How to Validate a Business Idea Before You Build Anything

Most idea validation is theatre. You describe the concept to friends, they say it sounds great, you take that as evidence and start building. Six months later you discover that sounding great and being worth paying for are unrelated properties.

Real validation answers one question: who currently spends money or effort solving this problem, and how?

Start with the sentence

Before anything else, fill in this sentence with real specifics:

[Specific customer] currently pays [amount] to [current solution] for [outcome].

If you can't complete it, that's your first finding. Every blank is a research task.

Notice what it forces. Not "small businesses need better bookkeeping" but "solo trades businesses pay a bookkeeper around $300 a month to reconcile receipts they photograph badly." The second sentence tells you the buyer, the budget, the incumbent, and the actual job. The first tells you nothing you can act on.

"Current solution" is often not a product. It's a spreadsheet, an assistant, a manual process, or living with the problem. Those are your real competitors, and they're harder to displace than software, because they're already paid for.

The evidence hierarchy

Not all validation signals are equal. Ranked from weakest to strongest:

| Signal | Worth | |---|---| | "That's a great idea" | Nothing. Politeness. | | "I would definitely use that" | Nearly nothing. Hypothetical. | | "We tried to solve this last year" | Real. Shows effort spent. | | "Here's the spreadsheet we use instead" | Strong. Shows a live workaround. | | "We pay $X a month for something similar" | Strong. Budget exists. | | "Can I get early access?" + email | Meaningful interest. | | A deposit, pre-order, or signed LOI | The only conclusive signal. |

Everything above the line costs the person nothing. Everything below costs them something. Only the second kind predicts behavior.

A single deposit is worth more than a hundred encouraging conversations. Money is the only feedback that has skin in it.

How to run the conversations

Fifteen conversations, two weeks, one rule: never ask about the future.

People are unreliable narrators of their own hypothetical behavior and reliably accurate about their past. So ask about the past.

Ask these:

• "Walk me through the last time this problem came up."
• "What did you do about it?"
• "How long did that take?"
• "What did it cost you — money, time, or a worse outcome?"
• "What have you already tried?"
• "Why didn't that work?"

Never ask these:

• "Would you use this?"
• "Would you pay $X for it?"
• "Do you think this is a good idea?"

The first set produces facts. The second produces encouragement, which feels identical during the conversation and is worthless afterward.

Don't pitch. The moment you describe your solution, the conversation stops being research and becomes a sales call where the other person is trying to be nice. Save the pitch for the last two minutes, and only to ask: "Would you want to hear when this exists?" — then see whether they give you an email address without being chased.

Where to find fifteen people

• Communities where they already complain — forums, subreddits, Slack and Discord groups, LinkedIn groups. Read before posting; the complaints are the research.
• Your existing network, two degrees out. Ask for introductions, not opinions.
• Competitors' reviews. One and two-star reviews of adjacent products are people telling you exactly what the current solution fails at.
• Job postings. If companies are hiring someone to do this manually, there's budget for it.

If you cannot find fifteen people who have this problem, that is itself the answer. A problem whose sufferers are hard to locate is a problem that will be expensive to sell into later.

Desk research that's actually worth doing

Before the conversations, spend a day on evidence you can gather alone:

Do paid competitors exist? If several companies charge for this and have been around a few years, demand is proven. That's good news, not bad — you now compete on positioning rather than on educating a market from scratch.

Is anyone searching for it? Search volume for the problem, not your solution. People search for problems.

Is there an active community? A busy forum means an acute, ongoing pain. A dead one usually means the problem is real but tolerable.

What do the workarounds look like? Templates, spreadsheets and manual processes people share with each other are a map of unmet demand.

The test that ends the debate

Once the conversations point one way, run a paid test before building anything:

• Pre-sell it. Describe the offer, name the price, ask for a deposit. Refund anyone who changes their mind — the point is the decision, not the money.
• Sell the service manually first. Deliver the outcome by hand for three customers. If nobody buys the manual version, nobody wants the automated one.
• Run a landing page with a real checkout. Not a "notify me" form — an actual payment step. The gap between signups and payments is the most honest number you'll get.

Manual delivery is the most underrated of the three. It validates demand, teaches you the process you'd eventually document, and generates revenue while you're still deciding.

Reading the result honestly

Strong signal: people describe the same painful workaround unprompted, have tried and abandoned alternatives, and at least a couple will pay before it exists.

Weak signal: everyone is encouraging, nobody has tried to solve it, and the enthusiasm evaporates when you mention price.

No signal: you struggled to find fifteen people who have the problem at all.

The last one is the most common and the most ignored. Difficulty finding people who have your problem is not a marketing challenge to be solved later. It's the finding.

What validation is not

• It isn't a guarantee. It removes the most expensive failure mode — building something nobody wants — and nothing more.
• It isn't building an MVP. If it takes three months, you're building.
• It isn't a survey. Surveys collect opinions at scale; you need reasons at depth.
• It isn't permanent. Markets shift. A validated idea still needs to be re-checked against reality once you're selling.

What to do next

Write the sentence at the top of this article with real specifics. Then list every blank you couldn't fill, and go find fifteen people who have the problem. Two weeks, no pitching, no building. The answer is usually obvious by conversation eight.