Customer discovery
Customer Discovery — How to Find and Talk to Potential Customers
Customer discovery is the process of finding out whether the problem you want to solve is real, frequent, and painful enough that people will pay for a solution. It starts with listening, not selling. The goal is to understand the customer's world — their problems, their workarounds, their language — before you build anything.
Last updated · September 5, 2026
Quick answer
What is customer discovery?
Customer discovery is the process of identifying, finding, and interviewing potential customers to validate the assumptions behind a startup idea. It is not sales — it is research. The goal is to understand whether the problem you want to solve is real, frequent, and painful enough that people will pay for a solution. Customer discovery happens before you build, not after. It replaces 'I think people need this' with 'Here is what people told me about their problem.'
Key takeaways
The core idea in30seconds
- Customer discovery is research, not sales. You are listening for problems, not pitching solutions.
- Talk to people who have the problem, not people who are polite to you. Friends and family will lie to protect your feelings.
- Ask about the past, not the future. 'Tell me about the last time you had this problem' is more reliable than 'Would you use a product that...?'
- Listen for the words they use, not the words you use. Their language reveals how they think about the problem.
- Five good interviews can reveal more than five hundred surveys. Quality of signal matters more than quantity of responses.
The process
The5-step customer discovery process
A structured approach to finding out whether your startup idea is worth building.
- 1
Name the assumption
Write down the core assumption your idea depends on. 'Busy professionals struggle to plan meals.' 'Small businesses cannot afford accounting software.' The assumption becomes the focus of your discovery.
- 2
Define who to talk to
Identify the specific type of person who has the problem. Not 'everyone' — a specific role, industry, or situation. 'Restaurant owners with5+ locations' is better than 'small business owners.'
- 3
Find5-10 people
Use LinkedIn, Reddit, industry communities, or your existing network to find people who match your target profile. Aim for5-10 conversations. Fewer than five is not enough data; more than ten is usually unnecessary for the first round.
- 4
Conduct the interviews
Ask open-ended questions about their current situation, not about your solution. 'Tell me about the last time you had this problem.' 'What did you do to solve it?' 'How much time/money did that cost?'
- 5
Look for patterns
After5-10 interviews, look for patterns. Do multiple people describe the same problem in similar words? Do they already spend time or money on workarounds? Patterns are signal. Outliers are noise.
Who to talk to
Finding the right people to talk to
Not everyone is a good candidate for customer discovery. Here is who to target.
People who have the problem right now
The best candidates are people who are actively dealing with the problem today. They can describe recent experiences, current workarounds, and real frustrations. People who had the problem last year may have solved it or given up.
People who are already spending money on workarounds
If someone is already paying for a solution — even a bad one — they have validated the problem with their wallet. They are more likely to pay for a better solution than someone who has never tried to solve the problem.
People who are not your friends or family
Friends and family will lie to protect your feelings. They will say 'That's a great idea' even if they would never use it. Find strangers who have no reason to be polite.
People in your target market, not adjacent markets
If you are building for restaurant owners, talk to restaurant owners, not food bloggers. Adjacent markets have similar problems but different contexts, budgets, and decision-making processes.
The interviews
How to conduct customer discovery interviews
The goal is to understand the customer's world, not to sell your solution.
Start with the last time, not the future
Ask 'Tell me about the last time you had this problem' instead of 'Would you use a product that...?' The past is concrete. The future is hypothetical. People are bad at predicting their own behavior.
Ask open-ended questions
Questions that start with 'Tell me about...', 'Walk me through...', or 'What happened when...' invite stories. Questions that start with 'Do you...', 'Would you...', or 'Is it true that...' invite yes/no answers, which are less useful.
Listen for the words they use
The words people use to describe their problem reveal how they think about it. If they say 'It's annoying' vs 'It costs me $500 a month', those are very different signals. Write down their exact words, not your interpretation.
Do not pitch your solution
The moment you start pitching, the interview stops being research and starts being sales. The person will shift from describing their problem to evaluating your solution, which is less useful information.
Ask about workarounds
What are they doing today to solve the problem? Spreadsheets? Hiring someone? Ignoring it? Workarounds reveal the pain level and the willingness to pay. If nobody is trying to solve the problem, it may not be painful enough.
