Marketing

Why Your Amazon Ads Keywords Are Wrong (And How to Fix It)

The keywords with the highest search volume are usually the worst performers. Here is the exact four-criteria check I use to find Amazon ads keywords that convert.

By Nezir Basar · 2026-05-16 · 9 min read

Most authors target the wrong keywords.

You go too broad, or you chase the biggest search volume, and you watch hundreds of dollars vanish on clicks that never turn into sales. I have done it. It is the difference between a profitable campaign and a slow leak in your budget.

Here is what nobody tells you. The Amazon ads keywords with the highest search volume are usually the worst performers. The competition is brutal, the cost per click is high, and your ad gets buried.

The keywords that actually make you money are hiding in plain sight: lower search volume, lower competition, higher conversion. I am going to show you exactly how to find them.

Why high-volume keywords drain your ad budget

Most people think keyword research is about quantity. Get a list of 100 keywords, dump them into a campaign, and see what sticks. That is how you go broke.

Almost 90% of those keywords will bleed your budget dry. Some have zero demand. Others are dominated by books with 5,000 or more reviews, run by authors with marketing budgets you cannot touch.

You cannot out-spend a book with a million-dollar budget. So stop trying to fight on their keywords.

The better approach is validation before spending. We only run keywords that pass four specific tests: relevance, demand, competition, and search volume. Fail one, and it is out.

I did not pull this framework out of thin air. We tried it across hundreds of book launches and burned through thousands in ad spend figuring out what works.

For this walkthrough, I am using a real book: Dear Son, This Is My Life Story, part of a guided journal series. We need keywords for a manual campaign. Not auto-targeting. Specific, validated terms we know will convert.

Set up Amazon for accurate keyword data first

Before you find a single keyword, you have to set up correctly. Most people skip this, so their data is wrong from the start.

Open an incognito window

Amazon personalizes search results based on your history. Search mystery thrillers all week, and Amazon assumes you want more of them. But your real customers see something completely different.

So open a private window. On Chrome, that is Ctrl+Shift+N on Windows or Cmd+Shift+N on Mac. Firefox uses the same shortcuts. Now you see what an actual customer sees, with no history and no personalization.

Set your location to the US market

If you sell on Amazon.com but browse from Austria, Germany, or the UK, Amazon limits what you see. Availability changes, ranking shifts, and your data goes sideways.

Fix it. Go to amazon.com, click Deliver to in the top left, and enter a US zip code. I use 10001, which is New York City. It is a major, neutral market and it unlocks the full catalog.

Install the Publbee Chrome extension

This puts the data you need directly inside Amazon's search results. It is free. Search for Publbee in the Chrome Web Store.

Here is what you get. At the top, the average book data for your keyword, so you know instantly whether people search this term at all. Then, for every book in the results, estimated daily sales and daily royalties.

No more clicking into each book and reverse-engineering numbers from reviews and rank. That alone saves you hours. The free plan caps historical data like BSR or price history at five lookups a month, but for keyword research you do not need it.

Harvest keywords before you validate anything

Setup done. Now we build a big, messy list. We trim it later.

Use the alphabet method

Start with a seed keyword. For our book, seeds could be mom journal, mother son gift, life story book, or memory journal.

Pick one. Select the Books category first, this is crucial. Type mom journal into the search bar, but do not hit enter. Amazon's autocomplete fires: mom story journal, "mom tell me your story" journal, and more.

Now add a letter. Type mom journal a and new suggestions appear, like about her life. Then mom journal b, and so on through Z. Every suggestion is a real search a real customer typed. All of them go on your list.

Dig through the "also bought" carousels

Most people hit the bestseller list and stop. That is a mistake. The best keywords hide deeper.

Go to Amazon Bestsellers, then into your category. For us that might be Parenting and Relationships, then deeper still. Click a book. Scroll to the "Customers who bought this item also bought" carousel.

Those titles are your keywords: mom's story memory and keepsake journal, dad's story, "grandma tell me your story". Amazon is literally showing you what customers buy together.

Click one of those books and repeat. Do it five or six times. Patterns appear: letters to my son, keepsake book, guided journal for moms. Book titles are often the exact phrases customers type.

