How to validate a startup idea in an afternoon

You cannot prove a startup idea in an afternoon, but you can find out whether it deserves a week. In one afternoon, name who has the problem, see who already solves it and how, check whether the numbers reach your goal, and list what would make it fail. End the afternoon with one test to run next.

What does it mean to validate a startup idea?

Validating a startup idea means collecting evidence that real people need it enough to pay, before you spend months building it. An afternoon gives you a first read, not proof. Proof comes from people using and paying for the product.

The goal of the afternoon is narrower: decide whether to keep going, change the idea, or drop it. All three are good results.

Who has the problem, and how badly?

Start with one sentence: who has this problem, and what does it cost them today. If you cannot name the people, the idea is probably weak.

Paul Graham's essay How to Get Startup Ideas puts the test sharply. Ask who wants this so much that they would use a rough first version from a startup they have never heard of. He also argues that an idea a small group needs badly beats an idea a large group finds mildly interesting.

Spend the first hour here. Write the sentence. Then name five real people or companies that fit it. If you know some of them, message them now and ask how they handle the problem today. Do not pitch. Ask what they did the last time it came up.

Who already competes, and how do they win?

Someone already solves the problem, even if the answer is a spreadsheet or doing nothing. Your job is to find out how they win today, not to prove you have no competition.

Spend the second hour listing every alternative you can find: products, services, and workarounds. For each one, write who it sells to, what it costs, and why people choose it. Then write one line on where you would win. If that line is "we are cheaper" or "we use AI," keep looking for a better one.

Do the numbers reach your goal?

The numbers work when a believable count of customers, at a believable price, reaches the revenue you need. Work backward from your goal, not forward from the size of the market.

A simple version takes fifteen minutes:

1. Write the monthly revenue you need for this to be worth doing.
2. Pick a price per customer per month that the alternatives make believable.
3. Divide. That is how many paying customers you need.
4. Ask whether you can name where that many customers would come from.

A large market number is easy to find and tells you little. The count of paying customers you need is more useful, because you can judge whether you could reach them.

What could make it fail, and what do you test first?

Every idea rests on a few beliefs that might be wrong. List them, then pick the one that would end the idea fastest if it were false. That is your first test.

Common ones: people do not feel the problem often enough, they will not pay the price, or you cannot reach them cheaply. A good first test is small and fast. Five conversations, a landing page with a price on it, or a manual version of the service for one customer.

How does Wizard by Ask The W help?

Wizard by Ask The W does the research half of this afternoon with you. It reads the market, lists competitors and how they win, estimates how many people could buy, and shows monthly revenue against your goal. It also asks what could make the idea fail and what to test first, and gives a verdict on the idea.

Wizard also learns your taste. You choose between ideas side by side, choose your domains, paste what you have built, or tell Wizard what to stop showing. Later picks keep refining what it learns. When you choose an idea, Wizard turns it into a project plan with outcomes, each marked as owned by a person, an agent, or both.

Wizard does not talk to your customers, and it does not decide for you. The conversations, the price you test, and the call to keep going stay with you. For more on how Wizard tests an idea, see validate a startup idea.

What should you have at the end of the afternoon?

You should have one page: the problem sentence, five named people, the alternatives and where you win, the customer count you need, and the first test. If the page looks weak, that is a result too. You saved months.