TL;DR
In 90 minutes, run this 5-step test to validate a book concept: (1) write the one-sentence promise, (2) mine 30 real audience queries, (3) sketch the table of contents, (4) send the promise + TOC to 5 ICP buyers and read their responses, (5) score against the buyer-promise-position rubric. If it clears 24 out of 30 points, write the book. Under 20 needs a full rework. This filter kills generic ideas before you burn a year on them, and sharpens surviving ideas into something people actually want.
The most expensive book mistake isn't the wrong path or the wrong publisher. It's the wrong idea. A founder who writes a solid manuscript on a concept nobody wanted has spent 12-18 months and S$8,000-$80,000 producing something that will underperform every metric.
The 90-minute test below runs before you commit. It's cheap enough that you have no excuse to skip it. And it's specific enough that failure at any step tells you exactly what to fix instead of leaving you guessing.
The 90-minute frame
Five steps, roughly 90 minutes total. Each has a specific output. At the end, you either have a validated concept you can start writing, or a clear diagnosis of which part of the concept needs to be reworked.
Step 1: Write the one-sentence promise
The formula: "This book teaches [specific buyer] how to [specific outcome] without [common failure mode]."
Bad: "This book teaches founders how to grow their business." (Buyer too broad, outcome too vague, no failure mode.)
Good: "This book teaches solo consultants how to reach S$400K/year in revenue without hiring a team or building a course." (Specific buyer, specific outcome, specific failure mode.)
Write yours. Read it aloud. If it feels weak, tighten. If it feels generic, add specificity to the buyer or outcome. The one-sentence promise is the anchor for everything else in the test.
Step 2: Mine 30 real audience queries
Go to where your buyers actually ask questions. Collect 30 verbatim questions in their words. Sources that consistently work:
- Reddit search. Search your topic in relevant subreddits. Sort by top for the past year. Read the questions people ask. Copy the ones that match your ICP.
- LinkedIn poll comments. Find polls on your topic in your niche. The comment section is a live focus group.
- Podcast Q&A segments. Search transcripts of the top 5 podcasts your buyers listen to. Note questions guests get asked repeatedly.
- Google "People Also Ask" boxes. Search your primary keyword. The PAA questions are ranked signals of common queries.
- ChatGPT/Perplexity shared conversations. Publicly-shared AI chats on your topic reveal what people are actually asking.
Copy 30 questions into a document. Don't rewrite them. The exact phrasing matters.
Step 3: Sketch the table of contents
Take the 30 questions. Cluster them into 8-12 groups. Each group becomes a chapter. Give each chapter a working title that answers the cluster's core question.
If a cluster has more than 5 questions, it's a strong chapter (probably needs subsections). If a cluster has only 1-2 questions, it might belong inside a larger chapter or get cut. If you have questions that don't cluster into any chapter, they're either signals you need another chapter, or signals the questions aren't book material.
This TOC is your book skeleton. Not the final version, but a working structure that reflects what your audience actually asks.
Step 4: 5-person ICP gut check
Take the one-sentence promise (Step 1) and the TOC (Step 3). Send them to five people who are actual buyers or influencers in your buying committee. Ask three questions:
- "Would you buy this book based on the promise alone?" (Yes/No)
- "Which chapter would you read first?"
- "What's missing that you'd expect to see?"
Send by email or DM. Don't do it in a group setting; the responses will be biased. Give them 24-48 hours to reply. Their answers tell you three things: whether the promise is strong (Q1), which chapters carry the highest interest (Q2), and what's missing from your outline (Q3). Q3 is the most valuable. What they say is missing often becomes the book's most-cited chapter.
Step 5: Score against the rubric
Score your concept 0-10 on three axes:
| Axis | Question | Score 0-10 |
|---|---|---|
| Buyer clarity | Can you name the exact reader in one specific sentence, not "entrepreneurs" or "leaders"? | ___ |
| Promise sharpness | Is the outcome specific and testable? Would your 5 ICP contacts describe it back accurately? | ___ |
| Market position | Is there an existing book you're clearly different from? Can you name the closest 2-3 comp titles and explain what you'd do differently? | ___ |
Total possible: 30. Pass: 24+. Rework: 20-23. Kill and pick a new concept: under 20.
