Most books published on Amazon do not sell. The numbers vary by source, but the rough estimate is that more than half of self-published titles sell fewer than a hundred copies in their lifetime. A small percentage do well, a smaller percentage do very well, and the rest fade.
The reasons books fail are usually not mysterious. They follow patterns. Authors who learn the patterns can diagnose what went wrong with one book and avoid the same problems with the next.
This guide walks through the most common reasons books fail on Amazon, in rough order of frequency, with notes on how to fix each one.
The Cover Does Not Sell
The single most common reason books fail is the cover. It is also the easiest to fix.
A weak cover does not signal genre, does not stand out in thumbnails, and does not look like the books that sell in its category. Shoppers scroll past it. The algorithm notices the low click-through rate and shows the book less often. The cycle continues until the book disappears.
Common cover problems start with covers that look DIY. Authors who designed their own cover in Canva or Photoshop usually produce something that does not match category standards.
Covers that look like the wrong genre confuse shoppers who would have bought it if they recognized the genre. A literary cover on a romance novel sends readers elsewhere.
Covers that do not work as thumbnails fail at scale. Most browsing on Amazon happens on phones, where the cover appears at about the size of a postage stamp. Covers with small text or busy imagery fail at that size.
Covers that look dated stop working. Genre conventions shift every few years. A cover that looked current in 2018 may look amateur in 2026.
The fix is usually to hire a cover designer who has produced covers for books that sold in your genre. Costs range from three hundred to fifteen hundred dollars.
The Description Does Not Convert
The second most common failure is the description. Shoppers click in, read a few lines, and leave without buying.
Common description problems include descriptions that summarize the plot. Plot summaries do not sell books. Hook, tension, and stakes sell books.
Descriptions that run too long lose readers. Most shoppers read the first paragraph and skip the rest. If the hook is buried after eight hundred words of setup, the shopper is already gone.
Descriptions with no formatting get skipped. Walls of text are passed over. Strong descriptions use bold lines, short paragraphs, and visual breaks.
Descriptions that sound like every other book in the category produce generic results. Specific, voice-driven copy stands out.
Descriptions that tell but do not promise lose nonfiction shoppers. Shoppers want to know what they will get from reading the book. Vague language loses to specific outcomes.
Fixing the description costs nothing and can be done in an afternoon. Authors who rewrite their description and see sales lift have proven the description was the bottleneck. Authors who rewrite and see no change need to look elsewhere.
The Categories Are Wrong
Categories control where Amazon shows your book in browse contexts. Wrong categories make the book invisible.
Common category problems start with using only two categories. KDP allows ten through support. Authors who used only the default two miss eighty percent of their visibility.
Categories that are too broad kill ranking chances. A book in “Fantasy” cannot rank. A book in “Sword and Sorcery Fantasy” can.
Categories that do not match the book backfire. Some authors stuff their books into categories that do not fit, hoping for visibility. Amazon removes the book from those categories, and reader reviews call out the mismatch.
Categories that are too competitive for the book’s current state hide it. A debut novel in “Top Romance” is invisible. The same novel in “Marriage of Convenience Romance” can hit the top ten.
The fix is to research categories using Publisher Rocket or manual analysis, pick a mix of broad and narrow ones, and submit the full request through KDP support.
There Are No Reviews
A book with three reviews looks risky to shoppers. A book with thirty reviews looks safe. The conversion gap between the two is dramatic.
Common review problems start when the author did not build an advanced reader team. Most working authors recruit twenty to one hundred readers before launch who commit to reading the ARC and leaving an honest review on launch day. Authors who skip this step launch with zero reviews and watch their sales stall.
The back of the book often does not ask for a review. Most Kindle readers reach the end and look for the next thing to read. A clear ask at the back catches a percentage of them.
Some authors buy reviews. Amazon detects paid reviews and removes them. The author often gets the account suspended. Reviews have to be earned, not bought.
Sometimes the book itself is not good enough to earn reviews. Books that disappoint readers do not earn reviews even when asked.
The fix is to build a review pipeline before the next launch. ARC team, back-of-book ask, and engaged readers who care enough to write reviews.
The Pricing Is Wrong
Pricing affects both conversion and algorithm signals. Wrong pricing breaks both.
Books priced above genre norms underperform. A romance ebook at 7.99 will not sell when the genre median is 3.99. Shoppers do not perceive added value; they just see overpriced.
Books priced below norms can signal low quality. Nonfiction priced at 0.99 signals junk. Premium nonfiction in business or self-help often does best at 7.99 to 14.99.
Print pricing too close to ebook breaks the math. If the ebook is 4.99 and the print is 9.99, the print seems overpriced. If the print is 14.99, the ebook looks like a deal.
