Before you build anything, find out what people actually need. It sounds obvious. But it's easy to skip, because building feels like progress and asking doesn't.
The cost of skipping it shows up in the post-mortems. In 2014 the research firm CB Insights went through 101 post-mortems written by founders of failed startups and counted the reasons they gave. The most common, cited in 42% of cases, was building something that didn't serve a market need. One founder summed it up: "I realized, essentially, that we had no customers because no one was really interested in the model we were pitching."
The problem in cases like that usually isn't the building. It's who the thing was built for: a customer who existed mostly in the founders' heads.
The fix sounds easy. Ask people. The trouble is that people lie to you, and not on purpose. Tell someone your idea and they'll say it's great, because they like you and being nice costs them nothing. That's why Rob Fitzpatrick called his book on customer conversations The Mom Test: your mom will tell you she loves your idea. The book promises to show you whether your business is a good idea "when everyone is lying to you."
Fitzpatrick's way around it is to stop asking about your idea and ask about their life instead. Not "would you use this?" but "when did this last happen to you, and what did you do about it?" Opinions about the future are cheap. Facts about the past are expensive, and they're the ones that tell you whether the problem is real.
I learned this at Basis, the first startup I co-founded, where Fitzpatrick mentored me. We were turning wearables data into health insights for busy professionals, with neuroscientists from Duke and a team from Harvard IACS behind the algorithms. That's exactly the kind of project where you're tempted to build first, because the science is interesting and the team is good. Instead I went and found our users myself, in gyms, in corporate offices and in meetings with executives.
Asking is slower at the start. It's faster overall, because you build fewer things nobody wants.
Building is the expensive way to find out you were wrong. Asking is the cheap one.