What to do with the result
The test spits out one of three outcomes. Each has a clear next step.
Pass (24-30 points)
The concept is book-ready. Move to the 30-day reader research sprint for deeper validation, then to the one-page book proposal. Then start drafting.
Rework (20-23 points)
Identify which axis scored lowest and target it. If buyer clarity is the low score, tighten who the book is for (usually by narrowing an "entrepreneurs" to a specific vertical, stage, or geography). If promise sharpness is low, add specificity to the outcome or failure mode. If market position is low, study the 2-3 closest comp titles and figure out what genuinely differentiates yours. Run the test again in 2-3 days after the rework.
Kill (under 20 points)
The concept isn't ready and can't be fixed with tweaks. Pick a different concept, run the test on that one instead. This is the outcome that saves you the most time. A killed weak concept costs you 90 minutes. A written weak book costs you a year.
Common failure modes and how to fix them
Three patterns show up in almost every failed idea test.
Buyer too broad. "Entrepreneurs," "founders," "leaders," "small business owners." These aren't buyers, they're categories. Narrow to a specific ICP: "solo consultants in Southeast Asia scaling from S$150K to S$400K in revenue" is a buyer. "Entrepreneurs" is not.
Promise too vague. "Grow your business," "build authority," "get more clients." These aren't outcomes, they're aspirations. The promise needs a specific number or milestone: reach S$400K/year, book 10 keynotes in year one, close 3 corporate coaching contracts per quarter.
No differentiated position. "A guide to X" is not a position when 15 other books are already called some version of "The X Guide." Position needs a specific angle: contrarian, methodology-based, targeted to a niche other books ignore, or built on original data. If you can't name the angle in one sentence, the market position isn't clear enough.
Frequently asked questions
How niche is too niche for a business book?
There's no floor. Books that target 5,000 specific buyers sell as well as books that target 5 million, because the narrow book converts a higher percentage of readers into paying clients. A book targeted at "B2B SaaS founders scaling from S$1M to S$10M ARR in Southeast Asia" will outperform a book targeted at "entrepreneurs" for the founder who serves that specific niche. Niche isn't the enemy. Generic is.
What if my audience isn't on Reddit?
Reddit is one mining source among many. Substitute LinkedIn (comments on niche-community posts, LinkedIn Groups), industry Slack communities, podcast episode Q&A segments, Twitter/X threads in your category, or comments on top-ranking blog posts. The point is to gather 100 real questions your buyers ask, in their own words. Where they hang out matters less than that you go where they hang out.
Should I do more validation than 5 people?
5 is the minimum. 10-15 is better if you have time. Anything below 5 is anecdote. The value comes from finding the pattern: what do all 5 (or 10 or 15) ask you back about? Which chapters do they want first? Which parts do they say "I already know that"? That pattern is more reliable than any individual person's opinion.
How do I know if the promise is strong enough?
Two tests. First, could someone repeat it back to you accurately after hearing it once? If yes, it's crisp enough. Second, would someone in your ICP forward the promise to a peer with "You should read this"? If yes, it's compelling enough. A promise that fails either test needs sharpening before the book gets written.
Ninety minutes now, or a year later
The 90-minute test is the cheapest, highest-leverage step in the entire book process. Everything downstream of a validated idea is easier; everything downstream of an unvalidated one is harder. Take the test before you take the year.
Passed the test? Move to the 30-day reader research sprint for the deeper validation, then to the one-page book proposal. Still deciding whether the book is worth it at all? Return to the cost-benefit reality check.
You're the expert. Now be the author.
Every article on this site exists to help founders and consultants turn hard-won expertise into the credential that unlocks keynote fees, board seats, and premium retainers. Ready to move?