โ† Back to methodology overview
๐Ÿšช

Customer Discovery

Steve Blank ยท steveblank.com

Steve Blank's foundational insight changed how startups are built: no business plan survives first contact with customers. Before Blank, the standard approach was to write a plan, raise money, build the product, and then find customers. The failure rate was enormous. Customer Discovery flips the sequence. You start by talking to customers, test whether your hypotheses about their problems are correct, and only then start building.

Customer Discovery is the first of Blank's Four Steps to the Epiphany: Discovery, Validation, Creation, and Building. In the discovery phase, your only goal is to answer one question: do real people actually have the problem you think they have? Not "do they like my solution?" Not "will they pay?" Just: does the problem exist, and is it painful enough that people are actively trying to solve it? This is harder than it sounds, because founders are naturally biased toward believing their own hypotheses.

The process involves structured conversations with potential customers, but it's not market research in the traditional sense. You're not gathering data to inform a report. You're testing specific hypotheses and updating them based on what you learn. Each conversation should change what you believe, even if only slightly. If you finish ten conversations and your assumptions haven't changed at all, you either had perfect intuition (unlikely) or you weren't really listening.

How Distil uses it

Customer Discovery is most active during the deep dive phase, where Distil assesses the strength of your market understanding. The critical questions become: have you spoken to real people who have this problem? How many? What patterns have emerged? What surprised you? If you haven't talked to anyone yet, that's a significant gap.

The framework feeds into all three scoring dimensions. For Problem Clarity: can you describe the problem better than the customer can? For Solution Potential: have customers confirmed that your proposed solution would actually help? For Idea Maturity: have you reached pattern saturation, where you can predict what the next interviewee will say?

Distil also uses Customer Discovery principles to check for pivot signals. If the people you've spoken to keep describing a different problem than the one you expected, or they care about different features, or a different customer segment responds more strongly, these are signals worth paying attention to. Distil will flag these when they appear in your answers.

Key principles

  • Talk to real people, not proxies. Industry reports and competitor analysis are not customer discovery. You need direct conversations with people who have the problem.
  • Seek patterns across conversations. One enthusiastic interviewee proves nothing. Five people describing the same problem in the same way is a signal worth acting on.
  • Your first idea about who the customer is will probably be wrong. Stay open to discovering that the real customer is someone you hadn't considered. Follow the evidence.
  • Distinguish between "nice to have" and "must have." If customers say "that would be cool" but haven't tried to solve the problem, it's a nice-to-have. Must-haves come with stories of failed workarounds.
  • Discovery never truly ends. Even after launch, keep talking to customers. Markets shift, needs evolve, and new jobs emerge.
  • Hypotheses are meant to be updated. Write down what you believe before each conversation. Update it afterwards. If your beliefs never change, you're not learning.
  • Get out of the building. You cannot discover customers from your desk. Surveys, analytics, and reports are supplements to conversation, not replacements for it.

Common mistakes

  • Treating discovery as a checkbox. "We talked to 5 people" isn't discovery. Discovery is understanding why those people behave the way they do, what they've tried before, and what would make them switch.
  • Only talking to early adopters. Early adopters are enthusiastic about everything. They're easy to find and easy to impress. But they're not representative of the broader market you'll need to reach.
  • Confirmation bias in interview selection. Choosing to interview people you already know will be supportive, or stopping interviews once you get the answers you wanted, defeats the entire purpose.
  • Not documenting findings. Memories fade and distort. Write up every conversation within 24 hours, using the customer's actual words, not your interpretation of them.
  • Skipping the "what have you tried?" question. If someone says they have a problem but hasn't tried to solve it, they're describing a mild inconvenience, not a real pain point.

Recommended reading

  • ๐Ÿ“– The Four Steps to the Epiphany by Steve Blank. The book that started the lean startup movement. Dense but essential.
  • ๐ŸŒ steveblank.com. Steve Blank's blog with years of articles on customer development and startup strategy.

Ready to discover whether your customers are real?

Distil brings Steve Blank's rigour to a 10-minute conversation about your idea.

Test My Idea