How to do book market research before you write

Introduction

Most authors treat market research as something to do after they've decided what to write. They pick a topic, draft a premise, open a blank document — and only after the manuscript is done wonder whether the idea fits the market at all. Properly understood, learning how to do book market research is an exercise you run before chapter one. The cost of skipping it is real: a polished book that nobody outside the author's circle was asking for.

This post walks through what book market research actually is for a book, how to do it cheaply in a single week, the signals that matter, the shortcuts that look like research but aren't, and the thresholds that say "ready." Throughout we'll link the AuthorPilot sample BookScore™ report so you can see what scored research looks like — and the validator itself when you're ready to score your own premise.

What "market research" actually means for a book

A book is not a SaaS product. The market research a product team runs — TAM, ICP, willingness-to-pay surveys — doesn't translate directly: authors who copy it over tend to over-engineer and under-answer. For a book, research is narrower: it's the empirical answer to three questions about the premise.

Who, specifically, is the reader. Not "anyone interested in productivity" — that's a definition with no shape. A real reader has a recurring problem they're already spending time, money, or attention trying to solve. If you can't name that person in a sentence — say, "first-time managers in their thirties who took the promotion to escape IC work and now feel under-qualified" — the research hasn't started.

What books sit next to yours, and where the search demand lives. The honest comp set, plus the search terms your reader is already typing into Amazon, Google, and podcast directories. A book that solves an undiscovered problem has no launch. One that solves a heavily-searched problem has a built-in demand curve.

The cheap-week research workflow

The method below is what our team at AuthorPilot uses when an author asks how to do book market research before committing six months to a manuscript. Five working days, less than a single weekend of writing time, defensible answers to the three questions above.

Day 1 — map the reader. In one sentence, describe the person you're writing for, the recurring problem they have, and what they're already doing to solve it. Then list the three books, courses, podcasts, or paid newsletters that person has paid for in the last year. If you can't fill that sentence, the premise is too vague. Tighten until you can.

Day 2 — category browse. Open the Kindle store and the bestseller list in your category. Read every top-20 title across the closest three sub-categories. For each, write a single sentence on what differentiates it from yours. Build a table: title, differentiation lever, table-of-contents strength. If the table has fewer than ten rows, the category doesn't yet have enough activity — that's a signal, not a blocker.

Day 3 — paid newsletter test. Write a 600-word essay arguing the premise. Pay two or three newsletters your reader already reads to send it with one call to action: "If this becomes a book, I'd like to know at launch." Click-through rate on a cold send to a 5,000+ list is the cleanest demand signal you'll see all year. Above 4%, the premise travels; below 2%, the reader isn't seeing themselves in it yet.

Day 4–5 — score the premise. Before chapter one, score the premise against the four axes the market is asking about — market fit, originality, audience fit, genre competition. The fastest path is to look at a sample BookScore™ report to see what scored research looks like across the same axes, then run the validation on your premise. The score is a check on your gut, not a replacement — but at the end of a research week you want a number attached to "this idea is ready," not a feeling.

Shortcuts that feel like research but aren't

A category of pre-launch activity masquerades as book market research: engagement, but not signal — inflating confidence in ideas the market has already rejected.

Amazon-rank stalking. Refreshing the bestseller rank of a comp title eight times a day. Rank is a noisy signal — driven by short bursts of promotion, page-read velocity, and seasonality. Useful rank research: rank over time, rank across categories, rank six months ago vs. now. Useless: today's number.

Pre-order list building. Cold pre-order lists typically convert to launch-day sale at under 4%, often under 2%. A 200-person list built around a specific title is signal; a 2,000-person general-interest list is not.

The unifying pattern: these shortcuts measure engagement with the research activity, not with the eventual book. Real book market research measures engagement with the premise — paid newsletter sends, survey replies, purchases, actions that cost the reader something.

When to write vs. when to iterate

Research produces a number, and the number is only useful paired with thresholds. Here are the cut-offs we use at AuthorPilot for the four-axis BookScore™ a validation produces.

Write when the score is above 70 across all four axes. Market fit above 75 means the topic maps cleanly onto a recurring problem. Originality above 70 means the book is differentiated from its comps on a single defensible lever. Audience fit above 70 means the audience is sized correctly for the price point. Genre competition above 70 means the category is alive — new entrants in the last six months — rather than saturated.

Iterate when one axis sits below 60, especially originality or audience fit. Low originality with strong market fit means the book is wanted but already over-served. Low audience fit with strong originality means it's fresh but has no clear buyer, and the launch will be slow.

Walk away when two or more axes sit below 50. Forcing it into a manuscript will cost the year you could have spent on the right book. Walking away is the entire point of book market research: cheap information in real time, before the expensive thing.

If you're on the boundary, the cheapest tie-breaker is to re-run the BookScore™ on the strongest pivot version of the premise. A pivot to a tighter niche or a sharper angle often pushes two of the four axes from red into green.

Closing: a week to save a year

Knowing how to do book market research before you write is the highest-leverage skill a self-publishing author can build — and the most-skipped, because research never feels productive the way writing does. The gap between "I have a strong idea" and "I have evidence the idea is strong" is the gap most authors would rather not look across.

The research week costs a fraction of the unvalidated manuscript, and the validated idea, nine times out of ten, beats the unvalidated one on launch-day conversion by enough to compound across a whole backlist. If you're past the research week and ready for the deeper pass — comp-by-comp positioning analysis, audience-fit interview loops, multi-angle pivot matrices — the AuthorPilot premium research workflow walks through that next layer.