March 13, 2026 · 10 min read

KDP Income Report: Real Numbers from 12 Weeks of AI-Assisted Publishing

Why I'm Publishing Real Numbers

The KDP income report space is a mess. You've seen the screenshots: €10,000 months, Lamborghini dashboards, "passive income" promises that conveniently leave out ad spend, returns, and the 200 books that earned nothing.

I'm not here to sell you a course. I'm here to show you what actually happened over 12 weeks of publishing AI-assisted books on Amazon KDP. The numbers are modest. Some weeks were bad. A few books flopped completely. That's the point.

If you're considering AI-assisted publishing, you deserve real data, not cherry-picked highlights from someone's best month ever. So here it is. No filters, no inflated screenshots, no guru promises.

The Setup

Here's what we're working with:

• Workflow: AI-assisted drafts using WriteAIBook, followed by a human editing pass. Every book gets at least one read-through and structural edit before publishing.
• Pen names: Multiple pen names across different genres. This keeps branding clean and lets me test niches without cross-contamination.
• Distribution: KDP Select (Kindle Unlimited exclusive). Page reads are the primary revenue driver, not direct sales.
• Starting point: I had an existing catalog before this 12-week window. These numbers are specifically from the latest batch of new titles.

Average time per book from idea to published: about 3-4 hours. That includes generation, editing, formatting, cover (using pre-made templates), and KDP upload. The AI does the heavy lifting on the draft. The human work is editing, keyword research, and category placement.

Week-by-Week Breakdown

Here's the actual data, broken into three 4-week blocks. Revenue includes both direct sales and Kindle Unlimited page reads. Ad spend is Amazon Ads only.

Yes, that's €350 profit in 12 weeks. I told you I wasn't going to inflate anything.

The upward trend matters more than the absolute number. Weeks 9-12 earned more than double weeks 1-4, mostly because earlier books started gaining traction in KU and a couple of series books began pulling readers through.

What Worked

A few clear patterns emerged:

• Series outperformed standalones by 3-4x. Readers who finish Book 1 often read Book 2 the same day. KU page reads compound when you have a series. My best-performing pen name has a 4-book cozy mystery series that accounts for nearly 40% of total page reads.
• KU page reads are the real revenue driver. Direct sales were minimal. Almost all revenue came from Kindle Unlimited borrows. This is typical for newer, unknown pen names.
• Category placement matters more than you think. Two nearly identical books in the same genre performed completely differently because one landed in a less competitive sub-category. Spend the 20 minutes researching categories.
• Romance and mystery sub-genres were the most consistent earners. Not groundbreaking insight, but the data confirmed it. These genres have voracious KU readers who binge-read series.
• Covers from pre-made templates did fine. I didn't spend €200 per cover. Pre-made templates in the €5-15 range performed well enough. The cover needs to be genre-appropriate, not award-winning.

What Didn't Work

Plenty of things failed. Here's what I'd avoid next time:

• One-off standalones in competitive genres. I published 3 standalone sci-fi novels. Combined revenue after 12 weeks: €11. Without a series to pull readers through, standalone books from unknown authors just don't get traction in KU.
• High return rates on one title. One book had a 28% return rate. The blurb promised something the content didn't deliver. Lesson: make sure your blurb accurately represents the book, especially the heat level in romance.
• Broad-match Amazon Ads in the first two weeks. I burned about €18 on broad-match campaigns that generated clicks but zero sales. Switched to exact-match targeting on specific comp titles, and ACOS dropped from 400% to about 80%.
• Publishing too fast without editing. Two early books went up with minimal editing. Both got 3-star reviews mentioning "repetitive" and "felt AI-generated." Those reviews tank your conversion rate. I pulled both, edited them properly, and re-uploaded. The damage was already done on one of them.
• Non-fiction in a saturated niche. I tried 2 non-fiction books (productivity/habits category). Combined earnings: €4. The category is dominated by established authors with hundreds of reviews. AI-assisted non-fiction needs a much sharper niche angle than I gave it.

The Honest Math

Let's be blunt about the numbers:

• 27 books published in 12 weeks
• €437 gross revenue
• €87 ad spend
• €350 net profit
• That's roughly €13 per book net

Not life-changing. Not even rent money. But here's what that number misses: those books keep earning.

KDP books don't stop selling after 12 weeks. The revenue curve flattens but doesn't die, especially for series with read-through. Based on the trailing trend, this batch of 27 books is projected to generate roughly €1,200-1,800 over the next 12 months with zero additional work.

The math only makes sense when you factor in the time investment. At 3-4 hours per book and 27 books, that's roughly 80-100 hours of work. If the annual projection holds at €1,500, that's about €15-19/hour for the initial work, plus ongoing passive income as the catalog ages.

What I'd Change

If I could rewind 12 weeks, here's what I'd do differently:

1. More series, fewer standalones. I'd allocate at least 80% of output to series books. The data is unambiguous: series earn more, retain readers, and compound faster.
2. Better keyword research upfront. I spent too little time on category and keyword research for the first batch. The books that performed best had the most deliberate keyword targeting. Use Publisher Rocket or at minimum browse KDP bestseller lists in your target categories.
3. Kill underperforming ads faster. I let losing ad campaigns run for 7-10 days before pausing them. Should have set a 3-day review cadence: if ACOS is above 150% after 3 days with decent impressions, pause and re-target.
4. More editing time on top performers. When a book starts getting traction, invest more time editing and polishing it. Add bonus content, fix any rough spots, and make sure the "Look Inside" preview is flawless. A book earning €20/month is worth 2 extra hours of editing.
5. Skip non-fiction entirely (for now). The AI-assisted non-fiction space needs a much stronger niche angle than I gave it. I'd focus 100% on fiction series in proven KU genres until the catalog hits 50+ titles.

The Bigger Picture

This is one quarter. Twelve weeks. If I keep this pace for a full year, that's 100+ books generating passive income. The compounding effect is what makes AI-assisted publishing viable as a business model.

Quarter 1: 27 books, €350 profit.
Projected Quarter 2: 30+ books (more series-focused), €500-700 profit from new titles, plus €300-450 from Q1 catalog still earning.
By end of Year 1: 100+ titles, projected €400-600/month passive income.

That's not a Lamborghini. It's a car payment. But it's real, it's documented, and it keeps growing without requiring proportional increases in work. The catalog is the asset.

The people showing you €10k months either have 500+ book catalogs built over years, are spending thousands on ads, or are lying. Probably a mix of all three. The honest path is slower, but it's the one that actually works for normal people with normal schedules.

If you want to try this yourself, the barrier to entry is genuinely low. Generate your first draft for free, edit it properly, and publish. Then do it again. And again. That's the entire strategy.