
Opinions Are Free. That Is Exactly the Problem.
Every failed business began with encouragement. Friends called it a great idea, family said they would definitely buy one, and a survey said 80% of people were interested — and then the product launched to silence, because none of those people were ever asked the only question that validates anything: will you pay, now, this much? Learning how to validate a business idea before spending real money means replacing opinions with evidence in a deliberate order: first, proof that a specific group of people has the problem badly enough to seek a solution; second, proof that they will exchange money (or its stand-ins — deposits, pre-orders, signed commitments) for your version of the fix; and only then, the building. This guide gives the working sequence: the customer conversations that produce truth instead of politeness (and the questions that ruin them), the cheap tests — landing pages, waitlists, pre-sales, concierge pilots — ranked by how much their results can be trusted, the numbers that count as a pass at each stage, and the kill signals founders most often explain away, because the purpose of validation is not to bless the idea but to give it every fair chance to fail cheaply before it can fail expensively. It precedes our guide to online business growth strategies that work — growth tactics belong to businesses whose core bet is already proven — and pairs with our guide to how to fix your sleep schedule, for founders whose most undervalued business asset is a functioning brain at 9 a.m.
Step One: Conversations That Tell the Truth
Before building anything — including a landing page — talk to fifteen to twenty people who plausibly have the problem, recruited from where such people already gather (communities, workplaces, existing customer lists of adjacent products), not from your friends. The method has one governing rule: ask about their past behaviour, never about your future product. “Would you buy an app that does X?” invites fiction; “Tell me about the last time this problem cost you money — what did you do, what did you try, what did it cost?” produces evidence. Useful questions share a shape: they are specific (“in the last month”), behavioural (“what did you actually do”), and economic (“what did that cost in money, time, or embarrassment”). You are hunting for three things: the problem occurring unprompted in their stories, existing spending — a workaround, a competitor, a person they pay — because money already flowing is the strongest signal in validation, and pain with a deadline (a cost that recurs or grows) rather than mild annoyance. Record the exact phrases they use; those phrases become your marketing copy later, in the customer’s own language. And apply the discipline that makes the exercise worth anything: compliments are not data, enthusiasm is not data, “send me the link when it launches” is not data. A conversation counts only when it surfaces past behaviour, current spending, or a concrete introduction to someone with the problem — everything else is the politeness you are paying to see through.
Step Two: Tests That Cost the Customer Something
With problem evidence in hand, escalate to tests in ascending order of trustworthiness. The landing page test: a single page describing the offer as if it existed, with a price and a call to action, driven by a small amount of targeted traffic — measure not clicks but the conversion to the strongest available commitment (email signup for a waitlist is weak; a refundable deposit is strong). A page converting cold traffic to signups at a few percent says interest exists; it says nothing yet about payment. The pre-sale: sell the thing before it exists — a founding-customer discount, a pilot cohort, a crowdfunding campaign — with honest delivery dates. Pre-sales are the gold standard of cheap validation because they convert the hypothetical into a transaction; even a handful of strangers paying real money outweighs a thousand survey approvals. The concierge pilot: deliver the service manually to three to five customers — you, personally, doing what the product will eventually automate — at a real price. This tests willingness to pay and teaches you the actual job, the exceptions, and the moments customers value most, at the cost of your weekends rather than your savings. The fake door: a button or listing for a feature/product that does not exist yet, measuring click-through before an apology and an explanation — acceptable in moderation, corrosive if it becomes a habit. At every stage, define the pass mark before running the test (“if fewer than 5 of 20 target customers pre-pay, we stop or pivot”) — thresholds invented after the results arrive are not thresholds; they are rationalisations with numbers on.
Reading the Results: Green Lights, Kill Signals, and the Honest Middle
Validation ends in a decision, and the signals sort into three piles. Genuine green lights: strangers paying, unprompted referrals (“you should talk to my sister, she has this problem worse”), customers who complain when the pilot pauses, and repeat purchase or renewal without being chased — behaviour that costs the customer something is the only applause that counts. Kill signals — the ones founders explain away: people love it but route around paying (“we’d need it free / our budget cycle / maybe next year”), every conversation needs the idea re-explained from zero (the problem is not felt, it is being taught), interest exists only at a price below your cost floor, and the customer who would pay is never the person you are talking to — the eternal “great for someone else” is the sound of a market that does not exist. The honest middle is the most common result: a real problem, warmly received, in a segment too small, too cheap, or too hard to reach. The mature responses are the pivot (same problem, different customer — or same customer, sharper slice of the problem), the pause (park it with your notes; markets change), or the deliberate lifestyle choice (a small, profitable niche run knowingly is a success, not a failed unicorn). What validation can never certify is that the idea will succeed — execution, timing, and competition still hold their votes. What it certifies, for the price of some conversations and a few honest tests, is that you are no longer spending real money on an untested hope. In a game where most losses come from building things nobody buys, that is the cheapest insurance a founder will ever be offered.
