This year I experimented focusing on partner first - I went through a list of good friends and bug them one by one with a different idea, tried to motivate them, and see which one kept the work going. So far I have 4 projects that survived (meaning are still worth investing time) out of a 10th or more.
I'm not interested in building a new startup now, though, that's why I did this specific experiment.
Experiments aside, in general I think you should focus on problems and how quickly you can get the first $ with your idea/solution.
So you list 10 ideas, you look at the core problem each tries to solve, think if you like to work on that problem/space (not the current solution, because you may change solution down the road), try to figure out who could pay you first, and estimate how quickly you can get something working which is worth your first users to pay for.
I'm not interested in building a new startup now, though, that's why I did this specific experiment.
Experiments aside, in general I think you should focus on problems and how quickly you can get the first $ with your idea/solution.
So you list 10 ideas, you look at the core problem each tries to solve, think if you like to work on that problem/space (not the current solution, because you may change solution down the road), try to figure out who could pay you first, and estimate how quickly you can get something working which is worth your first users to pay for.