How do you know if your ebook idea will sell?

Check 1: strangers are already searching

Type your topic into Etsy, Amazon, and Google and look for autocomplete suggestions and existing products. Strangers searching in their own words is demand you did not manufacture — the strongest signal available.

What you want to find: specific phrasing (“meal prep for night shift nurses”) with real products attached. What should worry you: total silence, or only giant broad books with nothing in your specific corner.

Check 2: they have paid for adjacent things

Look for money already changing hands near your topic: competing ebooks with reviews, courses with students, templates with sales, coaches selling the outcome. Reviews are especially valuable — three-star reviews on competing products are feature requests for yours.

Read those mid-tier reviews carefully and note every “I wish it covered…” That's your table of contents, dictated by buyers.

Check 3: people ask about it unprompted

Search Reddit, Facebook groups, and niche forums for questions matching your topic. Unprompted questions — asked to peers, not to you — prove the problem is alive enough that strangers spend social capital on it.

Collect ten real questions in the askers' own words. If you cannot find ten, the topic is either too narrow or too early. Both are survivable, but only with eyes open.

Check 4: a landing page pulls emails

Write the title, the one-sentence promise, and three chapter teasers on a single page, and ask for emails to hear when it launches. Fifty signups from relevant traffic is a pass; crickets from the right crowd is a fail worth believing.

Publish an edition explicitly when you are ready to share it. An account is required, and the current hosting limits apply; creating a draft does not make it public.

• Pass: signups from people who match your reader sentence.
• Weak: traffic but no signups — fix the promise, not the book.
• Fail: no traffic at all — fix distribution before writing.

Validate with three kinds of evidence

Search interest, direct conversations, and a small use test answer different questions. Search tools can show whether people use phrases near your topic, but a trend line is not a count of buyers. Customer conversations reveal where the problem appears in real work, but friends may be encouraging rather than representative. A prototype shows whether your proposed material helps someone act, but one test does not prove a market. Use all three as imperfect signals, then choose a modest first release rather than waiting for certainty no method can provide.

Start with the reader’s exact problem sentence. Search it in a marketplace, a search engine, and relevant communities. Note the language buyers use and what existing products promise. Read low-star reviews as carefully as high-star ones: complaints about missing templates, unclear steps, stale screenshots, or too much theory point to ways a better book could help. Do not copy titles or treat review counts as your future sales forecast. The goal is to understand the job and the expectations around it.

Ask five potential readers to describe the last time they faced the problem. Prompt for what they tried, what information they needed, and what remained unresolved. Avoid asking whether they would buy your proposed ebook; hypothetical praise is cheap. Their description of past behavior is more useful. If nobody has encountered the problem recently, the topic may be too speculative or the audience definition too broad. If the same obstacle appears repeatedly, use their language to shape the first chapter and the product page.

Build a ten-page prototype with one complete result: a checklist, example, calculation, or decision tree. Give it to readers with a task and watch where they hesitate. Ask them to explain the next action in their own words after using it. If they cannot, the book needs a clearer sequence or a different format. You are testing usefulness before production polish. A beautiful cover cannot rescue a method that nobody can apply.

Set a go, revise, or stop decision in advance

Write your criteria before seeing the first test result. For example: proceed if several independent readers describe the same recent problem and at least two can finish the prototype exercise unaided; revise if the problem is real but the material confuses them; stop if the audience cannot name the situation in which they would use the book. These are illustrative thresholds for a small qualitative test, not industry benchmarks. Choose criteria that fit your time and risk, then keep them visible while you listen.

If the idea passes, write the smallest complete edition. Include the core method, a worked example, a checklist, and the sources a skeptical reader would need. Do not broaden into every adjacent topic just because search results show them. A focused book is easier to finish and easier to describe. Publish where you already have a credible route to readers; a listing on a platform is not a traffic plan by itself.

If the idea needs revision, change the promise or the format based on observed friction. Readers may want an editable spreadsheet instead of a prose chapter, a short field guide instead of a course, or a case study before a template. Keep the evidence you gathered and test one revised prototype. Repeating the same question to more people without changing the asset may only produce more polite uncertainty.

If the idea fails, record why. The topic may have weak urgency, strong free alternatives, no reachable audience, or a problem too broad for an ebook. Those are useful findings, not wasted work. The interviews and draft can become a narrower article, a lead magnet for a different service, or the starting point for a better book. Validation protects your time by making the next decision clearer, not by guaranteeing sales.

Avoid the evidence traps in early validation

Search volume is an estimate about queries, not a purchase order. A topic may be searched for homework, entertainment, free advice, or a buying decision. Look at the actual results and ask what the searcher seems to want at that moment. Google Trends can show relative interest over time and place, but its normalized scale is not the number of ebook buyers. Use it to notice language and seasonality, then validate the buying or use case separately with readers.

Marketplace listings show competition and conventions, but they do not reveal the full economics. A bestseller badge, review count, or polished cover does not tell you profit, ad spend, refund rate, or production cost. Record what buyers praise and complain about instead. If multiple reviews ask for editable templates, that is a product clue; if they complain about stale information, that is a maintenance warning. Do not conclude that your book will sell because another listing has a high rank today.

Interviews can mislead when you ask leading questions. “Would you buy my ebook?” encourages kindness and imagination. Ask what the person tried last week, what failed, what they paid for, and what they would do if this problem remained unresolved. Listen for concrete workarounds. A problem that already consumes time or money is more urgent than one people agree sounds interesting. Keep the interview short and record exact phrases, but do not turn one quote into a market size estimate.

A prototype test can also create false confidence if the participants are too close to you. Friends may understand your shorthand and excuse missing instructions. Try a reader who did not see the draft evolve. Give them the same task and file a buyer would receive, then step back. Watch which page they open, where they hesitate, and whether they can state the outcome. Their behavior may challenge a beloved chapter; that is the value of testing before layout and launch make changes expensive.

Combine signals cautiously. If search interest exists but readers cannot use the prototype, improve the product. If the prototype works but you cannot reach the audience, improve distribution before writing a longer edition. If interviews reveal a different problem from the one your title promises, revise the promise and test again. Validation is a sequence of decisions under uncertainty, not a stamp that guarantees traffic. The useful result is a better next action with less unsupported confidence.