Nothing kills good products faster than brilliant conversations with the wrong people.
You're at a tech conference, at your booth, offering free swag and a demo to anyone peaking longer than 10 seconds at your rollup banner. You start up chats about your product with passers-by. They share some pain points and suggest features. The "it would be great if it did this" type of thing.
After the conference, you rush back to your team with these brilliant "user insights" that you claim came from real conversations.
Here's the thing.
That wasn't user research. That was networking plus confirmation bias.
Why?
Conference attendees aren't your typical users. They're early adopters, industry insiders or people with enough time and budget to attend events. They're probably more tech-savvy, more forgiving of rough edges and definitely more engaged with your industry than your actual user base.
The conversations happen in a specific context. People are networking, being polite or trying to impress. They're not using your product in their messy, real-world environment while juggling three other tasks and dealing with a kid screaming in the next room or their boss breathing down their neck.
People you talk with at conferences rarely expose their messy lives. They'll rarely disclose their wrists hurt and need a special keyboard at the end of their work day. They won't mention they can't tell red and green apart. Or that they use screen readers. Or that loud autoplay videos trigger their anxiety. Conference small talk doesn't cover the fact that someone's arthritis makes tiny touch targets impossible. Or that they rely on captions because they're in a noisy open office.
These aren't details people casually drop into networking conversations. But they're exactly the details that determine whether your product actually works for the millions of users.
The problem is conference conversations feel so insanely authentic that you skip any proper user research afterwards. I mean, why should you run usability tests when you already "know" what users want?
It doesn't feel like a shortcut, but it is. And it leads to products built for conference-goers instead of real users. Features that sound brilliant in a booth fall flat in real world scenarios.
It'd be better to use these conference conversations as inspiration, not instruction. They're great for generating hypotheses. But you still need to test those assumptions with actual users in their actual environment.
Real research means finding people who represent your users, not just the ones who happen to cross your path.