How to Validate an Online Course Idea Before You Create It

Validate an online course idea by proving people will pay before you record a single lesson. Work through five steps: research whether the topic already has demand, interview your audience about their problem, test a course outline against a real outcome, run a presale to a small cohort, and decide against a go/no-go number you set in advance. Payment is the only signal that counts.

Quick Answer: Enthusiasm is not demand. Presell the course to a real audience, and only build it once enough people have paid to clear a threshold you picked before you launched.

The most expensive way to find out nobody wants your course is to spend three months recording it first. Yet that is the default path for most creators: pick a topic you love, batch-record forty lessons, launch to silence, and conclude that "the algorithm" failed you. It didn't. You skipped validation.

This guide walks the sequence that inverts that risk. You will collect evidence in ascending order of cost and honesty — cheap signals first, a real payment last — so that by the time you press record, the course is already partly sold. Think of it as a coach's playbook: do the reps in order, and let the results, not your optimism, make the build decision.

Why polls and likes mislead course creators

Polls and likes mislead course creators because they measure approval, not intent to pay, and those are two entirely different behaviors. A like costs nothing. A poll vote costs nothing. Saying "I'd totally buy that" in your comments costs nothing and earns the commenter a small hit of goodwill for being supportive. None of it moves money, so none of it predicts revenue.

This is the gap between stated preference and revealed preference. What people say they will do and what they actually do diverge most sharply exactly when the stated action is free and socially rewarded. Your audience isn't lying to you. They are being agreeable, and agreeableness is not a market.

Pat Flynn's Will It Fly? frames the whole discipline around this trap: real validation means putting your idea in front of the market in a way that demands a real response, not a friendly nod. A show of hands in a webinar is a friendly nod. A checkout page is a real response.

There is a second, subtler failure. The people most likely to like and comment are the people already closest to you — your warmest followers, who would applaud almost anything you announced. They are the least representative sample of the broader market you actually need to reach. High engagement from a small, loyal core can look identical to demand right up until launch day proves it wasn't.

The rest of this guide replaces cheap signals with progressively costlier ones, ending at the only test that has ever been decisive: someone hands you money for a thing that does not exist yet.

Step 1: Research whether your course topic already has demand

Start by checking whether people are already trying to solve this problem and already paying to do it — before you talk to a single person. Existing demand leaves fingerprints, and your job in this step is to find them.

Look for the places money and attention already flow toward your topic:

• Existing paid products — other courses, books, coaching, templates, or paid communities on the same problem. Their existence is the single strongest desk-research signal.
• Active search behavior — people typing questions into search engines and forums, which tells you they are pulling for answers rather than waiting to be pushed one.
• Self-assembled solutions — folks stitching together free YouTube videos, Reddit threads, and blog posts to solve the problem the hard way. A messy DIY workaround is a market asking to be served properly.
• Recurring, specific complaints — the same frustration voiced over and over in communities, in the sufferer's own words.

Competition is confirmation, not a warning sign. New creators often panic when they find existing courses on their topic and conclude the space is "saturated." Reverse that instinct. A total absence of competition usually means no market, not an untapped one. What you want is a proven, paying audience that existing options serve incompletely, leaving room for your angle, your audience, or your depth.

The output of this step is a written verdict: does observable demand exist, and where is it underserved? If you can't find fingerprints anywhere, that is itself a result — treat it as a red flag before you invest interview time. For a deeper method here, see the companion guide on researching whether your course topic has demand, which breaks down the specific signals to weigh and how to weight them.

Step 2: Interview your audience about the problem, not the course

Interview to understand the person and their problem — never to pitch your course. The moment an interview becomes a sales presentation, the other person switches into "be nice to the creator" mode, and every answer after that is contaminated by politeness.

Arvid Kahl's The Embedded Entrepreneur makes audience-first the whole foundation: you build for a community you understand deeply, and that understanding comes from listening before building. Interviews are how you earn it. Talk to eight or ten people who plausibly have the problem, and keep the conversation anchored to their world, not your idea.

Ask about the past, because past behavior predicts future behavior far better than hypotheticals do:

• What have you already tried to solve this? Effort spent is proof the problem matters.
• What did you pay for — money, time, or tools? Money already spent is the clearest demand signal a conversation can surface.
• Where did those attempts fall short? The gap is your opening.
• When did you last run into this, and what happened? Specific recent stories beat general opinions every time.

