This week I listened to the founding team of an early-stage startup describe their current traction. They’ve built a software product and have thousands of (nonpaying) users. To determine what features their customers would pay for, they sent them a survey. Twenty-five percent of the people who responded said they’d pay for a particular feature. So, the company built that feature and launched it recently. The results weren’t good. Only three (yes, three) customers bought it.

So, what went wrong?

The team surveyed customers; they didn’t talk to customers. Talking to customers involves picking up the phone or meeting a customer in person (one customer at a time is ideal). A conversation, if you aren’t leading them, will allow you to dive deeper into the customer’s thought process and experience. It’s iterative. You hear something you weren’t expecting (or didn’t know) and ask questions about it. That leads to something else you didn’t expect, and you ask more questions. The result of that loop is new insights about the customer’s problem. Do that with multiple customers, and you start to see a pattern. You now have a better understanding of the problem and how to solve it for the customer.  

What I described can’t be done with surveys. Surveys may have biased questions that don’t help you understand your customers. And they foreclose iterative interaction with your customers. Sure, they’re efficient and allow you to get a lot of feedback quickly, but the quality is often low, and it can lead you down the wrong path.

The lesson this early-stage team learned was to stop doing surveys of their users and start getting them on the phone. Conversations lead to insights. You can’t have a conversation in a survey.