Brain-dump synonyms

Customers do not search like academics. They search from emotion and from problems.

So brainstorm. For life story, think my story, memoirs, biography. For journal, think book, diary, keepsake. Then combine: mom life story book, mother son memory book. Keep going to 20 or 30 variations. AI is handy here for quick lists.

When you are done, your sheet should hold 70 to 100 keywords. That is exactly what you want.

Validate every keyword against four criteria

Now the part that decides everything. Not all of those keywords are worth a cent. From 100, only 10 to 20 survive. This is where most people fail.

Criterion one: relevance

Type the keyword, hit enter, and look at the first 10 books on page one. Ask one question: does my book belong here?

Try mom journal aesthetics. Page one shows art journals, bullet journals, aesthetic planners. Does a life-story journal fit there? No. Mark it as a fail and move on.

Now try life story book for son. Page one shows guided journals for mothers, memory books, keepsakes. That fits.

The rule is simple. If 7 out of 10 books on page one point the same direction as your book, it is relevant. If not, discard it.

Criterion two: demand

Now check if people are actually buying. We hunt for what I call winning books.

A winning book has two traits: consistent sales in the Publbee extension, and fewer than 150 reviews. Why under 150? Because a book selling well without a mountain of reviews is selling on the keyword and topic, not on an old, established brand. That is the gap you can compete in.

Type a keyword and read page one:

Count the winners. Three or more on page one means good demand. One or two is medium, which can still work. Zero winners means no demand. Discard it.

You know that feeling, sitting alone at your laptop, wondering if you are even doing this right? That is why we built Publbee as a command center: keyword research, Niche Scout, sales insight, cover generator, and more. No more juggling ten subscriptions.

Criterion three: competition

This has two warning signs: dead books and authority figures.

Dead books carry lots of reviews but show little to no sales. Think 30-plus reviews and barely any movement. Customers clicked but did not buy. High clicks, low conversion. A money pit.

Count them per page. Zero to two is fine. Three to five, be careful. Six or more, discard the keyword.

Authority figures are worse. These are top-five books with over 1,000 reviews. When three or more of the top five clear that bar, they own the search. Customers click almost only on them, your ad gets buried, and your cost per click climbs.

Look at memoirs and you see books with 220,000 reviews, 87,000, 6,000, 3,000. That is a bloodbath. Unless you have a serious budget, avoid keywords like these.

Criterion four: search volume

Type the keyword and look at the total results in the top left of the page. That number matters more than you think:

A keyword sitting at 50,000 results is too much. I would refine it before I would ever bid on it.

A keyword has to pass all four. Relevant, real demand, low competition, reasonable volume. Fail one, it is gone. On this book, I started with 100 keywords and finished with 10 solid ones. That run took me about an hour, by hand, for 10 keywords.

Automate the same checks and reverse-engineer competitors

The manual method works, and you should learn it. But doing this for 50 or 100 keywords is hours of work. So I run the same four checks automatically.

In the Publbee keyword research tool, you enter a seed like mom life story, hit search, and wait 10 to 20 seconds. It pulls the top 10 books and calculates everything you just did by hand, plus an opportunity score and a difficulty level.

Those last two do the heavy lifting. The opportunity score blends volume, competition, and profit into a 0 to 100 number. Over 80 is high, under 40 is low, and the orange middle needs a closer look. For ads, I want both green, and I avoid anything under 40 to 50.

It even shows format distribution. A keyword dominated by paperbacks may not convert if you sell Kindle only.

Then there is the move that changes the whole game: reverse keyword research. You find a competitor who already ranks and sells, then pull every keyword they rank for. Grab their ASIN straight from the extension, drop it into the Reverse ASIN Lookup, and search.

On one mother son guided journal competitor, we found 314 keywords that book ranks for. The best part? You already know they convert, because that author is selling on them right now.

Find keywords that convert, then go build the campaign

Here is the whole thing in three lines.

High-volume keywords usually lose money. Validate every keyword against relevance, demand, competition, and search volume before you spend a cent. The best keywords come from competitors who are already winning.

Learn the manual method first so you understand why a keyword works. Then let the tools give you back your hours. Build your validated list this week, and your first campaign finally has something real to stand on.