How to write and sell an eBook in 72 hours using AI (step-by-step)
The global eBook market hit $26 billion in 2025 — here's your 72-hour roadmap to claiming a slice of it.
The ebook business has a version that doesn’t work: someone generates 10,000 words of AI slop, slaps a cover on it in ten minutes, and waits for passive income to materialize. It doesn’t. Then there’s the version that does work: you identify a specific problem a specific group of people desperately want solved, use AI to compress weeks of writing into a weekend, package it professionally, and put it in front of the right buyers before most people have finished overthinking their topic. The second version is not a fantasy. The global ebooks and audiobooks market reached $26 billion in 2025, according to market data compiled by Accio. That growth isn’t coming from novels people buy at airports. It’s coming from practical, problem-solving guides that people find at midnight because they have a problem they need solved by Tuesday.
This is the step-by-step version of that process, compressed into 72 hours. Not because speed is the point, but because the biggest enemy of your first digital product isn’t quality — it’s the decision paralysis that keeps you from starting.
Hour 0–8: pick a topic that actually has buyers
The most expensive mistake in this entire process costs you nothing at the time. It’s writing an ebook nobody wants. Most guides tell you to “follow your passion.” That’s romantic. It’s also how you end up with a beautifully written guide that earns $11 in its first month. 📚
The filter that actually works is this: does your topic solve a specific, painful problem for a defined group of people who are already spending money to solve it? “Productivity tips for everyone” fails this test. “A content calendar system for solo Etsy shop owners” passes it. The audience is specific. The problem is real. The buyer can immediately self-identify.
Inkfluence AI’s analysis of top-selling digital categories for 2025 and 2026 found that niche-specific AI and automation guides are the fastest-growing ebook category, with search volume doubling year-over-year. Profession-specific topics — things like “AI tools for accountants” or “workflow automation for real estate agents” — command pricing of $27–$97 and face almost no competition compared to broad productivity topics.
To validate your idea before writing a word, run this prompt in ChatGPT:
“I’m considering writing a practical guide for [specific audience] on [specific problem]. Search Reddit and Amazon reviews to identify the three biggest complaints people have about existing guides on this topic, and what they wish those guides included.”
You’re looking for gaps. People leaving one-star Amazon reviews saying “this was too theoretical, I needed actual templates” are telling you exactly what to write. 🔍
Strong topic categories right now include:
AI and automation guides for non-technical professionals (lawyers, teachers, marketers, tradespeople)
Personal finance for specific demographics — freelancers, nurses, recent grads — not “everyone who wants to save money”
Mental health workbooks with CBT-based exercises and journaling prompts
Niche business systems — social media planning, client onboarding processes, pricing strategy for specific service types
Whatever you pick, check that your topic has visible demand. Spend 20 minutes on Amazon’s bestseller lists in your category, scan relevant subreddits, and search your topic on Etsy. If you find products already selling on the same problem, that’s validation, not competition. It confirms buyers exist.
Hour 8–40: write the thing (with AI doing the heavy lifting)
Here’s what the process looks like when you do it right. ChatGPT doesn’t write your ebook. You do. ChatGPT handles structure, first drafts, and research synthesis, and you handle everything that makes the book worth buying: your examples, your specific knowledge, your editorial judgment, your voice. 🤖
Start with the outline. Open ChatGPT and prompt:
“I’m writing a practical guide for [specific audience] on [specific topic]. Give me an 8-chapter outline where each chapter solves one distinct problem they face. Each chapter should be actionable, not conceptual.”
Review every chapter before you touch the content. Move things around. Delete what’s redundant. Add chapters the AI missed because it doesn’t know your audience as well as you do. The outline is the architecture — if it’s weak, the whole thing wobbles.
Then go chapter by chapter with a prompt like:
“Write a 500-word draft for Chapter 2: [title]. Use a direct, conversational tone. Include at least one specific example, and avoid hedging phrases like ‘it’s important to note’ or ‘one might consider.’ End with one concrete action the reader can take immediately.”
The critical move — and the one most people skip — is editing each chapter before moving to the next. Read it aloud. Add your own examples. Cut anything generic. Replace AI hedging with direct claims. As one creator noted on Medium after testing AI proposals across dozens of projects, the output that sounds like you — honest, curious, specific — is the output that converts. Generic doesn’t convert. Your perspective does.
For a 5,000-word ebook, this chapter-by-chapter process takes roughly 12–15 hours if you’re working seriously. Not because the drafting is slow, but because editing is where the quality lives. 📝
A few prompts worth keeping in your workflow:
“What are 5 specific statistics or research findings I could include in Chapter 3 on [topic]?” — forces you to ground claims in facts rather than assertions
“Rewrite this paragraph to sound like a smart friend explaining it over coffee, not a corporate FAQ.” — fixes AI stiffness fast
“What’s the most common objection someone would have to the advice in this chapter? Write a paragraph that addresses it directly.” — makes the ebook feel like it anticipates real reader resistance
Once the draft is done, do one read-through for consistency. The AI sometimes forgets what it said three chapters ago. You won’t.
