topic, paste them into your 7 slots, repeat. That approach treats Amazon like Google in 2008. Amazon is a buyer-intent search engine where every query is paired with a credit card. The algorithm rewards phrases that match how real shoppers describe what they are about to buy, not loose tags that describe what the book is about.

Three failures show up in almost every audit I run on a new author account:

• Single-word stuffing. Slots filled with "coloring", "mandala", "relax", "stress", "adult". Each word is indexed, but none of those slots match a real phrase a shopper types.
• Title duplication. The same word appears in the title, subtitle, A+ content, and three of the seven backend slots. Amazon counts it once.
• No competitive screening. The keyword set sounds reasonable in a spreadsheet, but every phrase has 50,000+ competing results and the top books were published in 2019. Unwinnable.

The Two-Layer Keyword Model: Frontend vs Backend

Before you touch a slot, internalize this: KDP keywords live in two layers, and a winning book needs both.

Layer 1: Frontend keywords (visible)

These appear in the title, subtitle, series name, author bio, A+ content, and product description. Amazon weights them heavily because shoppers can see them. They double as conversion copy. Frontend slots are where you put the highest-volume buyer-intent phrases you can naturally fit. Example for a coloring book:

• Title: Mandala Coloring Book for Adults
• Subtitle: 50 Stress-Relieving Designs for Relaxation and Mindfulness

Between the two fields you have already captured "mandala coloring book", "coloring book for adults", "stress relieving designs", "relaxation coloring book", and "mindfulness coloring book". Notice what is missing: every word in those phrases. That is the point. The frontend is where you put the volume.

Layer 2: Backend keywords (hidden)

The 7 hidden slots are where you capture everything the frontend cannot reach without making the title unreadable. Audience descriptors, format details, secondary use cases, gift framings, niche misspellings, seasonal terms. The frontend says "what the book is", the backend says "all the other ways a shopper might search for it".

The 5-Stage Keyword Research Workflow

Here is the workflow I use for every new book. It takes 60 to 90 minutes the first time, and 20 minutes once you have a working spreadsheet template.

Stage 1: Seed list from your niche brief

Open a blank spreadsheet. Column A: "Phrase". Brain-dump every phrase a shopper might type to find your book. Aim for 40 to 60 seeds. Do not edit. Bad ideas are fine at this stage because the next four stages filter ruthlessly. Sources for the brain-dump:

• How you would describe the book in one sentence to a friend
• The pain or desire the book solves
• Who the book is for (age, role, life stage, skill level)
• When and where they would use it (gift, travel, bedtime, classroom)
• Adjacent niches (a mandala book overlaps with zentangle, geometric, meditative art)

Stage 2: Amazon autocomplete mining

Autocomplete is the single most underused free tool in KDP. It surfaces what Amazon shoppers actually type, ranked by frequency. The workflow:

1. Set the dropdown to the Books category (Books, Kindle Store, or All Departments depending on your goal).
2. Type your seed phrase and record every suggestion that appears.
3. Add a single letter to the end: "coloring book a", "coloring book b", through Z. Record everything.
4. Prepend modifiers: "for adults", "for women", "easy", "advanced", "large print". Record everything.
5. Repeat with the top 5 phrases from your seed list.

You will end Stage 2 with 100 to 200 phrases. Some will be duplicates. That is fine. The unique gold here is the modifiers shoppers add that you would never invent: "coloring book for women relaxing", "coloring book for seniors large print", "coloring book by black women". Every one of those is a real query someone is typing right now.

Stage 3: Competitor harvesting (reverse-ASIN lite)

Open the Amazon search results for your top 3 seed phrases. For each phrase, open the top 5 best-selling books that match your format and target reader. For each book, harvest:

• Title and subtitle (every meaningful word)
• Series name if present
• Category breadcrumbs
• Top 3 phrases that appear repeatedly in the product description and A+ content
• Top 3 phrases that appear in customer reviews

Paste them all into your spreadsheet. For deeper extraction, jump to our reverse-ASIN lookup guide and our reverse-ASIN keyword harvesting playbook. The second piece is the one to read if you want a step-by-step process for turning competitor ASINs directly into backend slots.

