How to Make Money with Amazon KDP – What It Actually Pays in 2026
While we were working on the Amazon Merch on Demand side hustle for this site, it got me thinking about Amazon’s other “upload it once and it sells itself” program: Kindle Direct Publishing, or KDP. Same idea as Merch – you make something once, Amazon handles the printing and the customer and the shipping, and you collect a royalty every time it sells. That’s the same hands-off mechanic behind most of the passive income ideas on this site – except instead of a t-shirt design, it’s a book.
Here’s what we found, the real math, the trap almost nobody mentions, and the one corner of this I found really interesting’
How the royalty actually breaks down
KDP has two separate royalty systems depending on what you publish, and the percentages everyone throws around aren’t the whole story.
For ebooks, you get 70% royalty if you price between $2.99 and $9.99, and 35% on anything priced outside that range. Sounds simple, except Amazon also takes a delivery fee of $0.15 per megabyte of your file size out of that 70% cut. If your file is small – just text – that barely matters. If it’s loaded with images, that “70%” can quietly shrink toward something closer to the 35% tier.
Print books pay differently, and the rules actually changed back in mid-2025. Now you get 60% of list price minus the printing cost if your book is priced at $9.99 or higher, but only 50% if you price it below that. Same printing cost either way – so pricing a hair too low doesn’t just mean a cheaper book, it means a real cut to your royalty rate per copy.
There’s a third option if you enroll in Kindle Unlimited through KDP Select: instead of getting paid per sale, you get paid per page read, through something Amazon calls KENP (Kindle Edition Normalized Pages). The going rate has been sitting around $0.004 to $0.005 per page. A 300-page book read cover to cover nets you somewhere around a dollar to a dollar fifty – less than a straight $4.99 sale would pay. It can work out if people binge through a whole series, but it’s not automatically the better deal.
The trap nobody talks about: AI disclosure
This is the part that would’ve bitten me if I hadn’t dug in, so pay attention here.
Amazon requires you to disclose it if AI generated the actual text, images, or translations in your book. You do not have to disclose using AI to brainstorm ideas, build an outline, or clean up grammar on something you wrote yourself – that’s “AI-assisted” and it’s fine. But if you lean on AI to draft the actual paragraphs that end up in the book with only light edits, that’s “AI-generated,” and Amazon wants you to check the box.
Skip the disclosure and get caught, and the penalties aren’t a slap on the wrist. Amazon can pull the book without warning, and it can withhold royalties already earned on that title. Amazon has also started going back through previously published books, flagging older titles that look AI-generated and demanding the disclosure get added retroactively. Repeated violations can get your whole account suspended. With detection getting better every year, this isn’t a corner worth cutting – if you’re using AI heavily to write the manuscript itself, just check the box. It costs you nothing in visibility and saves you the account.
The honest income picture
I’m not going to sell you a dream here, because the actual numbers tell a different story than the YouTube thumbnails do.
A decently positioned low-content title – think notebooks, planners, journals – might sell around 20 copies a month, earning somewhere between $20 and $70 for that one title. To clear something like $1,000 a month from low-content books alone, you’re typically looking at having 15 to 50 titles out there pulling their weight, not one or two. Most people who publish just a book or two with no real keyword strategy earn less than $100 total in their first few months.
And the obvious niches – generic notebooks, sketchbooks, basic planners – are about as crowded as it gets right now. Everyone with a laptop and an AI tool can crank one of those out in an afternoon, and a lot of people have. That’s not a reason to skip KDP entirely. It’s a reason to skip the obvious stuff.
Where I’d actually look, if I tried this
Here’s the part that connects back to why I looked at this in the first place. The saturated KDP niches are the ones anybody can make – generic journals, generic planners, the stuff with zero specific knowledge baked in. What’s not saturated, because it actually requires knowing something, is a log book or planner built around a specific trade.
Think about what a generic AI-content publisher has no way of getting right: an apprenticeship hour log formatted the way your state’s licensing board actually wants it tracked, a daily job-site log with the fields a foreman actually needs, a tool inventory log sized for what a mechanic or electrician actually carries, or a no-nonsense study companion for a trade certification exam. None of that is hard to build once you know the trade. But you have to know the trade to build it right, and that’s exactly the kind of edge a generic publisher chasing volume doesn’t have and can’t fake.
That’s a smaller, slower niche than “publish 50 notebooks and hope.” But it’s the kind of niche where the thing you already know from working the trade is worth more than another AI tool subscription.
If you want to try it
Setting up an account costs nothing – you can start a KDP account and look around before committing to anything. My honest advice: don’t start with a generic notebook just to test the waters. Pick one specific log or guide tied to your actual trade, build it carefully, and see what happens before deciding whether to scale it into a second or third title. This works the same way print on demand does – the people doing fine with it treated it like a real small project, not a lottery ticket.
If you want more side hustle ideas that actually fit a trade background instead of generic gig work, I put together a full list of side hustles for mechanics and tradesmen worth a look too.
Frequently Asked Questions
It can be, but not as a quick win. The generic low-content niches like basic notebooks and journals are heavily saturated. Specific, well-built niches tied to real knowledge – like trade-specific log books – still have room, but it takes research and patience either way.
Only if AI generated the actual text, images, or translations in the book. Using AI to brainstorm, outline, or check grammar on your own writing doesn’t need disclosure. If AI wrote the passages that ended up in the book, check the box – skipping it risks the book being removed or your account being flagged.
Most beginners earn under $100 in their first few months with just one or two titles. Building toward $1,000 a month from low-content books typically takes 15 to 50 titles performing well, not a single book. Specific niches with real demand can move faster, but there’s no shortcut around the early slow period.
Ebooks priced $2.99 to $9.99 earn 70% royalty minus a delivery fee based on file size; anything outside that range earns 35%. Print books earn 60% of list price minus printing costs if priced at $9.99 or above, but only 50% below that threshold.
Yes, more than you’d think. The most saturated KDP niches are generic notebooks and planners anyone can make. Trade-specific formats – apprenticeship hour logs, job-site logs, tool inventories, certification study guides – require knowing the trade to build correctly, which keeps the competition thinner.