Hour 40–55: design and format it properly
A professional-looking ebook isn’t optional. It’s table stakes. Buyers on Gumroad or Etsy are making a purchase decision before they read a single word — the cover and the visual presentation of the first page either build trust or destroy it. 🎨
Canva is the standard tool for good reason. It’s free, it has dozens of ebook templates, and it doesn’t require design skills. The actual process:
Search “ebook” in Canva’s template library and choose one that fits your topic’s tone (clean and professional for business, warmer for health or personal development)
Import your chapter text into the template. Don’t try to cram everything onto one page format — let the chapters breathe
Invest 20–30 minutes on the cover specifically. Change the font, adjust the colors, add a subtitle that names exactly who the book is for. The cover is your product photo. It earns or loses the click
Export as a PDF
Keep the ebook length honest. According to Inkfluence AI’s guidelines on ebook length, the sweet spot for a Gumroad product priced at $9.99–$29 is 8,000–20,000 words across 5–8 chapters. That’s a guide that solves a real problem without padding it out to justify a price tag. Readers on Gumroad aren’t buying word count. They’re buying a clear path from their problem to a solution.
Worth knowing: the BizWhat Membership covers this topic in a dedicated ebook, which is a lot more useful than a single article can be.
One formatting decision that matters more than most: the table of contents. Include it, make it specific (not “Chapter 1: Introduction” but “Chapter 1: Why your current approach to client onboarding costs you repeat business”), and hyperlink each chapter so PDF readers can navigate easily. Small detail. High impact on perceived professionalism.
Hour 55–68: set up your store and price it correctly
The platform choice is simpler than most guides make it. For a first ebook, Gumroad is still the fastest path to your first sale — no monthly fees, instant delivery, and a 10% transaction fee that you only pay when you earn. InsightRaider’s analysis of 1,049 Gumroad ebooks shows an average sale price of $50.91 on the platform — which is considerably higher than most beginners expect. ✅
Do not default to underpricing because it feels safer. The data doesn’t support it. Products under $10 account for only 0.8% of total Gumroad platform revenue despite representing about 35% of all products. The $30–49 price band converts 28% better than products priced under $10. Gumroad buyers are not browsing for bargains — they’re searching for solutions to specific problems they’re already motivated to fix.
Setting up your Gumroad product takes about 30 minutes:
Create an account at gumroad.com (free)
Click “New Product,” select “Digital product,” and upload your PDF
Write a product description that leads with the buyer’s problem, not your credentials. “Stop losing clients after one project” is more compelling than “I’ve been a consultant for 8 years”
Add a cover image — your Canva cover works perfectly here
Set your price. Start at $19–$29 for a first product and adjust based on results
Enable the email collection option. Gumroad gives you every buyer’s email. Use it
Payhip is a worthy alternative at 5% commission instead of 10%, particularly if you expect higher volume. The BizWhat article on building an online income that doesn’t stop when you stop working makes the point that ebooks are royalty-style income — they pay every time someone buys, even when you’re nowhere near your laptop. That only holds true if the product is actually visible to buyers. 💡
Get your product page URL before moving to the next step. You’ll need it.
Hour 68–72: sell it (without an existing audience)
Here’s the part that makes most people freeze: they finish the ebook, list it on Gumroad, and then wait. Nothing happens. This is predictable. Gumroad has minimal organic discovery for brand-new products. You need to drive the first traffic yourself.
The approach that actually works without an existing audience:
Reddit: Find the subreddits where your target buyer already hangs out. Spend a week or two genuinely participating in threads before you mention your product. When you do, frame it as “I wrote this because I kept answering the same question in this community.” That framing converts. Blatant promotion gets removed.
LinkedIn and X: Write one post that gives away the best insight from Chapter 1. At the end, mention the full guide is available. People who wanted more depth will click. This is content marketing at its most efficient. 📲
Facebook Groups: Find niche groups related to your topic, join them, answer questions helpfully for a few days, then mention your ebook when it’s genuinely relevant.
Your own email list: Even a tiny list converts better than any social platform because the audience already trusts you. If you have 50 people on a list, email them. Offer a launch discount for the first 48 hours.
The realistic first-week benchmark for a brand-new creator with no existing audience: 2–5 sales. That might feel underwhelming if you were hoping for 50. Here’s why it matters anyway: you now have a product, a sales page, a distribution system, and actual buyers who you can email for feedback. The second ebook takes half the time to create and launches to an audience instead of a cold market. As BizWhat’s ebook economics piece points out, “a mediocre ebook on a great topic will outsell a brilliant ebook on a bad topic.” Your first job is proof of concept, not perfection. 🚀
What’s the most specific problem in your own professional world that you keep explaining to other people — the thing you wish someone had written a clear guide about when you were starting out?