Value

How to validate a business idea before you build anything (the rule that would have saved me months)

Strangers pre-commit time, email or money before you build a thing. Here is how to collect that proof in a week, what counts, what does not, and what to do when the answer is no.

Youness ES-SEBIY Published August 2026 11 min read

Most people who search for how to validate a business idea are really asking for permission. They have a course half outlined, a document with a product name at the top, and a quiet fear that three months of building will land on silence. The fear is correct. The fix is not more confidence. It is a rule I wish someone had handed me in 2019: do not build until someone has, in some form, paid.

Paid does not only mean money. It means a stranger gave up something real (a calendar slot, an email address, a deposit) for a promise you have not built yet. That is the whole method. The rest is the how.

I run an online education company that has coached more than 2,000 students since 2019, built from a laptop with no office, no investors and no debt. I got the Value variable mostly right because I had been my own customer. I still would have validated earlier.

The three months I would take back

The pattern is so common that I can describe it before the person finishes their message. Someone with a real skill decides to make a course. A productivity course, say. They outline twelve modules, buy a microphone, record on weekends for three months, and launch to forty people, most of them relatives. Then silence. Conclusion: online business does not work.

It worked fine. The equation did exactly what it does. Revenue is Value times Storefront times Marketing times Sales times Delivery, and in a multiplication one zero makes everything zero. Nobody was actively trying to buy what they built, so the zero sat in variable one while they polished the other four on top of it.

My own version was milder. When I started recording courses as a student, I knew exactly where students got stuck, because it had been my problem two years earlier. But I still recorded for months before a stranger had a chance to tell me yes or no. Being your own customer gives you a head start on the person and the problem. It does not give you proof.

What does it mean to validate a business idea?

To validate a business idea means getting strangers, not friends, to pre-commit something real before you build: their time on a call, their email on a waiting list, or their money on a pilot. Opinions, likes and encouragement are not validation. A commitment that cost the person something is. You build only after that commitment exists.

The word pre-commit carries the whole idea. A commitment is something the person would have to undo: cancel the call, ignore your emails, ask for the deposit back. A survey answer costs nothing to give and nothing to abandon, which is why surveys lie so politely. Validation is a transaction in miniature, run before the product exists.

Person, problem, proof, then product

Value is not what you make. Value is a problem, solved, for a person who wanted it solved. The working order is person, problem, proof, then product, and proof is meaningless if you collect it for the wrong person.

Person first. Pick someone you can describe from the inside: the version of you before you learned your skill, the colleagues who ask you for help, the community whose complaints you can recite. Describe the person at that level of detail and your storefront writes itself.

Problem second. The problems that sustain a business usually cost the person money (or block them from earning it), time, health, or a key relationship. The stronger the pain, the less convincing you will need to do later.

Proof third: the rest of this article. Product fourth: by then the market has made half your decisions, because your headline is a sentence someone said on a call. This ordering is the part the Value mind map does for you on one page. Once proof exists, the build is a thirty-day job, which I cover in how to turn your expertise into a digital product.

What counts as proof, and what does not

Payment takes many shapes. Here is what counts, strongest first.

• Money. A pilot client at a modest price, a pre-order, a deposit on a service you have described but not yet delivered.
• Calendar time. Fifteen strangers, not friends, who accept a call about the problem. Fifteen minutes of a stranger's week says the problem is real for them.
• Email. Two hundred sign-ups to a waiting list from one honest post is a strong signal. Far fewer can be enough if the ratio is right.
• Repetition. The same fears, in the same words, across several calls. That phrase is the headline of your storefront.

What does not count, however good it feels: likes, comments, encouragement from family, a survey where nobody had to give anything, and your own conviction at 2 a.m. The line between the two lists is cost. Friends saying "great idea!" is social kindness. Strangers giving emails, calendar slots or deposits is validation. Only count the second kind.

The cheapest validation instrument ever invented

A sales page for a product that does not exist yet, with a waiting list where the buy button will be. I call it the promise page. It takes one afternoon and costs nothing. If you cannot get strangers to leave an email for the promise, you will not get them to leave money for the product. Better to learn that in week one than in month four.

The page has four parts. The headline is your one-sentence, rewritten in the person's words. Under it, three to five lines describing the smallest version that would fully deliver the bridge. Then the price you intend to charge, stated plainly, because a waiting list for a free thing tests nothing. Then the email field, labeled honestly: join the waiting list, the first version opens later this season, nothing exists yet.

That last line matters more than the design. Pretending the product exists, faking a countdown or inventing testimonials would pass the conversion test and fail the Integrity Filter. Say it is a waiting list. The people who sign up anyway are the ones you want.

Then put the page in front of the person: one honest post in the places where they already gather, describing the problem in their language and linking to the page. One post, ten places, a count. The real page, when it is time, uses the structure in the one-page sales page that converts; the promise page is that structure with a waiting list where the button goes.

