Pick a niche and test demand

Most failed books fail at the first decision: nobody was looking for them. This lesson helps you choose with evidence, not hope. It will not promise sales: no method can.

What you will build

A shortlist of three tested niches and one chosen niche, with the evidence written down.

Figure 1. Narrow in stages, and test the last three against real listings.

Step 1: List 30 ideas

Use three lenses and write ten per lens:

1. People you understand (parents, teachers, runners, gardeners, a hobby you know).
2. Occasions (birthdays, holidays, back to school, a season).
3. Formats (coloring pages, planners, logbooks, activity books, story books).

Ask Claude to widen the list, not to choose:

I will give you a topic area. List 30 specific book ideas in the form
"format + reader + situation", for example "weekly meal planner for new parents".
Do not rank them and do not claim demand. Topic area: [yours]

Step 2: Cut to ten you can serve

Remove ideas you cannot make well, ideas that need credentials you do not have, and anything where you cannot honestly say who it is for.

Step 3: Test three against real listings

For each candidate, search on Amazon the way a buyer would and write down what you see:

Check	Evidence to record
Are there books like this at all?	Count of similar titles on the first page
Do they look professionally made?	Cover and preview quality, 1 to 5
Do any have many recent reviews?	Rough review counts and dates
What do buyers complain about?	Three themes from low-star reviews
Could you make a clearly better one?	One specific improvement

A niche with no similar books may have no demand; a niche with hundreds of identical books needs a clear difference. See our guide to keyword research for KDP for finding the phrases buyers use.

Step 4: Check the numbers honestly

Estimate what you would earn per sale using KDP’s royalty rules and your print cost, then ask whether the effort makes sense. Our royalties and pricing explainer walks through the calculation.

Step 5: Choose one and write the promise

This book helps [reader] to [outcome] by [format and approach].
It is different from existing books because [specific improvement].

Worked example

A teacher considered “coloring book for kids”. Too broad. Narrowing by occasion and reader produced “first-day-of-school coloring pages for 5 to 7 year olds, with a sticker-style name tag page”. Listings showed few books with that exact angle, and the low-star reviews of generic books complained about thin paper-sized pages and no teacher use. That became the difference.

Common mistakes

• Choosing because a video said a niche is “hot”.
• Skipping the buyer-complaint research.
• Picking a niche you will be bored of by page 20.

Checklist

• 30 ideas listed, cut to 10, tested 3
• Evidence table filled for each tested niche
• One-line promise written