Why Is My Book Not Selling on Amazon? And Why Almost Every Answer You've Been Given Is Wrong

You have already typed "why is my book not selling on Amazon" into a search bar. That is probably how you got here.

And you already know what you found. Give the book away free for five days. Get more reviews. Run some ads. Post more on social media. Somewhere in the results, if you went deep enough, you found a forum thread telling you that the real reason to publish a book is the sense of accomplishment, that any sales beyond that are gravy, and that if you write literary fiction or serious nonfiction without a traditional publisher behind you, you should quietly adjust your expectations.

You closed the tab feeling worse than when you opened it.

I want to say something plainly, and it is the reason I am writing this at all. Almost every one of those answers is addressing the wrong problem. Not badly. Wrongly. They are all answers to the question how do I push my book harder, when the question that actually decides whether a book lives or dies is a different one entirely.

Here it is. Can the people already looking for your book find it?

Not readers in general. Not the market. The specific person who, this week, went looking for exactly what your book delivers, and came away with something else.

The thing nobody says out loud

There is a line buried in the technical documentation that publishers and platforms rarely put in front of authors, because it is uncomfortable and it does not sell anything. It goes like this: invisible books fail regardless of craft.

Regardless of craft.

Your book can be the best thing in its category. It can be reviewed by Kirkus. It can be the thing you gave four years of your life to. None of that produces a single sale if the book does not appear when the reader looks. Quality is necessary. It has never been sufficient. And every piece of advice about free days and ad spend is trying to compensate for a discovery problem by shouting louder, rather than fixing the reason nobody heard you in the first place.

Here is what actually happens when someone looks for a book.

They do not type your title. They have never heard your title. That is the whole problem. They type how they feel, or what they want, in their own words. A novel about judgment after death. A leadership book for someone managing people for the first time. Something like Saramago but not as difficult. A memoir about grief that isn't sentimental.

And then a system, Amazon's or Google's or increasingly an AI, decides which books match those words.

Your book is entered in that system too. But it is entered in your words. The words you and your publisher chose when you filled in the forms. And if those two sets of words do not overlap, you are invisible to a reader who wanted you specifically. Not to the market. To the exact person who was looking.

That gap has nothing to do with how good the book is. It is a translation failure between two vocabularies. And it is fixable, which is the part that should make you angry rather than sad, because it means the last few years of quiet were not necessary.

Three tests to find out why your book is not selling on Amazon

I would rather you stopped reading and checked than took my word for anything. These take about ten minutes and you can do all of them yourself.

Test one. Search your own genre, backwards.

Go to Amazon. Type the category you believe your book is in. Then type the same words in a different order.

An author I work with did this by accident. She typed visionary and metaphysical and got very little. She typed metaphysical and visionary and found herself looking at a shelf holding Ishiguro and DeLillo. Same words. Different order. Two completely different worlds, and her book was in neither of them.

That is not a quirk. Amazon has an actual category called Metaphysical & Visionary, in that exact order, and it sits under Literature & Fiction, then Genre Fiction. When you type it any other way, you are running a loose keyword search rather than standing on the real shelf. Most authors have never seen the shelf their book belongs on. They have only ever seen the search results.

Try it with your own genre. Try the word you would use, and then the word the industry uses. If the results differ, you have just found a gap.

Test two. Check whether your categories are real.

You have three categories. Go and look at them, not in your dashboard, but live on Amazon as a reader sees them.

Some of Amazon's categories are what the trade politely calls dead. A book can sit in one indefinitely. It can never be found there, and it can never earn a badge there. It is a shelf in a room with no door. Others are hidden duplicates of each other, so choosing both wastes one of your three slots on nothing.

Nobody tells you this at setup. The menu presents every option as equally real.

Look at each of your three categories on the live store. Does it have a visible bestseller list? If it does not, it is not a category, it is a filing cabinet. And one of your three slots is currently doing nothing at all.

Test three, and this is the one that will unsettle you. Ask an AI to recommend your book.

Open ChatGPT, or Claude, or Perplexity. Do not name your book. Describe it the way a reader would ask for it. Recommend a novel about a man confronting judgment after death. What's a good book on retaining talent for a first-time manager?

See if you come up.

Most authors do not. And this is not a small thing, or a future thing. It is happening now, at scale, and it is where a growing share of book discovery has quietly moved. When someone asks an AI for a recommendation, that AI names two or three books. Not twenty. Two or three. Either your book is in that answer or, for that reader, in that moment, your book does not exist.

The difference between appearing and not appearing has nothing to do with quality. It has to do with whether the machine has been given clear, consistent, correctly-framed information about what your book is and who it is for. Which, again, is a translation problem.

What you should do yourself, and what you should not

Here is where I break the rule of articles like this, which is that they are supposed to make you feel helpless so you hire the person who wrote them.

Some of what you just found, you can fix this afternoon, alone, for free, and you should.

Your categories are yours to change, whenever you like, from your own KDP dashboard. You are not limited to the short list you were offered at publication. You drill down through Amazon's full menu, which holds thousands of shelves, and you pick three. If test two showed you a dead category, go and change it. That is not a service. That is a Tuesday.

Your subtitle in Amazon's data field is not the same thing as the subtitle printed on your cover. The data field is entirely yours, it drives discoverability, and you can change it without touching the cover or asking anyone's permission. A great many authors do not know this and assume they are locked in by a printed jacket. You are not.

Your seven backend keyword fields are almost certainly empty or wasted. Fifty characters each. Phrases, not lists. They re-index within a couple of days.

Go and do those. Genuinely. If that is all you ever take from this article, it will have been worth writing.

What you should not do alone is the part where those changes have to be correct. Knowing that categories are editable is easy. Knowing which three, out of thousands, put you where your readers actually browse, avoiding the dead shelves and the duplicates, is not. Knowing your subtitle field is free is easy. Knowing what should go in it, in the language readers are actually typing, is the entire job.

That is the difference between knowing the levers exist and knowing which way to pull them. The first is public. The second is research.

The book was never the problem

In 2023 a man called Allen Levi self-published his first novel. He was in his sixties. No publisher, no machine, no platform. He sold it by hand, person to person, and it reached something in the region of a hundred and fifty thousand copies that way. Then Simon & Schuster bought it, republished it, and it went to number one on the New York Times list and stayed there for weeks.

Read that slowly, because there is one detail in it that matters more than the rest.

The book did not become good when the publisher arrived.

It was always good. Every copy it eventually sold, it deserved on the day it was written. What changed was that enough of the right readers found it, and then they carried it the rest of the way.

That is the whole thing. That is what every free-book strategy and ad campaign and review-buying rationalisation is fumbling toward without naming. The book is not the variable. The finding is.

So if your book is quiet, I would gently suggest that the least likely explanation, and the one you have probably been sitting with alone at 2am, is the one about the writing.

The far more likely explanation is that somewhere between how you described your book and how your reader described what they wanted, there is a gap. And it has been open for a while.

The question worth sitting with is not what that has cost so far.

It is what another year of the same silence costs.

---

InsightIgnitee studies why strong books stay invisible and identifies the specific gap between how a book presents itself and how its readers search for it. If you want to know what we would find in your book, we look before we talk.

If you want to know what we would find in your book, write to me at info@insightignitee.net. I will look first, and then I will tell you what I found, whether or not you ever work with us.

---