Glossary
What Is Customer Discovery?
Updated · Team Fonda
Short answer
Customer discovery is the process of testing your assumptions about who your customers are and what problem they have by talking to them before you build a product. The term comes from Steve Blank, who made it the first step of his customer development method.
Key takeaways
- You are testing guesses about the customer and the problem, not selling a product.
- Write your guesses down first, with what result would prove each one wrong.
- Evidence is what people have done and spent, not what they say they might do.
- It ends with a decision: go ahead, change something or drop the idea.
What customer discovery means in plain words
Every new idea rests on beliefs you have not checked: that a certain group of people has a certain problem, that it bothers them enough to pay, and that your approach would help. Customer discovery is going out to check those beliefs with the people concerned. Steve Blank, who created the customer development method, sums up the advice as get out of the building, because the answers are with customers and not at your desk.
It is different from selling. In a sales conversation you try to convince. In discovery you try to find out whether you are wrong, as cheaply and early as you can.
How customer discovery works
- Write your hypotheses. Who is the customer, what is the problem, how do they handle it today, and what would they pay to fix it? A lean canvas is a quick way to get these onto one page.
- Set a pass mark for each. Decide before you start what result would count as support and what would count as a miss.
- Find people who fit. Go where they already gather: online groups, trade events, local businesses, your wider network. Avoid close friends.
- Run [customer interviews](/resources/glossary/customer-interview). Ask about their past behavior around the problem. Do not pitch.
- Look for patterns after each batch. Tally how many people raised the problem unprompted, how many have tried to solve it and how many have spent money on it.
- Decide. Carry on, adjust the customer or the problem and run another round, or stop.
A hypothesis you can test
"I believe [customer group] struggles with [problem] when [situation], and today they cope by [current workaround]. I will know I am right if at least [number] of [number] people describe this without being prompted."
A worked example: late invoices for freelance designers
The figures below are made up to show how the tally leads to a decision. Say you believe freelance designers lose a lot of time chasing late payments and would pay for a tool that does the chasing.
You set a pass mark: at least 10 of 15 designers must have had a late payment in the last three months, and at least 8 must have tried something to fix it. Then you interview 15 designers.
| What you tallied | Result | Pass mark | Outcome |
|---|---|---|---|
| Had a late payment in the last three months | 12 of 15 | 10 | Passed |
| Tried a tool or service to fix it | 3 of 15 | 8 | Missed |
| Named finding steady clients as a bigger worry | 9 of 15 | Not set | Surprise |
The problem is real but most people shrug it off, so few would pay for a fix. Meanwhile a different problem came up again and again without you asking. The sensible move is to refine: run a second round focused on how designers find clients. You learned this from 15 conversations instead of from a product nobody bought.
Why it matters and when you do it
Customer discovery sits between having an idea and building an MVP. It is the main activity of idea validation, and when it goes well you reach problem-solution fit: evidence that a specific group has the problem and wants your kind of solution.
For a first-time founder it is also the lowest-cost stage. You can do it in evenings while still employed, and it needs conversations, not capital. The people who care most about the problem often become your early adopters and first customers.
Fonda builds this in. The Validate phase includes an Interviews step, so you talk to real customers before you get a Verdict of go, refine or pivot, with reasons and a next step.
Common mistakes
- Pitching the idea. Once you describe your solution, people react to you and stop telling you about themselves.
- Counting compliments. "Sounds great" is not evidence. Time, money or an introduction is.
- Stopping at the first few yeses. Three friendly conversations can mislead you. Keep going until the answers start repeating.
- Not writing the pass mark first. Without it, any result can be read as good news.
- Handing it to someone else. Founders need to hear customers first-hand, since tone and hesitation carry as much as the words.
Frequently asked questions
How many people should I talk to in customer discovery?
There is no fixed number. Keep going until new conversations stop telling you anything new about the question you are testing. If answers are still all over the place after a dozen or so, your customer group is probably too broad and needs narrowing.
What is the difference between customer discovery and customer validation?
In Steve Blank's method, discovery tests whether the problem and the proposed solution are real for a group of customers. Validation comes next and tests whether you can sell to those customers in a repeatable way.
What if customer discovery shows my idea is wrong?
Then it worked. You have saved the time and money you would have spent building it. Look at what people did complain about, since a better idea is often sitting in your notes, and consider a pivot toward that problem.