How to Validate a Business Idea (Without Wasting a Year)

Most "idea validation" advice is theatre. Surveys, friends saying "I'd totally use that," landing pages collecting emails that never convert. None of that is validation. Validation is one thing: evidence that strangers will pay you (or give up something scarce) to solve the problem you think you're solving.

This is a practical guide to getting that evidence in weeks, not years.

What validation actually is

Three things have to be true for a business idea to deserve another month of your life:

1. The problem is real and painful — people are already trying to solve it, badly.
2. Your solution makes their life measurably better — they can describe the difference.
3. They will give up something scarce — money, time, a calendar slot, a public commitment — to get it.

If you can't show all three with real people who don't love you, you don't have a validated idea. You have a hypothesis.

The validation pyramid

You don't validate everything at once. You stack evidence from cheapest to most expensive:

1. Problem evidence (free, ~1 week): Are people complaining about this problem in the wild?
2. Behavioural evidence (cheap, 1–2 weeks): What workarounds are they already using?
3. Commitment evidence (medium, 2–4 weeks): Will someone book time, sign a letter of intent, or pre-pay?
4. Repeatability evidence (the real test, weeks–months): Can you do it again, with a stranger, without rewriting everything?

Step 1 — find five people in the problem

Not five friends. Five strangers who, in the last 90 days, have personally felt the problem you think you're solving. If you can't find five in two weeks, the problem is either too rare or you don't know where these people live. Either way, that's information.

Good places: Reddit threads where people complain, LinkedIn comment sections, Slack/Discord communities, online reviews of competitors. Bad places: your own network and your co-founder's mum.

Step 2 — interview them the right way

The single biggest mistake first-time founders make is asking people about the future: "Would you use this?" "Would you pay for it?" The answer is always "yes" because being nice is free.

Instead, ask about the past:

• "Tell me about the last time you ran into this problem."
• "What did you do about it?"
• "What did that cost you — in money, time, or frustration?"
• "What have you already tried? Why did you stop?"

If the answer to "what did you do" is "nothing" — the problem isn't painful enough. Move on. This is the gift you give yourself in week two instead of finding out in year two.

Step 3 — look for the workaround

Real problems generate workarounds: ugly spreadsheets, WhatsApp groups, manual processes, paid hacks. The workaround is your competitor. If there is no workaround, the problem isn't painful enough yet. If the workaround is elaborate, you're onto something.

Step 4 — ask for a commitment

This is where most "validation" stops being theatre. Pick the smallest real commitment you can ask for:

• A 30-minute paid consult.
• A letter of intent for a pilot.
• A pre-order at a real price.
• A reserved calendar slot for a demo.
• An introduction to two other people with the problem.

"I'd love to try it" is not a commitment. "Here's my calendar — book me in" is. The exact ask matters less than the principle: did they give up something scarce?

Step 5 — do it manually before you build anything

Before you write a line of code or design a logo, deliver the value by hand to your first three customers. Concierge it. Use email and Google Sheets. The point isn't to be sustainable — it's to find out what they actually need versus what they said they needed.

Half of what you "validated" in interviews will turn out to be wrong here. That's not a setback; that's the validation. The half that survives is gold.

The effectual angle

This whole process is effectual thinking in motion. You're not predicting a market — you're co-creating one through commitments. Each "yes" from a real person changes what the venture is. That's not failure of planning. That's the plan revealing itself.

A validation checklist for this week

• [ ] I can describe the problem in one sentence without using my product's name.
• [ ] I have talked to 5 strangers who felt this problem in the last 90 days.
• [ ] I can describe, in their words, the workaround they're already using.
• [ ] At least one stranger has given me something scarce (time, money, a calendar slot).
• [ ] I've delivered value manually to at least one of them.
• [ ] I know exactly what to test next.

If you can tick those six boxes, you're past idea stage. If you can't, no amount of building will save you.

How to practice this on Playground

The Smart Stuff module turns your real conversations into an editable Business Model Canvas and hypothesis cards, so you can see which assumptions are still untested. The Sales Roleplay module lets you practice the customer conversations themselves with AI customers who push back like real humans do. Both are designed for exactly this loop.

Market validation, in one table

"Market validation" and "idea validation" describe the same job from two ends: does a market of people already spend something on this problem, and is your specific bet inside that market true? Turn each belief into a test with a pass mark you set before you run it.

Assumption	Test	Pass mark
The problem happens often enough to matter	Five problem interviews about the last occurrence	Four of five place it in the past month, unprompted
They already pay something to solve it	Ask what they spend today, workarounds and staff hours included	At least half name a real current cost
They will switch	Offer a paid pilot or pre-order at real price	One stranger commits money or a signed slot
You can reach them repeatedly	One channel test: post, cold email, or ad to a landing page	Response above your channel baseline
The economics can work	Price minus delivery cost across ten customers	A positive number without assuming scale

The step-by-step version, with what to do when a test fails, lives on validate your idea.

The rest of the validation toolkit

• Customer interview questions — the exact questions to ask at each stage, and the ones that produce useless answers.
• Business Model Canvas example — map the nine blocks and mark which ones are still assumptions.
• Business idea generator — if validation kills the idea, generate the next batch properly.
• AI sales roleplay — rehearse the conversation before you spend a real meeting on it.
• Entrepreneurship games — drill the judgement calls validation keeps asking of you.
• AI startup coach — talk the results through with a coach that pushes back instead of agreeing.