Notice what is missing from that list: "Would you buy my course?" Never ask it. It invites a free, agreeable yes that means nothing, and it tips the interview into a pitch. Your goal is a vivid picture of the problem, the person's real language for it, and evidence of what they have already been willing to spend. If nobody in your interviews has ever paid to relieve this pain, be honest with yourself about what that implies.

Step 3: Test your course outline against a real outcome

Turn the problem you heard into a transformation, then test whether your outline actually delivers that transformation — not just whether it covers the topic. People do not buy courses to consume lessons. They buy a specific change: from stuck to shipping, from confused to confident, from where they are to where they want to be.

Write the outline as a path between two points. State the starting state ("you have never written a line of code"), the ending state ("you have deployed your first working app"), and the ordered steps that get someone from one to the other. If a proposed module doesn't move the learner along that path, it is content you love, not curriculum they need — cut it.

Then put the outline back in front of a few of your interviewees. The test is not "does this look good?" — it is "where does this miss?" Ask them which step they are most afraid of, which part they would skip, and what is missing that would stop them from succeeding. You are looking for the outline to survive contact with the people who will pay for it.

This step is also where scope discipline pays off. A tight outline built around one clear outcome is far easier to presell than a sprawling "everything about X" syllabus, because a specific promise is easier to believe and easier to price. Resist the urge to make it comprehensive. Make it decisive.

Step 4: Run a presale offer before you record anything

Run a presale, because it is the only validation test that uses real money, and money is the only signal that has never lied. Everything before this step was preparation. This is the exam. You offer the course for sale — with a real checkout, a real price, and a real delivery date — before it exists.

A presale works because it collapses the gap between saying and doing. There is no politeness discount at a payment page. Someone either believes in the outcome enough to pay for it, or they don't, and either answer is worth more than a hundred enthusiastic comments. The full mechanics — founding-cohort framing, pricing, delivery dates, and refund handling — are worth studying in the dedicated walkthrough on how to presell an online course; the essentials are below.

Structure the offer so a yes is meaningful and a no is clean:

Introduced honestly, this is not a trick — you are letting early buyers fund and shape a course they want, and protecting them fully if it doesn't happen. The takeaway from the table: a presale is only a valid test if a yes costs the buyer something real and a no leaves them no worse off. Anything softer measures interest, not demand.

Be honest about delivery on your side, too. Selling something you fully intend to build, with a clear refund path if you don't, is straightforward. Collecting money with no intention to deliver is not validation — it is a problem, and consumer-protection rules around it vary by jurisdiction, so if you have any doubt about how presales or refunds work where you operate, verify the local rules before you launch.

Step 5: Set a go/no-go threshold and decide

Decide your go/no-go number before you launch the presale, so the result makes the decision instead of your feelings. This is the single most important discipline in the whole sequence, because without a pre-committed threshold you will rationalize any outcome. Three sales feels encouraging when you are attached to the idea. It shouldn't be encouraging if you needed twenty.

Set two numbers in advance:

• A minimum number of paying customers that means "build it." Choose a number that reflects real demand for your audience size and price — enough that the revenue and the signal both justify the work ahead.
• A time box — the window the presale runs. Open-ended presales let you keep waiting for one more sale to save a failing idea.

When the window closes, you are in one of three lanes, and each has a clear next move:

1. Cleared the threshold — build it, on the date you promised, for the people who paid.
2. Missed it badly — refund everyone promptly, thank them, and treat it as a cheap, early save. You just avoided months of wasted production for the price of a few conversations.
3. Landed just short — go back to the buyers and the near-misses and ask what would have made it a yes. Often the offer, price, or promise needs one turn, not the whole idea abandoned.

A no is a gift, not a failure. A presale that misses cost you a launch page and a week; a finished course that misses cost you a quarter of your life. This step is where course validation connects to the wider practice of testing any business idea before building — the same pre-set-threshold logic runs through the complete guide to startup idea validation, and it is worth internalizing beyond courses.

Validation signals ranked from weakest to strongest

Not all "yes" signals carry the same weight, and treating them as if they do is how creators fool themselves. The ladder below ranks the responses you might collect by how much they actually cost the person giving them — and cost is the best proxy for honesty. Read it from top to bottom as increasing evidence.