Series book one not discounted loses funnel readers. Series authors permafree or 0.99 book one and full-price the rest. Authors who full-price book one of a new series lose readers who would have entered the funnel cheaply.
The fix is to survey the top hundred in your category, find the median price, and align with it. Test small price changes of a dollar up or down over four-week windows to see what converts best.
The Sample Does Not Hook
Amazon shows shoppers a free sample of every book. For Kindle, the sample is usually the first ten percent. If the sample does not hook, the shopper does not buy.
The book may open slowly. Long setup, world-building, or backstory in the first chapter loses readers who expected to be hooked.
The first paragraphs may be weak. The hook needs to land in the first page, often in the first paragraph. Writers who buried the hook lose conversions.
The sample may end mid-scene. A sample that ends with a strong cliffhanger or open question converts at higher rates than one that closes a scene tidily and gives the reader an easy exit.
Front matter may eat the sample. If the table of contents, copyright, dedication, and acknowledgments take up half the sample, the actual book content is too short to hook.
The fix is to look at your sample as a shopper would. Trim front matter, sharpen the opening, and end on a hook.
There Is No External Traffic
Amazon’s current algorithm favors books that bring outside traffic. Books that exist only on Amazon and depend entirely on Amazon’s own systems for visibility often fail.
The author may have no email list. They have nowhere to point readers and no way to launch with a sales spike.
The author may have no social presence. The book has no organic conversation around it, so no one shows up to buy.
The author may have no podcast or media appearances. They never reach audiences outside their existing reach.
The author may have no content marketing. They publish a book and go silent.
The fix is to build outside channels. Email list first. Then one social platform like Bookstagram, BookTok, YouTube, or whatever fits your genre. Then content or media that brings new readers in.
The Book Itself Is Not Ready
Sometimes the issue is upstream. The book itself is not ready for the market.
Editing may be weak. The story or argument has holes. The prose is rough. The book gets one-star reviews from readers who finished it and were disappointed.
Structure may be weak. The book starts strong but loses focus. Readers abandon midway.
Genre fit may be wrong. The book is positioned as one genre but reads like another. Readers who bought based on the cover and description feel misled.
The market may be too small. The book is well-written but the audience is too small to support meaningful sales.
The fix here depends on the diagnosis. Books with editing issues can be revised and re-released. Books in genres with small audiences may need a category change to find their actual readers. Books that fundamentally do not work sometimes have to be retired and the next book written better.
The Author Stopped Marketing
A book launches, sells modestly, and then the author moves on. Sales decay, the algorithm notices, and the book fades.
Single launch push, then silence is the most common pattern. Sales decline as the algorithm reads silence as fading demand.
No newsletter promotion to existing readers is another. Authors who already have lists often forget to mention old titles. Backlist sales depend on cross-promotion.
No ongoing ads is a third. Books that ran ads at launch but stopped lose visibility that the ads were providing.
No content production is the last. Authors who used to publish regularly and have stopped lose audience momentum.
The fix is to treat marketing as ongoing rather than one-time. A monthly hour devoted to refreshing the listing, sending one email, and reviewing ads keeps the book alive.
The Author Tried to Game the System
Some authors fail because they tried shortcuts that triggered Amazon’s enforcement systems.
Bought reviews on Fiverr get detected. Reviews get removed. Account can be suspended.
KU page-read manipulation gets caught. Some authors paid services to fake page reads. Amazon banned thousands of accounts over this.
Keyword stuffing in the wrong places gets demoted. Hiding keywords in white text or stuffing them into the wrong fields. Amazon catches this and removes books.
Misrepresenting categories or genre backfires. Marketing a book as something it is not produces refunds, one-star reviews, and sometimes platform action.
There are no shortcuts. The platform rewards real demand for real books, and every shortcut is a trade against long-term sales.
How to Diagnose Your Own Book
If your book is failing, the diagnosis is usually one of the issues above. Some checks to run.
Look at your Amazon page as a shopper. Does the cover stand out. Does the description hook in the first paragraph. Are there reviews. Is the sample strong.
Pull your sales data. Are you getting impressions but no clicks, which is a cover issue, or clicks but no buys, which is a page issue, or buys but no read-through, which is a book issue.
Compare to books selling well in your category. What do they have that you do not. Where is the gap.
Talk to readers who bought and finished. Their feedback usually points to the real problem faster than guessing.
The Path Back
Most failed books can come back. Books that sat at zero sales for six months sometimes start moving when the author updates the cover, rewrites the description, requests more categories, builds reviews, and runs a small ad campaign.
Some books cannot recover. Books with fundamental issues in the writing itself need rewrites. But more often, the issue is in the marketing layer, and the marketing layer is fixable.
The authors who succeed long-term are not the ones whose first books worked. They are the ones who fixed what went wrong and applied the lessons to the next book.
