Today I met with another early-stage founder, and the topic of customer discovery came up. He said he’d done discovery, so I dug in and asked more questions. Two things stood out to me that we discussed in detail.

First, most of his customer discovery came from sending out surveys. (I shared my thoughts on surveys as customer discovery earlier this week—see here.) We discussed the benefits of doing more one-on-one conversations.

Second, a lot of his insights were based on a conversation he’d had with one customer. He’d built a lot of his product offering and website messaging in response to what he’d heard in this conversation. Learning from a conversation with a potential customer was a big plus. But one customer doesn’t confirm a pattern that leads to previously undiscovered insights.

Surveys are good, but not for initial discovery. Conversations are the best way to learn about pain that potential customers are experiencing. And you need lots of conversations with lots of people to uncover patterns and gain a deep understanding of problems and why they’re so painful.