What separates the self-published books that earn from the ones that do not
The writing is the easy part. Niche, cover, category, distribution — that is where the math actually decides.
Most people who self-publish spend months on the manuscript and about four hours on everything else. That ratio is backwards. The writing is the easiest part to get right. The niche, the cover, the category selection, the distribution: those are where the money either happens or it does not.
The niche decision
The default instinct is to write about something you know. That is a fine starting point, but it is not a publishing strategy. What actually sells is the intersection of genuine demand and low competition: a topic people are actively searching for where the existing books are mediocre, outdated, or missing the real audience.
Amazon's bestseller rankings tell you what is selling right now. Reddit tells you what people actually struggle with. Competitor reviews tell you what the existing books get wrong. Most self-publishers check none of these before they write. They publish into a void and wonder why nothing moves.
The cover decision
Covers are not art. They are signals. A reader scanning Amazon results at 160 pixels wide needs to instantly feel: this is for me, this looks credible, this is worth clicking. Most self-published covers fail that test in under a second.
Genre conventions exist for a reason. Readers of business books expect one visual language. Readers of wellness books expect another. Stepping outside those conventions, even with something genuinely beautiful, reads as amateur and tanks click-through rates before the title gets a chance.
The books that earn are not the best-written ones. They are the ones that look right, land in the right category, and solve a problem people are already searching for.
The distribution decision
Publishing exclusively to Amazon is leaving most of your potential revenue on the table. Apple Books, Kobo, Barnes and Noble, Scribd: readers exist on all of them, and many never touch Amazon. Setting up distribution to 20-plus stores manually means separate accounts, separate uploads, separate metadata, separate pricing. Most authors do not bother. The ones who do often get the metadata wrong anyway.
What changes when you fix all three
A book in the right niche, with a cover that converts, distributed everywhere: that is a different product from one that gets one thing right. The ceiling is completely different. So is the floor: it does not die in obscurity because nobody could find it.
The split between authors who earn and authors who do not is rarely about prose quality. It is about whether the book is positioned to be found by the people who would want it, in the form that signals "this is the one." Get those three decisions right before the writing starts, and the writing becomes the easy part it was supposed to be.