Should I give customers what they ask for, or what I believe they need?

You don't ask customers. They don't know. People can tell you what's wrong with what they have. They can't tell you what to make next, because they're stuck in the frame of what already exists. If I'd asked people in 2000 what they wanted, they'd have said a better Walkman. They wouldn't have said, "I want a thousand songs in my pocket." Same with the Mac. Nobody asked for a graphical interface. Nobody came to us and said, "Please make computing visual." They were used to terminals. They would've asked for a faster terminal, a cheaper terminal. We knew better. You have to start with the customer experience and work backwards to the technology. Not the other way around. And that experience — you have to feel it, see it before they do. It's your job to connect dots they can't see yet, between what they struggle with now and what's possible. Focus groups are fine for incremental improvements. For breakthroughs, they're worse than useless. They'll steer you toward mediocrity, because people vote for the familiar. So no — you give them what they need. But you better be right. Because if you're just giving them what you think is clever, and it turns out you didn't understand the need at all, you've built something nobody wants. That's the risk. You're on the hook. That's why this is hard.