Customer discovery
Market validation: how to test a business idea before you build it
How to validate a market before building: set pass or fail criteria, run problem interviews and expert calls, then ask for a commitment such as a paid pilot.
Market validation means collecting evidence, before you build much, that a specific group of buyers has a problem, already spends time or money on it, and will commit to your solution. The strongest evidence is something the buyer gives up: money, time, or their standing with colleagues when they introduce you. Compliments and "I would definitely use that" are weak evidence, because people say them to be polite.
If you are still working out what to research and in what order, market research to start a business covers that. This page is about the validation step itself: deciding in advance what would count as a yes, and then collecting that proof.
A worked example: food waste tracking for restaurants
Suppose you want to build an app that lets kitchen staff log food they throw away, so owners can see what goes to waste and order less of it. You have worked in restaurants and believe waste is expensive, but you do not yet know whether owners will pay to track it. This example is hypothetical and used throughout the page.
Decide what counts as validation before you start
Write down pass and fail criteria before the first conversation. Without them, it is easy to read every result as encouraging. For the restaurant app, the criteria might be:
- Most owners we interview can describe a recent week when they threw away food they later had to reorder.
- Owners, not only chefs, say waste is a cost they watch.
- At least 3 of the first 20 restaurants agree to a paid 60-day pilot at our proposed price.
These numbers are examples, not a standard. Pick thresholds that would justify the time you are about to spend, and do not lower them after the results come in.
Know which evidence is strong
| Evidence | What it shows | Strength |
|---|---|---|
| "Great idea," "I'd use that" | Politeness | Weak |
| A detailed story about a recent incident | The problem happens | Moderate |
| Money or time already spent on a workaround | The problem costs something | Stronger |
| A second meeting, an introduction to the decision maker, or data shared | It is a priority for them | Stronger |
| A deposit, pre-order, paid pilot or signed letter of intent | They will pay | Strongest before launch |
Rob Fitzpatrick's The Mom Test is a short book on exactly this problem: how to learn whether your business is a good idea "when everyone is lying to you," including how to avoid biased feedback and tell whether someone will actually buy.
Run problem interviews
Start with the people who would use and pay for the product. Ask about the last time the problem happened, not about your idea:
Tell me about the last time you threw out food you had ordered that week. What was it, how did you notice, and what did you change afterward?
Listen for specifics: what was thrown away, what it cost, who noticed, whether anything changed. An owner who cannot remember a recent example is telling you something.
How many conversations you need depends on how clear the pattern is. NSF's I-Corps program, which trains research teams to judge a technology's commercial potential, requires a minimum of 100 potential customer interviews during its seven-week program. You do not need 100 to spot a clear no, but it shows the scale a structured program expects. How many customer interviews do you need covers when to stop, and problem vs. solution interviews covers when to start showing your idea.
Add expert calls for what buyers cannot tell you
Owners know their own kitchens. They usually cannot tell you how the rest of the market works: how food distributors price and deliver, which software restaurants already pay for, why earlier waste-tracking products did or did not stick. For that, talk to people who see many restaurants at once, such as a distributor's sales rep, a purchasing manager at a restaurant group, or someone who used to sell restaurant software.
Treat what they say as leads to check with owners, not as validation on its own. Market research interviews with customers, suppliers and operators explains how to combine those views. If someone used to work for a competitor, tell them up front that you do not want confidential information; what to do when an expert cannot share confidential information covers the moment they cannot answer.
Ask for a commitment
Interviews tell you the problem is real. A commitment tells you the market will pay. Common forms:
- Paid pilot: a short, paid trial with a clear end date and success measure. This is often the most natural option for businesses.
- Letter of intent: a signed statement that the buyer plans to purchase at a stated price if the product does what you describe. It is weaker than money, but stronger than a verbal yes.
- Pre-order or deposit: money up front for something you will deliver later.
- Doing it by hand first: Paul Graham's Do Things that Don't Scale describes doing by hand what you plan to automate later. That can let you charge for the result before the software exists. For the restaurant app, you could count waste yourself for a few weeks and deliver a weekly report.
Be clear with buyers that the product is not finished, and put the delivery date and refund terms in writing. If you take pre-orders for physical goods by mail, phone or online in the U.S., the FTC's Mail, Internet, or Telephone Order Merchandise Rule requires a reasonable basis for the shipping time you state, or for shipping within 30 days if you state none. If you cannot ship on time, you must ask the buyer to agree to the delay or, if you will not ship at all, cancel and promptly refund them. The rule does not cover services, and other rules may apply to them and to software. This is general information, not legal advice, so check what applies to you.
Compare the results with your criteria
When you have run the planned conversations and asked for commitments, compare the results with the criteria you wrote first:
- Pass: build the smallest version that serves the customers who committed, and keep talking to them.
- Mixed: look for a sub-group that passed. Perhaps multi-location restaurant groups committed and single restaurants did not. Rewrite the criteria for that group and test again.
- Fail: change the segment or the problem, or stop. A clear no after a few weeks costs far less than a product nobody pays for.
Your next step
Write three pass-or-fail criteria for your idea today, including one that requires a commitment. Then book the first five problem interviews.
If the buyers or industry people you need are outside your network, Instant Expert can find people who match a description you write, such as "owners of independent restaurants with two to five locations" or "sales reps at food distributors." You review who it finds, it sends your invitations, and you pay only for calls that get booked. The directory pages for procurement professionals in restaurants and sales professionals in food distribution are one place to start.