The most expensive mistake in business isn’t a bad logo or the wrong tool — it’s spending six months building something nobody actually wanted. And almost everyone does it.
It feels productive to build. It feels like progress. So people disappear into building mode for months, emerge with a finished product, and only then discover the market shrugs. The painful part is how avoidable it is. You can find out whether people want your thing in days, for almost no money, before you build a single feature. Validation isn’t a step you do after building — it’s what you do instead of building, until the evidence earns the build. Here’s how to do it properly.
Why “build it and they will come” is a lie
The fantasy is that a great product sells itself — make something good enough and customers appear. Reality is the reverse: the graveyard of failed businesses is full of well-built products that solved a problem nobody was willing to pay to fix. Building first inverts the risk. You commit your scarcest resources — time and money — to an unproven assumption, and you only learn the truth after it’s too late to cheaply change course. Validation flips that: you spend cheaply to learn the truth first, then build only what the evidence supports. The goal of the early stage isn’t to build a product; it’s to reduce your uncertainty about whether anyone wants it.
Start with the problem, not your solution
The instinct is to fall in love with your solution and look for people to sell it to. Backwards. Start with the problem and confirm it’s real, painful, and common before you mention what you’ve built. A real problem is one people already spend time, money, or stress trying to solve — with clunky workarounds, spreadsheets, or a competitor they complain about. If the problem is real and painful, a solution has a market. If you have to convince someone they have the problem, you don’t have a business; you have an education campaign you can’t afford. The first thing to validate is never “do they like my idea” — it’s “does this problem genuinely hurt.”
Have the conversations (and shut up)
The single highest-value validation activity is talking to potential customers — and the single most common way to ruin it is pitching instead of listening. Don’t describe your idea and ask “would you use this?” People are polite; they’ll say yes and never buy. Instead, ask about their actual experience: How do you handle this today? When did you last run into this problem? What did you do about it? How much does it cost you? You’re hunting for evidence of real pain and real spending, not compliments on your concept. Aim to talk to a couple of dozen people who fit your target, and listen far more than you talk. The truth is in what they already do, not in what they say they might do.
The questions that get honest answers
Notice none of these mention your idea. The best validation conversations can happen before your product exists at all.
The ultimate test: will they pay before it exists?
Talk is cheap, and even genuine enthusiasm doesn’t pay the bills. The strongest possible validation is someone paying — or committing to pay — before the thing is fully built. This is the pre-sale, and it cuts through all the politeness instantly. You can offer an early-access deal, take a deposit, sell a “founding member” spot, or simply ask “if I build this, will you buy it at this price — here’s where to pay.” The moment money is on the line, vague interest either becomes a real commitment or evaporates, and either answer is gold. People reaching for their wallet is the only validation that fully counts; everything else is a leading indicator. If you can pre-sell even a handful, you have something. If no one will pay before it exists, that’s the cheapest “no” you’ll ever get.
Reading the signals: real interest vs politeness
Validation is noisy, so learn to tell strong signals from weak ones. Weak signals: “that’s a great idea,” “I’d definitely use that,” “let me know when it’s ready” — all free to say and worth nothing. Strong signals: someone pays a deposit, refers you to a colleague unprompted, asks detailed how-do-I-start questions, or gets visibly frustrated describing their current workaround. The rule of thumb is that anything that costs the person something — money, time, a reputation by referring you — is a real signal, and anything free to say is noise. Beginners hear enthusiasm and feel validated; experienced founders discount the words and watch the costly actions. Calibrate to actions and you won’t fool yourself.
The smallest thing you can test with
You don’t need the full product to validate — you need the smallest thing that lets someone experience the value or commit to it. That might be a simple landing page describing the offer with a “buy” or “join the waitlist” button, a manual version where you deliver the service by hand before automating anything, or a single core feature instead of the whole vision. The point is to get real behavior against a real offer as fast and cheaply as possible. Delivering manually first is especially powerful: it validates demand and teaches you exactly what to build, all while earning revenue. Build the minimum that produces a real yes or no, not the maximum you can imagine.
When to actually start building
Validation gives you permission to build — but only once the evidence is there. You’re ready when you’ve confirmed the problem is real and painful through honest conversations, when you’ve seen strong signals (ideally money) rather than polite enthusiasm, and when you understand the solution well enough from talking to real people that you’re building to a known need rather than a guess. At that point building is no longer a gamble; it’s executing on demand you’ve already proven exists. If you’re not there yet, more validation is cheaper than more building. The discipline of staying in validation until the evidence earns the build is what separates the businesses that survive from the well-built ones that quietly die.
Validation never really ends
It’s tempting to think of validation as a gate you pass through once and then forget — prove the idea, build it, done. In reality the mindset that validates an idea is the same one that keeps a business alive for years. Markets shift, customers’ problems evolve, and what people wanted last year isn’t always what they’ll pay for next year. The founders who last keep a thread of validation running permanently: they stay close to customers, keep asking about real behavior, and test new offers small before committing to them. A new feature, a new product line, a new market — each deserves the same cheap proof you gave the original idea, rather than a leap of faith funded by past success. This is also the cheapest insurance against the slow drift that kills established businesses, where a company keeps building what used to work long after the market quietly moved on. Treat validation not as a phase but as a habit: a standing willingness to ask “do people actually want this?” before you pour resources into the answer. The discipline that saved you from building the wrong first product is the same one that will save you from building the wrong tenth. Keep talking to the people you serve, keep watching what they do rather than what they say, and you’ll keep building things the market actually rewards instead of things that merely seemed like good ideas from inside your own head.