Before I build anything, there are four questions I ask myself:

1. Who is it for?
2. What do they get out of it?
3. What does the ideal result look like for them?
4. Would I enjoy building it - every step, including the failures, the debugging, the rethinking?

The first three are about the product. Get them wrong and you can fix them later. The fourth is about me, and it’s the one that decides whether I’m still working on this in six months.