How I Used Claude to Write 12 Ebooks in 2026

I published my first ebook in March 2026 and made back my monthly Claude subscription in 11 days. I want to be upfront about that number: it wasn’t a fortune. But for a one-person real estate consulting business in Madeira that had never sold a digital product before, clearing costs in under two weeks felt like proof that this approach was worth documenting honestly.

What I want to walk you through here isn’t a fantasy about passive income. It’s the exact process I used — the prompts, the workflow, the stumbles, and the real numbers — so you can decide whether writing and selling ebooks with Claude makes sense for your solo operation.

Why I Decided to Write an Ebook at All

My consulting business runs on referrals and repeat clients. That model works, but it’s entirely dependent on my time. Every euro I earn requires me to show up, take calls, tour properties, write reports. There’s no product that keeps selling while I’m having lunch in Funchal.

In late 2025, I started noticing the same questions showing up again and again — from friends, from LinkedIn connections, from people who found me through my newsletter. “How does buying property in Madeira actually work if I’m not Portuguese?” “What do foreigners always get wrong about the process?” I was answering these questions individually, sometimes spending 30 minutes on a detailed email reply that only one person would ever read.

That’s the problem Claude helped me solve. Not by writing something generic, but by helping me turn seven years of client experience into a structured, useful guide that I could sell once and deliver automatically.

The Exact Process I Used: From Blank Page to Published Ebook

Step 1: Using Claude to Build the Structure First

I did not start by telling Claude to “write me an ebook about buying property in Madeira.” That produces generic content that reads like a Wikipedia summary. Instead, I had a conversation.

I opened a Claude Pro session (I’m on the $20/month plan, which I’ve been using since early 2024) and gave it this context: my background, my typical client profile (Northern European retirees and remote workers, usually first-time buyers in Portugal), the top 8 questions I get asked repeatedly, and the mistakes I see clients make most often.

Then I asked Claude to propose a chapter structure for a practical guide — not a textbook, not a legal manual, but something a nervous buyer could read on a flight to Lisbon and feel genuinely more prepared. It gave me a 9-chapter outline in about 90 seconds. I pushed back on two chapters, merged another two, and we landed on a 7-chapter structure in roughly 15 minutes of back-and-forth.

Step 2: Writing Chapter by Chapter With Context Loaded In

This is where Claude’s long context window actually earns its keep. I kept the full outline in every conversation, added my personal notes for each chapter (bullet points, client stories, specific warnings), and asked Claude to draft each chapter in my voice — direct, first-person where relevant, no legal hedging, no padding.

After each draft I’d read it out loud. Anything that sounded like an AI wrote it, I flagged and asked Claude to rewrite with more specificity. “Don’t say ‘it’s important to consult a lawyer.’ Tell them what kind of lawyer, what to ask, and roughly what it costs in Madeira.” That kind of redirecting turned generic drafts into genuinely useful content.

Seven chapters, each between 900 and 1,400 words. Total writing time with Claude: about 6 hours across 4 days. My own editing and fact-checking on top of that: another 3 hours. Compare that to my estimate of 25–30 hours if I’d written this from scratch on my own. Even being conservative, Claude cut the production time by roughly 70%.

Step 3: Generating the Sales Page and Email Sequence

Once the ebook was done, I used Claude to write the sales page copy, a 4-email welcome and delivery sequence, and a short “teaser” post for my newsletter. For the sales page, I gave Claude the chapter list, the target buyer profile, and three specific objections I knew readers would have (“Is this just for wealthy investors?” / “I don’t speak Portuguese, is this relevant?” / “Can’t I find this on Google?”). The copy it produced needed light editing — maybe 20 minutes of tweaking — and it was live.

My Real-World Experience: Numbers, Messiness, and What Actually Happened

The ebook is called Buying Property in Madeira: What Nobody Tells You. I priced it at €27. I sell it through Gumroad, which takes a small cut, and I drive traffic through my newsletter (about 1,100 subscribers at the time of launch) and occasional LinkedIn posts.

In the first 30 days: 47 sales. That’s €1,269 in gross revenue before Gumroad fees. My total cost to produce the ebook was my Claude Pro subscription for that month (€20) plus maybe 2 hours of formatting time in Canva for the PDF layout. I did not hire a designer. I did not run ads.

What surprised me wasn’t the revenue — it’s modest, I won’t pretend otherwise. What surprised me was the quality of leads it generated. Three of those 47 buyers later booked paid consulting calls with me. One is now a serious buyer I’m actively working with on a property search. The ebook became a qualifier. People who read it arrive on a call already trusting me, already aligned with how I work. That’s worth more than the €27.

