
The real economics of self-publishing a children's picture book
Most authors find out how picture book economics work after they've already spent the money. This post is for the ones who want to know before.
The math is not encouraging. But it's better to understand it clearly than to discover it the hard way.
The royalty problem
Amazon KDP is where most self-published picture books land. It's easy to set up, free to list, and reaches a global audience. The catch is the royalty structure.
For fixed-layout ebooks (which is what every picture book is), KDP effectively pays 35% of the list price. Technically, authors can choose a 70% tier, but it charges a per-MB delivery fee that makes it unviable for image-heavy files. A 20MB picture book can earn negative royalties on the 70% plan. Most picture book authors end up at 35%. That's not 35% of profit. That's 35% of what the customer pays.
Price your book at $4.99, which is a reasonable entry point for a children's ebook, and you earn $1.75 per sale.
To earn $35,000 in a year at that rate, you need to sell just over 20,000 copies.
Industry estimates consistently suggest the median self-published picture book sells fewer than 100 copies in its lifetime.
This isn't a failure of ambition. It's a structural problem with the model.
The full cost stack
Royalties are only part of the picture. Before your book earns a single dollar, you've likely spent on several of these:
Illustration. This is the biggest line item for most picture books. A professional illustrator for a 32-page picture book typically charges between $5,000 and $20,000, depending on style, experience, and scope. Some work for less. Very few work for more than that on a first-time author collaboration.
Editing and proofreading. A developmental editor who understands picture books runs $500 to $2,000. Copy editing and proofreading add another $200 to $500. Skipping this step is a false economy.
Layout and formatting. Fixed-layout ebooks require precise formatting that most authors can't do themselves. A professional formatter charges $300 to $800 for a picture book.
ISBN and distribution setup. A single ISBN through Bowker costs $125. A block of ten is $295. Some platforms provide free ISBNs, but owning your own gives you more control.
Marketing. There is no floor here and no ceiling. Authors who spend nothing on marketing sell to family and friends. Authors who spend on ads, ARCs, blog tours, and social content spend anywhere from a few hundred to several thousand dollars, often with limited measurable return.
A realistic minimum budget for a professionally produced self-published picture book, before any marketing: $6,000 to $8,000.
At $1.75 per sale, you need to sell between 3,400 and 4,600 copies just to break even.
The translation question
At some point, most picture book authors wonder about translation. The story works. The illustrations are beautiful. Wouldn't it reach more readers in other languages?
The answer is yes. The economics, under the current model, make it nearly impossible.
Professional literary translation for a picture book runs $300 to $2,300 per language, depending on the language pair and translator. That's for text only. Narration, if you want audio, adds $200 to $500 or more per finished hour, per language.
Translating into 10 languages: $3,000 to $23,000, before narration.
Translating into 28 languages: $8,400 to $64,400, before narration.
Narration adds another $5,000 to $14,000 for 28 languages at conservative estimates. That's a separate line item that Redda also covers, at no cost to the author.
For an author who has already spent $8,000 producing the original and is earning $1.75 per sale, this is not a real option. It's a fantasy.
This matters because multilingual families are one of the most underserved and highest-intent audiences for children's books. They are actively looking for stories in their languages. They aren't finding them, because the economics make it nearly impossible for most authors to get there.
What the data on discovery actually looks like
KDP's catalog has millions of titles. A new picture book from an unknown author, with no pre-existing audience and no marketing budget, is competing in one of the most crowded retail environments ever built.
Amazon's algorithm rewards sales velocity. Books that sell quickly get visibility. Books that don't sell quickly don't get visibility. For a debut author with no audience, this is close to a closed loop.
Some authors break through. They usually have one or more of: a large social media following before publication, a PR budget, a professional marketing background, or exceptional luck with timing and topic. These are not replicable formulas.
The honest assessment: discovery on Amazon for a debut picture book, without significant marketing investment, is very unlikely.
A different model
The economics above describe the dominant model. They aren't the only model.
The structural problems are specific: 35% royalties on fixed-layout ebooks, one-time sales with no recurring revenue, translation costs that scale linearly per language, and distribution to a general audience rather than a targeted one.
A model that addressed those problems would look different. It would pay per read rather than per sale, so revenue compounds with every reading. It would handle translation and narration at no upfront cost to the author, funding itself through the revenue share. And it would distribute to an audience that is already looking for what the author has made: multilingual families searching for stories in their languages.
That model exists. It's what Redda is built around. Authors on Redda earn on every read, in every language. A story read 1,000 times across 28 languages is 28,000 earning events, not one sale.
If you've done the math above and felt the ceiling close in, it's worth understanding what a different structure looks like before you decide the traditional route is your only option.
Publish your story in 28 languages at no upfront cost. Learn more at reddastories.com/for-authors.
Sources: Amazon KDP eBook Royalties. KDP Digital Book Pricing.