Stage 4: Score every phrase

You now have a sheet with 150 to 300 candidate phrases. Most of them are wrong. The job of Stage 4 is to score them and surface the top 20 to 30 worth fighting for. Add four columns to your sheet:

• Relevance (1 to 5): Does this phrase describe what a shopper would buy from your book? A 5 is a phrase your ideal reader would type without prompting.
• Competition (1 to 5, higher is better): Lower competing-results count and weaker top-rank books score higher. Under 1,000 results scores 5. Over 50,000 scores 1.
• Intent (1 to 5): Does the phrase signal an active buyer (5) or a casual browser (1)? "Best coloring books" is a 2. "Coloring book for adults stress relief large print" is a 5.
• Velocity (1 to 5): Are the top results selling? If the BSR of the top 5 books is under 100,000 in their main category, score 5. If everything ranks above BSR 500,000, score 1.

Sum the four scores. Any phrase scoring 16 or higher is a candidate for your title, subtitle, or one of the 7 backend slots. Anything below 12 gets cut. Anything between is parked for AMS targeting.

Stage 5: Slot assignment

Take your top 7 highest-scoring phrases that are not already in your title or subtitle. Assign one phrase per slot. Each slot is a single coherent phrase, not a tag cloud. If a phrase fits in 30 characters, fill the remaining 20 with a tightly related phrase that shares no words with anything else in your listing. Example for an adult coloring book:

Long-Tail vs Short-Tail: The Real Trade-Off

Almost every guide tells you to focus on long-tail keywords. Almost every guide is half-right. The trade-off is not "long-tail good, short-tail bad", it is volume against winnability.

• Short-tail ("coloring book"): Tens of thousands of searches per day, but millions of competing results and entrenched bestsellers. You will not rank in your first year. You can pay for it through ads.
• Mid-tail ("mandala coloring book"): Thousands of searches, hundreds of thousands of results. Rankable in 6 to 12 months with consistent reviews and ad support.
• Long-tail ("mandala coloring book for adults stress relief"): Hundreds of searches, but those searches are people who are literally typing your book. Rankable in 4 to 12 weeks. Higher conversion. Higher review rate per click.

For a new book: 1 short-tail in the title, 2 mid-tail in the subtitle and series, and 7 long-tail in the backend slots. As reviews accumulate and BSR improves, you can swap one of the long-tail backend slots for a more aggressive mid-tail and start pushing for a higher-volume phrase.

Competitive Density: How to Avoid Unwinnable Niches

Competitive density is the ratio of demand to supply for a given keyword. A high-density phrase has many shoppers chasing few books (good). A low-density phrase has few shoppers chasing many books (bad). For each phrase you are scoring, run this 3-minute check:

1. Search the phrase on Amazon Books. Note the number of results.
2. Open the top 5 results. Note their BSR (paperback or Kindle).
3. If any of the top 5 has a BSR worse than 200,000, the phrase is winnable. The bestsellers are not selling enough to defend the position.
4. If all 5 have BSR under 50,000, you are walking into a fortified market. Move on or attack from a long-tail flank.

The Amazon Autocomplete Mining Workflow (Detailed)

Autocomplete deserves its own section because most authors do it for 5 minutes and quit. Done thoroughly, it replaces an entire paid tool. The complete workflow:

1. Set the right dropdown. For paperback and hardcover books, set the search dropdown to "Books". For Kindle ebooks, set it to "Kindle Store". The suggestions differ.
2. Run the alphabet sweep. Type your seed phrase plus a space, then cycle through a, b, c... z. Record every suggestion that fits your niche.
3. Run the modifier sweep. Before your seed: "best", "easy", "advanced", "large print", "spiral bound", "for women", "for kids", "for seniors". After your seed: "with answers", "book 1", "volume 2", "2026 edition".
4. Run the season sweep. "christmas coloring book", "summer puzzle book", "back to school workbook". Seasonal modifiers reveal annual demand spikes.
5. Run the cross-niche sweep. If you are publishing a coloring book, also search "activity book", "puzzle book", "journal", "workbook". Shoppers cross over.