The takeaway is blunt: everything above the last row is a leading indicator at best and a comforting illusion at worst. Only the bottom rung — a completed payment for a course that doesn't exist — reliably predicts that people will buy the finished thing. Collect the cheaper signals to guide your offer, but never mistake them for the verdict.

Common course validation mistakes

The most common validation mistakes all share one root: substituting a cheap signal for the expensive one and calling it proof. Spot yourself in this list before it costs you a production cycle.

• Building first, selling second. The default trap. You record everything, then look for buyers, having already spent the resource you were trying to protect.
• Validating with friends and family. Their job is to love you, which makes them structurally incapable of giving you a market signal. Their yes is real affection and worthless data.
• Asking "would you buy?" instead of selling. A hypothetical purchase is free to promise. Replace the question with a checkout page and watch the answer change.
• Pricing the presale near zero. A one-dollar or free presale tests curiosity, not demand. If a real price scares people off, that is the finding, not a reason to discount reality away.
• Ignoring the no-shows. Ten "definitely" replies and zero payments is not a mixed result. It is a clear no, and the payments are the only half that counts.
• Moving the goalposts. Deciding the threshold after seeing the result. If you set twenty and got six, six is a miss — no matter how much you want it to be a start.
• Over-producing the presale itself. You do not need a polished sales funnel to test demand. A clear promise, an honest offer, and a working payment link are enough. Save the production budget for the course.

One more, specific to certain niches: be careful with outcome and credential claims. If your course promises professional results, financial gains, health improvements, or a certification, the language you can legally use around those promises may be regulated, and the rules vary by jurisdiction. This is US-centric guidance and not legal advice — if your topic touches a licensed field, verify what you can claim locally before you sell.

Key Takeaways

• Enthusiasm is not demand. Likes, poll votes, and supportive comments measure approval, cost nothing, and reliably predict nothing about who will actually pay.
• Validate in ascending order of cost. Move from desk research to interviews to a real presale, so cheap signals shape the offer and one expensive signal makes the decision.
• Competition proves the market exists. Finding existing paid courses on your topic is confirmation you have a real audience, not evidence the space is too crowded to enter.
• Interview about the problem, never the pitch. Anchor conversations to what people have already tried and paid for; the question "would you buy my course?" produces only agreeable, meaningless yeses.
• A presale is the only decisive test. A completed payment for a course that does not yet exist is the single signal that has never lied, because it is the only one that costs the buyer something real.
• Set the go/no-go number before you launch. A threshold chosen in advance lets the result decide; a threshold chosen after the fact lets your optimism override the market.
• A no is the cheapest win available. A failed presale costs a launch page and a week — a failed finished course costs a quarter of your life you can't refund.

Frequently Asked Questions

How many presale sales mean my course idea is validated?

There is no universal number — validation is whether you cleared the threshold you set in advance, given your audience size and price. Set that number before launching, based on the revenue and demand signal that justify the build for you. Ten sales might validate a premium cohort course and fall far short for a low-priced one. The discipline is the pre-commitment, not a magic figure.

Can I validate a course idea without an existing audience?

Yes, but the early steps carry more weight and take longer. Without an audience, lean harder on topic-demand research and one-on-one interviews to confirm the problem is real and painful. Then borrow reach for the presale — partner with someone who has the audience, post in relevant communities where allowed, or run a small ad to a landing page. You need buyers, not necessarily your own followers.

Should I create the course before or after I presell it?

After. The entire point of validation is to avoid producing a course nobody wants, so building first defeats it. Presell against a specific future delivery date, then record once you have cleared your threshold — often with your founding buyers helping shape the content. Have your outline ready and be genuinely prepared to build, but do not record the lessons until paying customers have told you it is worth building.

What should I do if people say they'll buy but don't pay?

Treat the gap as your real answer and trust the payments, not the promises. A wall of "I'd definitely buy this" followed by zero purchases is a clear no, not a mixed signal — the stated interest was free and the payment wasn't. Go back and ask the non-buyers what stopped them: price, timing, trust, or the promise itself. Their reasons tell you whether to adjust the offer or drop the idea.

Is it okay to sell an online course before it's made?

Yes, when you do it honestly: sell something you fully intend to build, commit to a clear delivery date, and offer a full refund if you don't deliver. That is a legitimate presale, and framing buyers as founding members who shape the course makes it a fair exchange. Collecting money with no intent to deliver is a different thing entirely, and consumer-protection rules around presales and refunds vary by jurisdiction, so verify the rules where you operate.