The Mom Test (Rob Fitzpatrick)
The Mom Test is a set of rules for customer discovery interviews. The core rule: ask about their life, not your idea. If you mention your idea and they say 'That's cool', you have learned nothing. If you ask about their last experience with the problem and they describe spending3hours on a workaround, you have learned something real.
What to listen for
Signals that the problem is real
Five patterns that indicate the problem is worth solving.
They describe the problem in their own words
When multiple people independently describe the same problem using similar language, it is a strong signal. The problem is real enough that people have their own way of talking about it.
They have tried to solve it themselves
Workarounds — spreadsheets, manual processes, hiring someone — indicate that the problem is painful enough to spend time or money on. The more workarounds they have tried, the more painful the problem.
They can describe a recent specific moment
When someone says 'Last Tuesday, I spent2hours trying to...', that is a concrete signal. When they say 'Sometimes it's hard to...', that is vague. Specific recent moments are more reliable than general statements.
They are already paying for a partial solution
If someone is already paying for a tool, a service, or a person to help with the problem, they have validated the problem with their wallet. They are more likely to pay for a better solution.
They ask when your solution will be ready
When someone asks 'When can I try it?' or 'Where do I sign up?', that is a strong signal. But be careful — verbal interest is not the same as payment. The next step is to test willingness to pay.
Common mistakes
Common customer discovery mistakes
Four mistakes that waste time or produce misleading signals.
Talking to friends and family
Friends and family will say 'That's a great idea' because they want to support you, not because they would use the product. Find strangers who have no reason to be polite.
Pitching instead of listening
The moment you start pitching, the person shifts from describing their problem to evaluating your solution. You get 'That's cool' instead of 'Last week I spent3hours on this.'
Asking leading questions
Questions like 'Don't you think it would be great if...?' or 'Wouldn't you pay for...?' lead the person to say yes. Ask neutral questions: 'What do you do when...?' or 'Tell me about the last time...'
Stopping after2-3 interviews
Two or three interviews are not enough to see patterns. Aim for5-10. If the first three are consistent, keep going. If they are inconsistent, you need more data to understand why.
How Yibud helps
How Yibud connects to customer discovery
Yibud's tools help you decide what to discover and how to structure the process.
Start with evaluation
Before you start customer discovery, evaluate your idea. The evaluator identifies which assumptions are most critical, so you know what to focus on in your interviews.
Try the evaluator →Use the Experiment Brief
Every Startup MRI report includes an Experiment Brief that recommends the cheapest way to test your highest-risk assumption. For many assumptions, customer discovery interviews are the recommended experiment.
See the Decision Guide →FAQ
Frequently asked questions
How many people should I talk to?
Aim for5-10 interviews in the first round. Fewer than five is not enough to see patterns. More than ten is usually unnecessary — if the first ten are consistent, you have a signal. If they are inconsistent, you need to understand why before talking to more people.
How long should each interview be?
20-30 minutes is ideal. Shorter than20 minutes and you may not get enough detail. Longer than30 minutes and the person may get tired or start giving you less honest answers.
What if nobody I talk to has the problem?
That is a signal — either the problem is not real, or you are talking to the wrong people. Before giving up on the idea, check whether you are targeting the right audience. If you are, and nobody has the problem, the assumption may be wrong.
What if people say they would use my product but won't pay?
Verbal interest is not willingness to pay. The next step is to test willingness to pay with a specific price and a real payment method. If people agree to the price but won't actually pay, the problem may not be painful enough to justify the cost.
Can I do customer discovery online?
Yes. Video calls, phone calls, and even text-based conversations can work. The key is to have a real conversation where you can ask follow-up questions, not just send a survey.
What if I can't find enough people to talk to?
If you cannot find5-10 people who match your target profile, that may be a signal that the market is too small or too hard to reach. Consider whether the distribution channel is viable before investing more time.
Is customer discovery the same as market research?
No. Market research is broad — it includes surveys, industry reports, and competitor analysis. Customer discovery is specific — it is about talking to individual potential customers to understand their problems, workarounds, and language.
Do I need to be technical to do customer discovery?
No. Customer discovery is about listening and asking questions, not about technical knowledge. The best customer discovery interviews are conducted by people who are curious about the problem, not by people who already have a solution in mind.
Start discovering your customers
Evaluate your idea first, then use the Experiment Brief to structure your customer discovery interviews.
Evaluate my idea →