How many people do you need to validate a business idea?

Ten to fifteen real conversations with strangers who have the problem, plus a measurable response to one honest post. On the post, count interest as a ratio: under ten percent of the people who read the promise leaving an email means the person or the problem is wrong. Above it, you have enough to build a small first version.

The ten percent is my rule of thumb, not a law of nature, and it applies to the people who actually read the promise, not to everyone who scrolled past. Forty readers and five emails puts you above the line. Four hundred readers and eight emails does not.

The conversations are where the real information lives. Fifteen minutes each, and you do not pitch. You ask: what have you tried, what did it cost you in money, time or sleep, what would a solution need to include for you to trust it, what would you expect to pay. Then you stay quiet and write down the exact words. Selling is diagnosing, and so is validating: the question bank in the Diagnosis Selling Playbook doubles as a validation interview script. The mindset is in how to sell without being pushy.

Why ten to fifteen and not fifty? Because by the tenth call the words stop changing and new calls confirm rather than teach. If every call describes a different problem, you do not have a segment yet. You have a crowd.

Three filters before you commit

Proof tells you people want the bridge. It does not tell you whether you should build it. Three filters before you spend a week on the product.

Are they already paying?

Competitors are not bad news; they are proof of a paying market. An empty field usually means the pain is too weak to pay for. Enter proven demand and serve one segment better, more specifically, than the incumbents do.

Can you reach them?

A perfect problem held by people you have no way to contact is a research topic, not a business. The test is the list of ten: ten places, groups, newsletters, events or people where your person already gathers. If you cannot fill the list, you cannot run the promise page, or the business.

Does helping them pass the Integrity Filter?

Does the customer end up genuinely better off, in a way they would confirm themselves? Is the exchange honest, with no hidden traps and no interest-bearing debt engineered into the model? Would the business still make sense if it only won when the customer won? If any answer is no, the idea is broken at the root, whatever the validation score says. This standard comes from my faith, and you do not need to share it to share the standard.

A worked example, end to end

Theory hides; examples confess. Sara is an imaginary reader, but I have met many versions of her. She is an accountant with nine years of experience, and she almost starts "a course about accounting," which is the productivity-course mistake wearing a calculator. Instead she works the order. Person first: who does she understand from the inside? Freelancers who dread tax season, because half her clients arrive every spring as panicked freelancers. Her sentence: I help freelance designers go from shoebox-of-receipts panic to taxes filed calmly, with a system they keep.

Problem check: it costs them money, time and sleep, every year, on a legal deadline, which is the strongest kind of problem, recurring and unavoidable. Already paying? Yes, accountants exist; her edge is a system built for designers' actual income patterns, in designer language. Reachable? Two communities where designers gather. Integrity Filter: a designer who files calmly and keeps the system is plainly better off.

Validation before building. She posts honestly in the two communities, offering fifteen-minute calls about tax pain. Eleven people book. On the calls she hears the same four fears in the same words, and she writes them down; they will become the headlines of her storefront. Then she drafts a one-page description of a small offer: templates, three short lessons and one live question-and-answer session, at a modest launch price for the first twenty buyers. She sends it to the eleven. Seven say "take my money" in some dialect.

Now she builds, and only now. Money spent: zero. Guesses made: almost none, because the market answered every question she was tempted to answer herself. The same four beats work for a plumber teaching apartment repairs or a developer productizing website audits, and they are how my own company started.

What to do when validation fails

Say Sara's post had produced two calls instead of eleven, or four emails from two hundred readers. Under ten percent interest, the rule is simple: change the person or the problem, not the fonts.

A failed validation is a diagnosis, and it points at one of three places. The person is wrong: the problem is real, but the people you reached do not feel it, or the people who feel it gather somewhere else. The problem is wrong: they have it but do not feel it, the way everyone should learn statistics and nobody searches for it at midnight; sell the felt version and deliver the important version inside it. Or the promise is wrong: "be more organized" instead of a specific A and a specific B.

What it almost never means is that the page needs a video or a better color palette. Those are fonts. Redesigning the page after a silent launch is the most expensive way to avoid the real question, and building anyway "to prove them wrong" is the second.

Change one thing and run the loop again. A loop costs an afternoon and one post; three loops cost a month. Results depend on your market and on the work you put in, and nothing here changes that. What the method changes is the price of being wrong: an afternoon instead of three months.

Do this tonight

That is the whole method. It is slower for the first two weeks and faster forever after, because everything you build afterwards sits on a Value variable that is not a zero. Version one of your business should be a row of honest 6s, and validation is how you make sure the first 6 is real before you spend ninety days building the other four on top of it.

Youness ES-SEBIY

Engineer (EHTP) and founder of YOUNESS SCHOOL, an online education company built since 2019 with no office, no investors and no debt. More than 2,000 students coached directly, tens of thousands reached with free content, students in four countries. Author of The Business Equation. Read the full story.