Glossary

What is customer discovery?

Customer discovery is the practice of talking to potential customers before building a solution.

It tests whether the problem you believe exists is real, who has it, and what they currently do about it.

The term comes from Steve Blank's Four Steps to the Epiphany, where it is the first step of customer development. The core discipline is asking about the customer's past behavior rather than their future intentions: what they did last time the problem occurred, not whether they would buy something that solved it.

Why it matters

The most expensive mistake in product is building something well that nobody needed. Discovery is cheap by comparison: a dozen conversations against months of engineering. It is also the only part of the process that can tell you you are wrong before the cost of being wrong is sunk.

Example

A founder plans to build scheduling software for clinics. Twelve conversations reveal that clinics do not experience scheduling as a problem. Billing reconciliation is the pain, and scheduling is where the errors show up. The product changes before a line of code is written.

In UpSight

UpSight ships a Customer Discovery lens that reads each interview for problems, workarounds, and what breaks, with a citation on every field. Interviews, survey answers, and notes attach to the people who gave them, themes regroup after each new conversation, and your research questions show what is answered and what is still open. The twelfth interview lands on top of the first eleven.

Related terms

FAQ

What is customer discovery in simple terms?
Asking the people you think have a problem what they actually do about it today, before you build anything.
How many customer discovery interviews do you need?
Most teams start seeing repeated patterns between 12 and 20 conversations within a single segment. Fewer than about 8 rarely distinguishes a pattern from a coincidence.
What's the difference between customer discovery and validation?
Discovery looks for what the problem is. Validation tests whether a specific proposed solution addresses it. Doing them in the wrong order is the classic failure.