Plan on 30 to 45 minutes for a complete sweep. You will end with a phrase list that no paid tool would have given you in full, because autocomplete is closer to Amazon than any third-party scrape.

Backend Slot Anti-Patterns (Avoid These)

Even with a strong phrase list, the wrong slot deployment kills the result. Avoid the following:

• Commas, semicolons, and quotation marks. Wasted characters. Spaces are the only separator Amazon uses.
• Repeating words across slots. Each indexed word counts once. If "coloring" appears in 3 slots, you have wasted 2 of them.
• Repeating words from title or subtitle. Same rule. Slot space is precious.
• Subjective claims. "Best", "amazing", "#1", "must-read". Amazon strips them in ranking and they violate the keyword terms of service.
• Competitor names, brand names, trademarks. Fast route to suppression. Never put another author's name in your slots.
• Time-bound claims. "New 2024", "now available", "just released". Amazon prohibits them and they age out fast anyway.
• ASINs or product codes. Strict TOS violation. Suspension risk.
• Foreign language overlap. If you publish only in English, do not stuff Spanish or German keywords. Use the foreign-language KDP store instead.

Categories: The Other Half of Discoverability

Keywords answer "what is this book about?". Categories answer "what shelf does this book belong on?". You can have perfect keywords and still get buried if Amazon shelves you wrong. Two rules:

• Match your format and price band first. If you publish an 8.5x11 paperback workbook for $9.99 and Amazon puts you in a hardcover-dominated category at $24.99 average price, your conversion will collapse and your BSR will drag your keywords down with it.
• Pick the deepest specific category, not the broadest. "Books > Crafts, Hobbies & Home > Crafts & Hobbies > Coloring Books for Grown-Ups > Mandalas" outperforms "Books > Coloring Books for Grown-Ups" every time. Specific categories rank you against fewer books and surface you to higher-intent shoppers.

For deeper category strategy, see our KDP niche research system, which pairs naturally with this guide.

Tool Comparison: Which Keyword Tools Are Worth Paying For?

I have run hundreds of dollars through every major KDP keyword tool. Here is the honest comparison, focused on what actually returns its cost.

Publisher Rocket (one-time $97)

The legacy standard. Strong on backend keyword suggestions and category browse-node lookup. Weak on velocity data and ranking insight. Worth it once if you publish 5+ books per year and want a deep category-mining workflow. The keyword suggestions are conservative compared to autocomplete plus competitor extraction.

Helium 10 (subscription, $99 per month and up)

Built for general Amazon FBA, with a Books extension. Strong on search volume estimates and trend data. Overkill for most KDP authors unless you also sell physical products. The free Chrome extension (X-Ray) is worth installing even if you never pay for the full tool.

KDPEasy keyword research (built in, free trial)

Our own tool, so calibrate the bias. Built specifically for KDP, scored on the same 5-stage workflow this article describes, and outputs ready-to-paste backend slots. The 2026 update adds AMS-keyword export so the same research feeds both organic and paid. Start with the free trial before you commit to a paid subscription elsewhere.

From Backend to Amazon Ads: The Keyword Recycling Loop

The phrases that scored 13 to 15 in your worksheet are the unsung heroes of KDP. They are not strong enough to burn a backend slot on, but they are perfect for Amazon Ads. The recycling loop:

1. Move your top 7 phrases into backend slots.
2. Move phrases scoring 13 to 15 into a manual Sponsored Products campaign as exact and phrase match keywords.
3. Run the campaign for 14 days with a modest bid ($0.25 to $0.50).
4. Pull the search term report. Promote any keyword that drove a sale at acceptable ACoS into a more aggressive bid tier.
5. If a paid keyword consistently converts above your organic average, swap it into a backend slot, freeing up an underperformer.

This loop turns your keyword research into a compounding asset. For the full ads workflow, see our Amazon Ads for KDP strategy guide and KDP Amazon Ads keywords playbook.

The Keyword Scoring Worksheet (Copy This)

Here is the exact column structure I use in Google Sheets or Excel. Build it once, clone it for every book.

