I’ve always followed Paul Graham’s advice: “live in the future, then build what’s missing.” If you’re really good at a few different things, and smart enough to learn how to do everything else, then usually the optimal startup idea is somewhere in the middle of that space. For me I’m a good programmer and I love learning languages, so I started building tools that I always wanted but no one else was selling. Once I got a prototype and some users to get feedback from, everything began to slowly snowball from there.
What’s important to clarify is that I’m not building exactly what I had in mind when I first dreamt all this up long ago, but my first idea would never have worked anyway. The best thing I ever did was just launch, because your users (or lack thereof) will quickly prove which ideas are good and which are not. So my advice is just pick the best idea you have right now. Or if you don’t have any idea just solve a problem you or your friends have. Then launch as soon as you can and pivot, pivot, pivot. It’s always going to be an iterative process.
What’s important to clarify is that I’m not building exactly what I had in mind when I first dreamt all this up long ago, but my first idea would never have worked anyway. The best thing I ever did was just launch, because your users (or lack thereof) will quickly prove which ideas are good and which are not. So my advice is just pick the best idea you have right now. Or if you don’t have any idea just solve a problem you or your friends have. Then launch as soon as you can and pivot, pivot, pivot. It’s always going to be an iterative process.