The part that didn’t go smoothly: the first draft of Chapter 4 — about the legal process and fiscal representation — was noticeably weaker than the others. Claude kept defaulting to cautious, hedge-everything language (“consult a qualified professional before taking any action”) that would have frustrated any reader looking for real guidance. I had to rewrite that chapter more heavily myself, probably 40% of it from scratch. Claude was genuinely useful for structure and flow, but it struggled with content that sits at the edge of legal advice. That’s a real limitation you need to plan for if your ebook touches anything regulatory, financial, or jurisdiction-specific.

I also found that Claude occasionally lost the thread of my voice between sessions. When I came back to a new conversation after a day away, the tone would drift slightly more formal, more generic. My fix was to paste in a short “voice reference” paragraph at the start of each new session — a sample of my writing — so Claude had something to calibrate against. It added maybe 3 minutes per session but made a real difference in consistency.

Six months on, the ebook has made €2,890 total with zero ongoing effort from me. That’s not life-changing money. But it’s real revenue from something that took less than two full working days to produce, and it keeps coming in while I’m doing everything else.

How This Workflow Compares to Other Approaches I Tried

Approach	Time Investment	Cost	Quality Control	My Experience
Writing from scratch	25–30 hours	$0 (just my time)	Full control	Never finished the one I started in 2023
Hiring a ghostwriter	5–8 hours (briefing + review)	€800–€2,000+	Depends on writer	Got one quote of €1,400. Passed.
Claude Pro (what I use)	9 hours total	€20/month	You edit everything	Published in 4 days, €2,890 in 6 months
ChatGPT (tested briefly)	Similar to Claude	$20/month	You edit everything	More generic output; I needed more rewrites

The Genuine Limitations You Should Know Before Starting

Claude is not a subject-matter expert. It’s an extremely capable writing collaborator. The difference matters. For my ebook to be accurate, I had to supply the expertise — the real client stories, the specific legal steps in Portugal, the correct names of documents, the actual fees involved. Claude organized and articulated that knowledge. It did not create it.

If you try to write an ebook on a topic you don’t genuinely know well, Claude will produce something that reads confidently but contains errors. Your readers will find them. That’s a trust problem that’s very hard to recover from.

A few other honest observations from 4+ months of using this workflow:

• Voice drift is real. Claude will subtly shift tone between sessions unless you actively manage it with a voice reference.
• Regulatory content needs heavy human review. Any chapter touching legal, tax, or compliance topics requires you to verify every claim independently. Claude will sound confident even when it’s slightly wrong.
• The ebook still needs human editing. Plan for at least 2–3 hours of editing per 10,000 words. Claude gets you 70% of the way there, not 100%.
• Marketing is entirely on you. Claude can write your sales page. It cannot build your audience. If you don’t have a newsletter, a social following, or some existing platform, the ebook will sit there quietly.

What I’d Do Differently If I Started Over Today

I’d record myself first. Before writing a single word, I’d spend 45 minutes answering my own chapter questions out loud — into a voice memo or Otter.ai — and then feed those transcripts to Claude as raw material. The output would have sounded more naturally like me from the start, and I’d have spent less time correcting the tone in editing.

I’d also set up a simple landing page before the ebook was finished to collect email interest. I waited until the product was done to announce it, which meant I had no pre-launch momentum. Even 20 people on a waitlist changes how a launch feels.

And I’d price higher. €27 felt safe. In hindsight, the buyers who went on to book consulting calls would have paid €47 without blinking. I left money on the table being cautious about a product I hadn’t tested yet.

My Rating: Claude for Ebook Writing as a Solopreneur

4.2 out of 5. Claude cut my ebook production time from an estimated 30 hours to 9 hours and produced content that was genuinely useful — not just readable — because I supplied the real expertise and pushed back on anything too generic. The gap from 4.2 to 5 is entirely explained by the voice drift between sessions and the weakness on regulatory content, both of which added friction I hadn’t anticipated.

Practical Summary: What to Take Away From This

Here’s what the process actually looks like, stripped down:

1. Identify the questions you already answer for free. Your ebook topic is hiding in your sent folder and your DMs.
2. Feed Claude your expertise, not the other way around. Give it your notes, your client stories, your real experience. It structures and writes. You supply the substance.
3. Use a voice reference at the start of every new session. Paste in 150 words of your own writing to anchor the tone.
4. Edit heavily on anything legal, financial, or jurisdiction-specific. Don’t publish it until you’ve personally verified every claim in those sections.
5. Write the sales page and email sequence in the same Claude workflow. It already knows the ebook. Use that context.
6. Get it out fast.** Don’t spend six weeks polishing. A finished ebook at 85% perfect earns more than a perfect ebook that never ships.

If you’ve been sitting on expertise that you give away for free in emails and calls, this is a genuinely low-cost way to productize it. My total investment was €20 and 9 hours of focused work. The ebook is still selling.

Want to see exactly which Claude prompts I used for each stage — outline, chapter drafts, sales page, and email sequence? I’ve put together a short prompt pack based on what actually worked. Grab it here and you’ll have a working template instead of starting from scratch.