Customer Discovery
Customer discovery means leaving your desk to talk with potential customers and test the assumptions behind your idea. It is the cheapest, fastest way to find out whether you are solving a real problem for people who will pay, before you spend months building.
Steve Blank's customer development
Steve Blank noticed that many startups failed not because they could not build products, but because they built products nobody wanted. In his book The Four Steps to the Epiphany he proposed customer development, a process that runs alongside product development. It has four steps:
- Customer discovery: test whether the problem and your proposed solution matter to customers.
- Customer validation: test whether you have a repeatable way to sell, for example by getting real orders.
- Customer creation: build demand and scale marketing once the model works.
- Company building: shift from a searching startup to an organisation built to execute.
Blank's famous advice is to "get out of the building": the facts you need live with customers, not inside your office. The first two steps are about searching; the last two are about executing.
Turn assumptions into testable hypotheses
List what must be true for your idea to work, then turn each assumption into a hypothesis that could be proved wrong:
| Vague assumption | Testable hypothesis |
|---|---|
| Students need help with fees | At least 6 of 10 parents we interview have paid school fees late in the past year |
| Salons want booking software | Salon owners lose at least two customers a week because of double-booked or forgotten appointments |
| People will pay | At least 3 of 15 salon owners will pay a KES 500 deposit for a one-month trial |
The Mom Test: asking questions that give honest answers
Rob Fitzpatrick's book The Mom Test is named after a simple idea: ask questions so well that even your mother, who loves you and wants to be supportive, cannot mislead you. His three rules, in our words:
- Talk about their life, not your idea. Once you pitch, people become polite.
- Ask about specific past behaviour, not opinions or predictions about the future. "Would you use this?" invites a kind lie; "When did this last happen?" invites a fact.
- Talk less and listen more.
| Weak question | Better question |
|---|---|
| Do you think this app is a good idea? | How do you currently keep track of appointments? |
| Would you pay for this? | What have you tried or paid for to solve this before? |
| Is this a problem for you? | Tell me about the last time this went wrong. What happened next? |
| How much would you pay? | What does this problem cost you each month in time or money? |
Fitzpatrick also warns about three kinds of misleading data: compliments ("great idea!"), fluff (generic claims like "I always..." or future promises) and ideas (feature requests you should dig behind to find the real need). The strongest signal is commitment: someone giving time, reputation or money, such as agreeing to a trial, introducing you to their boss, or paying a deposit.
A customer discovery plan
- Pick a specific segment: "salon owners in Nakuru town with 2–6 staff", not "small businesses".
- Write 5–8 open questions about their current behaviour and problems.
- Find people: walk in during quiet hours, ask for referrals at the end of each interview, use relevant WhatsApp or Facebook groups.
- Interview 10–20 people, 15–30 minutes each. Take notes immediately afterwards.
- Score your hypotheses: confirmed, rejected or unclear.
- Ask for a commitment in later interviews: a pilot, a deposit, an introduction.
- Decide: persevere, adjust or pivot.
Lydia, 20, believed salon owners in Nakuru needed booking software. She interviewed 14 owners using Mom Test questions. Only two had problems with double bookings, because most customers simply walked in. But 11 of 14 described a different, painful problem: buying hair products on credit from distributors and losing track of what they owed. Three asked whether she could help. Her original hypothesis was rejected, but she discovered a stronger problem. She ran a second round of interviews on supplier credit before building anything.
Counting "Yes, I'd definitely use that!" as validation. Opinions about the future are unreliable. Another mistake is interviewing only people who are easy to reach, such as classmates, when they are not your customers. Count commitments, not compliments.
Key points
- Steve Blank's customer development has four steps: discovery, validation, creation and company building.
- Turn assumptions into hypotheses that could be proved wrong, then test them by talking to customers.
- Rob Fitzpatrick's Mom Test: talk about their life, ask about specific past behaviour and listen more than you talk.
- Compliments and opinions are weak evidence; commitments of time, reputation or money are strong evidence.
Tip: use the ← and → keys to move between pages.
