The barrier that fell was distribution. Anyone can put a book on sale in an afternoon, in a hundred countries, for nothing.
The barrier that did not fall is discovery, and most self-publishing advice quietly assumes it did. That single mismatch explains almost every disappointed first launch: the writer solved the problem the internet already solved, and then met the one it did not.
Everything useful about publishing independently in 2026 follows from taking that seriously.
What you actually get
Rights. You keep them, all of them, including the ones a traditional contract takes without anyone in the room quite noticing: audio, translation, film option, and the right to keep the book in print when the publisher decides it is done.
Timing. You publish when the book is ready rather than eighteen months later in a slot chosen for someone else’s catalogue.
Price control, which matters more than it sounds. You can run a first-in-series at 0.99 for a week and watch what happens to book two. No one has to approve that.
And a direct relationship with readers. Their email addresses belong to you rather than to a retailer.
The arithmetic, honestly
The royalty split is the number everyone quotes and the one most often misunderstood. Amazon’s KDP pays 70% in its main price band and 35% outside it, minus a delivery fee that scales with file size, which is why image-heavy books earn less than authors expect. Outside that band, a 9.99 book and a 2.99 book can net you similar amounts. Read the current terms yourself rather than trusting a blog from three years ago, including this one.
Against that, a traditional deal typically pays 25% of net on ebooks, and an advance you have to earn out before you see anything further.
So the per-copy maths favours indie by a wide margin, and that is exactly what makes the discovery problem so expensive. 70% of nothing is nothing. The traditional publisher is selling you access to shelves, reviewers, and a sales force, and the honest question is not which royalty rate is higher but whether you can reach readers without those things.
What it costs to do properly
Editing is the line people cut and should not. A developmental edit and a copy-edit are different jobs and a good copy-editor will not fix a structural problem.
A cover that signals genre correctly. This is not about beauty. It is about a reader recognising in half a second what kind of book this is, and a wrong-genre cover kills a launch more reliably than a weak first chapter.
Formatting, which is cheap or free depending on your patience.
And then the part with no ceiling: marketing, forever. That is the real cost, and it is paid in time rather than money.
Serialisation changed the maths for a lot of people
Publishing a chapter at a time to an audience that accumulates solves a version of the discovery problem, because the platform does some of the finding for you and the reader relationship builds while you write.
It works when your genre has a serial audience and you can sustain the pace. Progression fantasy, romance, anything with a strong episodic pull. It works badly for literary fiction, for anything whose pleasure is structural and only visible whole, and for anyone who cannot commit to a schedule for a year. A serial that goes quiet for three weeks loses readers it does not get back.
The trade is real: you are exchanging some control and some revenue for discovery you cannot otherwise buy.
Direct sales, newsletters, and what the platform is offering you
The case for selling direct is that you keep more and you own the customer list. The case against is that you become responsible for payment, delivery, tax handling, piracy, and support, and none of that is writing.
A newsletter is the durable asset. Platforms change their terms, their algorithms and their exclusivity rules; a list of readers who asked to hear from you does not. Most working indie authors will tell you their list matters more than any single platform relationship.
What the platform offers in exchange for exclusivity is visibility, and it is a real offer. Whether it is a good trade depends on whether your readers are already there. That is an empirical question about your genre, not a matter of principle.
How long it takes
This is the part most guides omit, and the omission is why so many writers conclude they failed.
A first book by an unknown author, published well, usually sells in the low hundreds in its first year. Not because it is bad. Because nobody knows it exists, and one book gives a reader nowhere to go next.
The thing that changes the curve is a backlist. Three books in a series, with the first priced low, is a different commercial object from one book. Readers who finish book one and immediately buy two and three are the entire model. Almost everything about indie publishing that looks like magic is that.
Plan on that timeline. A first year that produces a small number of sales and a finished second manuscript is a good first year, and it does not feel like one.
What to decide before you publish anything
Whether this book is a series or a standalone, because that changes almost every other decision.
Whether your genre’s readers are on a serialisation platform, on retailers, or both.
Whether you can write the next one. Not “would you like to” — whether the second book exists in any form. If it does not, the first one is going to sit alone for a while, and that is worth knowing in advance rather than discovering in month seven.
None of that requires a decision about tools, and it is worth making before the part everyone starts with, which is picking a cover.