• Column A: Phrase
• Column B: Source (autocomplete, competitor, brain-dump, review mining, ads report)
• Column C: Character count (=LEN(A2))
• Column D: Competing results count (manual lookup, paste in)
• Column E: Top-rank BSR (paste BSR of #1 result)
• Column F: Relevance (1 to 5)
• Column G: Competition (1 to 5)
• Column H: Intent (1 to 5)
• Column I: Velocity (1 to 5)
• Column J: Total (=F2+G2+H2+I2)
• Column K: Decision (title, subtitle, slot 1-7, AMS only, cut)

Sort descending on Column J. The top 7 phrases that fit in 50 characters and are not already in your title or subtitle become your backend slots. The next 15 to 25 go into your AMS campaign launch list. Everything below row 30 gets archived.

Measuring Whether Your Keywords Are Working

Keywords are not "set and forget", but they are also not "tweak weekly". Run the following review cadence:

• Days 1 to 30: Do nothing. Let Amazon index and let early reviews accumulate. Track BSR daily for the trend, not the level.
• Day 30 check-in: Search each of your 7 backend phrases plus your title phrases on Amazon. Note what page you appear on. If you are buried past page 5 for everything, the issue is more likely your cover or price band, not your keywords.
• Day 60 review: Pull the Amazon Ads search term report. Move any high-converting search terms into your backend slots, retiring underperformers.
• Day 90 refresh: Update for seasonality. Swap one or two slots for upcoming Q4, Mothers Day, back to school, planner season, etc.
• Quarterly: Re-run the full 5-stage workflow on your top 3 selling books. Markets shift, language shifts, what worked in Q1 may not work in Q4.

Putting It Together: The 90-Minute First Run

If you have never done structured keyword research before, here is what your first session looks like end to end. Block 90 minutes.

1. 0 to 10 min: Write your one-sentence book brief and brain-dump 40 to 60 seed phrases.
2. 10 to 40 min: Run Amazon autocomplete mining (alphabet, modifier, season sweeps).
3. 40 to 60 min: Open the top 5 books for your 3 strongest seed phrases. Harvest titles, subtitles, descriptions, reviews.
4. 60 to 80 min: Score every phrase across relevance, competition, intent, velocity. Sum and sort.
5. 80 to 90 min: Assign top 7 phrases to backend slots. Park the next 20 for an AMS launch campaign.

After the first run, you have a template. Every subsequent book starts with a known good seed set and finishes in 20 minutes. That is how authors who publish 30+ books a year actually scale.

Common Mistakes That Tank Otherwise Solid Books

• Researching keywords after writing the book. Keyword research is market validation. Doing it after the manuscript is finished means you are forcing a square book into a round niche.
• Optimizing for search volume only. A high-volume phrase you cannot rank for sends zero traffic. A low-volume phrase you own sends every shopper who types it.
• Copying competitor backend slots wholesale. You cannot actually see their slots (Amazon hides them). Anyone selling you a "spy on backend slots" tool is selling you their best guess at the title + description. Use that data as a hint, not a copy target.
• Ignoring the title. The single most important keyword placement is the title. A perfect backend slot loadout cannot rescue a title that mentions zero search phrases.
• Treating Kindle and paperback as the same. The two formats have different category trees, different bestseller patterns, and slightly different autocomplete suggestions. Score them separately when it matters.
• Forgetting category alignment. Discussed above and worth repeating. Keywords without aligned categories are darts thrown into the wrong room.

Your Next 7 Days

If you take one thing away from this guide, take this: keyword research is a system, not a search. Build the worksheet once, run the 5 stages every time, and your hit rate compounds book by book. Here is a concrete 7-day plan:

1. Day 1: Set up your scoring spreadsheet using the column structure above.
2. Day 2: Pick one book in your catalog (or one planned book). Brain-dump seeds and run autocomplete mining.
3. Day 3: Harvest 5 competitor listings end to end.
4. Day 4: Score every phrase. Sort and pick your top 7.
5. Day 5: Update the title, subtitle, and backend slots inside KDP. Save and walk away.
6. Day 6: Launch a low-budget AMS campaign on your next 20 phrases. Set bids at $0.25.
7. Day 7: Document the workflow as your repeatable template. Every future book runs the same loop.