How to validate a business idea before spending money

Level

Foundational

Reading time

13 minutes

Updated

November 3, 2026

Sources

6 cited

Useful for

All business types

What validating an idea means

Every business idea is a bundle of guesses. You're guessing that a certain group of people has a certain problem, that it bothers them enough to pay for a fix, that they'll pay your price, and that you can reach them without spending more than they're worth.

Validation is the work of replacing those guesses with evidence before you put serious money behind them. It can't prove your business will succeed. It shrinks the chance that you spend six months and your savings on something nobody buys.

The U.S. Small Business Administration describes market research as a way to reduce risk while a business is still only an idea.^[1] Validation is the most hands-on form of that research: instead of reading about a market, you test it. The Lean Startup methodology calls the result validated learning and treats it as the unit of progress for a new venture, gathered by running experiments that test each part of your plan.^[2]

Why it matters before you spend money

Most of the money a new business spends early is hard to get back. A lease, a batch of inventory, or a custom app costs the same whether customers show up or not. Conversations and small tests usually cost a few weeks of your time and a few hundred dollars or less. Make your biggest commitments after you have evidence, not before.

Why commitments count more than compliments

People are polite. Describe your idea and most will say something encouraging, because that's easier than explaining why it doesn't fit their life. Rob Fitzpatrick's book The Mom Test is built around this problem: how to learn something useful from customer conversations when people's answers are polite rather than true.^[3]

The fix is to watch what people give up, not what they say. A useful way to rank evidence is by what it costs the person giving it.

Time, reputation, and money are the three currencies of a real commitment. Design every test to ask for at least one of them.

A step-by-step method

This method works for any kind of business. The section on business types below covers how the details change.

1. 01

   Write down the problem and exactly who has it

   Fill in one sentence: [Specific customer] struggles with [problem] when [situation], and today they handle it by [current workaround].

   Be narrow. "Busy parents" is not a customer. "Parents with two or more children under ten, both working full time" is. A narrow group is easier to find, interview, and sell to. MIT's Disciplined Entrepreneurship framework, a 24-step method for building a new venture, opens with the question of who your customer is.^[4]

2. 02

   List your assumptions and pick the riskiest three

   Write down everything that has to be true for the business to work. Most assumptions fall into five groups:

   • Problem: the people you described actually have this problem, and often.
   • Payment: they'll pay to solve it, at a price that works for you.
   • Reach: you can find and reach them at a reasonable cost.
   • Delivery: you can deliver what you promise and still make a profit.
   • Preference: they'll choose you over what they do today.

   Then rank them. The riskiest assumptions are the ones that would sink the business if wrong and that you have the least evidence for. Test those first.

3. 03

   Set a pass or fail threshold before you test

   For each risky assumption, write a statement with a number in it, and decide in advance what result counts as a pass. For example: "At least 15 people will pay a $20 refundable deposit within 14 days."

   Setting the bar first protects you from your own optimism. Once results come in, it's tempting to decide that whatever happened was good enough. A written threshold can't quietly move.

4. 04

   Talk to at least ten people who have the problem

   Interview people who match your customer description and have dealt with the problem recently. Ask about what they've done, not what they think of your idea: the last time the problem came up, what it cost, and what they use today. If you're still hearing new things after ten, keep going. Our guide to running customer interviews that get honest answers walks through the method, and the customer interview guide gives you a script and a question list.

5. 05

   Map what people use today

   Your real competition is whatever customers do now: a competitor, a spreadsheet, a relative who helps, or putting up with the problem. List the most common alternatives, what each costs, and where each falls short.

   If people already pay for a clumsy workaround, that's good news: the problem is costly enough to spend money on. If nobody has tried to solve it, ask whether they truly care about it.

6. 06

   Ask for a real commitment

   Now test whether people will act. Pick the smallest test that asks for time, reputation, or money:

   • A pre-order or deposit, with a clear delivery date and a refund promise.
   • A landing page with a price and a button to reserve or pre-order, promoted to people who match your customer.
   • A paid pilot or fixed-price first project for a business customer.
   • A letter of intent stating what a larger organization would buy, and at what price, if you deliver.

   Show the real price. A free waitlist tells you about curiosity. A price tells you about demand.

7. 07

   Check that the numbers can work

   Demand isn't enough if every sale loses money. Estimate what one sale earns after its direct costs, and how many sales cover your fixed costs. The SBA's break-even formula is a simple starting point: divide your fixed costs by the difference between your price and your variable cost per unit.^[1]

   Rough numbers are fine. The unit economics calculator lets you try different prices and costs.

8. 08

   Decide: proceed, pivot, or stop

   Compare each result with the threshold you set. Then make a decision and write it down:

   • Proceed if your riskiest assumptions passed. Move on to the next-riskiest ones, or start small.
   • Pivot if some assumptions passed and others didn't. Change one major element, such as the customer, the price, or the offer, and test again.
   • Stop if the core problem or willingness to pay didn't hold up. Stopping early is the cheapest way to fail.

   The Lean Startup method frames this as knowing when to turn and when to persevere. When results show the plan isn't working, it calls for a pivot: a structural course correction, not more of the same.^[2]

A worked example

Common mistakes

How validation differs by business type

The principle is the same everywhere. What changes is the riskiest assumption, the best commitment to ask for, and what a strong result looks like.

The limits of validation

Use AI to prepare, not to replace customers

AI tools help with the thinking around validation: spotting assumptions you missed, drafting interview questions, and finding patterns in notes. They can't validate demand, because only customers can pay you.

For more ways to use AI in early research, see how to research a business idea with AI.

Your next step

Pick your three riskiest assumptions today and write a threshold for each. Then book your first five interviews.

For a structured route, the Validate Your Business Idea Founder Path takes you through this guide one stage at a time and ends with a Validation Decision Memo: a one-page record of what you tested, what customers committed, and what you decided.

Once you have paying customers, the next questions are about fit and price: read what product-market fit means and how to price your first product. For a course alongside this guide, Y Combinator's Startup School is a free online course on starting a company, including a session on planning a minimum viable product.^[6]

If the evidence says build and your product is software, you can build the first version yourself with AI tools. This free guide walks you through it, from planning to a deployed app: