Per-copy royalties from Amazon KDP (about $7 net per paperback, per author); catalogue of titles in one niche
Summary
Write short nonfiction books (50–150 pages) that each solve one specific reader problem. Draft them with AI, then self-publish as paperback, Kindle and hardcover on Amazon KDP. Amazon prints, ships and handles support. You build a catalogue of books in one niche so each new title helps sell the older ones. Source: Tommi Pedruzzi's 11-step thread (29 Oct 2026, AEDT).
Who buys and why
Amazon book buyers searching for help with a specific, long-lasting problem. His examples are "sleep after menopause", "caring for aging parents" and "budgeting on a single income". Buyers already have cards saved on Amazon (he cites 310 million Prime members), so you publish into demand that already exists rather than building an audience first.
Step-by-step setup
- Find a lasting problem. Search the topic in Google Trends over 10+ years. A flat or rising line means a lasting market; a spike that fell means a fad.
- Go specific. One reader, one problem (not "health" but "sleep after menopause").
- Prove people buy. Search the topic in Amazon Books and check Best Sellers Rank. Under 100,000 means it sells; over 1,000,000 means nobody buys. Confirm a paperback nets $5+.
- Mine 3-star reviews of the top books. Copy 10 complaints into a doc. These gaps are your book's reason to exist.
- Outline with ChatGPT. Give it the gaps and ask for a 10-chapter plan for one reader, each chapter fixing a gap (about 15 minutes).
- Write with Claude, one chapter at a time, about 1,000 words each, with your tone.
- Polish with Grammarly and Hemingway (about 1 hour): fix grammar and shorten long sentences.
- Design the cover in Ideogram (free). Match the fonts and colours of the top 5 bestsellers, with a different title.
- Upload to KDP under a pen name. List paperback, Kindle and hardcover. Tick the AI-disclosure box (it is required). Price the paperback at $12.99–16.99. One upload reaches 17 marketplaces.
- Launch at $0.99 for 5 days. Share in 3 niche Facebook groups and message 10–15 people for honest reviews. At 15–25 reviews Amazon starts recommending the book. Never buy reviews.
- Repeat in the same niche so the titles cross-sell.
Tools and AI agents used
Google Trends, Amazon Best Sellers Rank, ChatGPT (outline), Claude (chapters), Grammarly, Hemingway Editor, Ideogram (cover), Amazon KDP.
Costs (startup and monthly)
- The thread gives no cost figures.
- The tools named are free or low-cost: Ideogram is described as free, and KDP has no upload fee.
- The $0.99 launch week gives up royalty income.
- Our estimate: under $50 a month for ChatGPT/Claude/Grammarly subscriptions.
Reported results (claims, credibility, course?)
- The author says the system "added $743,000" in the last 11 months.
- He links three case studies: a Barcelona couple making $100k from one book, an actor making $8,500 a month, and an IT professional making $4,000 in a month.
- His own maths: a paperback nets about $7. One copy a day in each of 4 markets is about $840 a month per book; 12 books is about $10,080 a month.
- He says "I'm not saying everyone hits that number", and that anyone following the system can build toward $3,000 a month.
- Credibility: the claims are not substantiated. He is a course seller; the thread ends with a free live masterclass that leads into his paid programme. The 11 steps themselves are given in full and free.
Time to profit (reported or estimated, assumptions, weekly effort, ongoing work)
- Reported: new books take 2–3 weeks to rank, "so expect a ramp".
- Estimate (ours): 1–3 months to cover the small tool costs, assuming one book is published in the first month and gets 15–25 reviews.
- Effort (ours): a book takes about a week of part-time work using the times in the thread (15-minute outline, hours of drafting, 1 hour of polish, plus research and cover).
- Ongoing work: publish more titles in the niche, gather reviews, and later the ads he mentions.
Skills and interests it suits
People who like research and writing, and anyone with real knowledge of a niche problem (health, parenting, money, care-giving). No audience or coding needed.
Smallest useful test
- Live on stream: pick one niche, run the Trends, BSR and 3-star-review checks, and generate the 10-chapter outline plus chapter 1 with Claude.
- Publish one book within a week and track BSR and reviews over 30 days.
Links
Our results
<empty-block/>