Keywords & Niches

The 7 KDP Keywords Strategy That Got My Book to Page 1

Amazon gives you seven backend keyword boxes. Most authors stuff them wrong and stay invisible. Here is the 3-2-2 split I use to rank, plus how I find the keywords.

By Nezir Basar · 2026-05-23 · 7 min read

You have seven boxes. Most authors waste all of them.

When you upload a book to KDP, Amazon hands you seven backend keyword fields. This is where you tell the algorithm what your book is about. And this is exactly where beginners make the mistake that keeps their book buried where no one will ever find it.

Your KDP keywords are not a place to dump every word you can think of. They are a place to make a decision. Focus or spread. Strong rankings on a few terms, or weak rankings on many.

I'm going to show you the exact way I fill out those seven boxes, how I find the keywords in the first place, and the four mistakes that can actually get you in trouble with Amazon.

How Amazon reads your seven keyword boxes

Here is the part nobody explains clearly.

Each box gives you up to 50 characters, letters and spaces included. So the obvious question is: one keyword per box, or pack in as many words as you can fit?

The answer is: it depends. Let me explain why.

Amazon takes every word inside a single box and automatically builds combinations from them. Write mom memory book gift journal in one box and you start ranking for "mom memory book," "mom gift," "memory journal," "mom memory gift," and more.

Sounds great. More words, more combinations, more traffic. But there's a catch.

The more words you cram into one box, the weaker your ranking power becomes for each individual keyword.

Think of it this way. If a box only says mom memory book, you're telling Amazon: my book is exactly about this. All your ranking power points at one target.

Pack eight words in there instead, and you dilute the message. That same power now spreads thin across every combination. You rank for more terms and rank well for none of them.

The 3-2-2 strategy for your KDP keywords

So the fix isn't "more" or "less." It's balance.

This is the split I use, and it's the one most successful KDP publishers land on too. Three jobs, seven boxes.

Boxes 1 to 3: your main keywords

These are the terms you genuinely want to rank for. The ones that matter most.

Here you write one targeted keyword or short phrase per box. Nothing extra. No filler. This is where your ranking power stays focused and strong.

Boxes 4 and 5: category keywords

Amazon sometimes shuffles your book into categories based on your keywords. That can be annoying, but you can steer it.

Use these two boxes to nudge Amazon toward the categories you actually want to sit in. For a mom memory book, that might be parenting, relationships, or self-help journals.

Boxes 6 and 7: filler keywords

This is where you pack everything that's left. As many relevant words as fit.

These are the nice-to-have terms. You'd like to rank for them, but they aren't critical. Cram them in and you pick up extra reach without weakening your three main targets.

The result: strong rankings where it counts, and bonus visibility everywhere else.

How I find keywords manually on Amazon

You can't fill those boxes until you know what real readers search for. And the best place to find that out is Amazon itself.

I open Amazon.com in incognito mode. Why incognito? Because I want to see what real customers see, not a feed shaped by my own browsing history.

I type my main term into the search bar. Something basic, like mom memory book. Amazon instantly suggests related searches. These aren't random. They're the terms real customers type most.

For that one seed, I saw suggestions like "mom memory book to fill out," "memory book for son," and even a Spanish version that's surprisingly popular. Good leads, all of them.

Read the competition, not just the suggestions

Here's the trick most people skip. Don't stop at suggestions. Check the actual search results.

Look at the top left of the results page. It tells you how many results exist. Fewer results means less competition. My rule of thumb:

A term showing 20,000 results is a hard fight. Not worth your strongest box.

But the result count alone isn't enough. Open the first few books and read the room.

If the top books all have thousands of reviews, that keyword is locked up by bestsellers. One book with 7,000 reviews is a wall you won't climb soon. You want competitors sitting around 100 to 500 reviews. Beatable, but proven.

And watch the other extreme too. If every book for a keyword has fewer than 10 reviews, that term probably isn't selling at all. Dead traffic.

When I searched guided journal from mom to son, I found about 3,000 results and books with review counts like 130, 300, 50, and 119. Reachable competitors with real demand. That one I saved.

One more source people ignore. Click into a book and scroll to the "related books" section.

These are the topics your audience also cares about. I found things like "grandma's story" and "memory and keepsake" this way, terms I hadn't thought of.

It also plants a seed for later. Publish the mom version, then a grandma version, a dad version, a daughter version. Same research, more books.

Done properly, this manual hunt takes me one to two hours. I list everything in a spreadsheet and rank it by competition and review counts.

Doing the same research in minutes

That hour or two of manual work is real. So this is where I lean on a tool to compress it.

Inside Publbee, I drop the same seed term, mom memory book, into the keyword research tool and search. One pass returned 231 related keywords. By hand, that's hours of digging.

For any keyword that interests me, I can load its data: search volume, average sales per day, number of competitors, average reviews, and average price.

It also gives two scores. An opportunity score from 0 to 100, and a difficulty level. The combination matters more than either alone. An opportunity score of 100 with a "hard" difficulty still means you're fighting uphill against entrenched competitors.

I read those numbers the same way I read Amazon's results page. Low difficulty, real demand, beatable reviews. The tool just gets me there faster.

Filling the seven boxes for real

Once I've collected my keywords, I slot them into the 3-2-2 structure.

Main keywords (boxes 1-3), one targeted phrase each:

Category keywords (boxes 4-5), to steer Amazon's categories:

Filler keywords (boxes 6-7), packed with good leftovers:

Notice the filler boxes don't have to read like a sentence. The words just need to be relevant to your book.

Four mistakes that can get you in trouble

Before you publish, a few hard rules. Some of these aren't just bad strategy. They can actually get you flagged.

  1. Don't repeat your title or subtitle. If "mom memory book" is already in your title, you rank for it anyway. Putting it in the keywords too is a wasted box.
  2. No words like "free," "bestseller," "new," or "sale." These violate Amazon's terms of service and can land you in real trouble.
  3. No brand names or other authors' names. You can't borrow a well-known author's name as a keyword. Amazon treats that as metadata manipulation.
  4. No random words just to fill space. Every keyword should genuinely fit your book. Amazon notices when you're trying to game the system.

The whole strategy in three lines

The seven boxes aren't a dumping ground. They're a decision about where your ranking power goes.

Use boxes 1-3 for your most important keywords, one per box. Use 4-5 to steer your categories. Use 6-7 for relevant filler. Find the terms on Amazon, validate them by competition and reviews, and never break the four rules above.

Want to skip the one-to-two-hour manual hunt? Start with the KDP Keyword Research tool and pull your shortlist in minutes.