Yep, that's the approach that worked for us: we started a consulting company and built up a war chest to self-fund the initial development of our "real" idea.

Yep, that's the approach that worked for us: we started a consulting company and built up a war chest to self-fund the initial development of our "real" idea. Rather than pulling the trigger as an event, it was a process: we funded some prototype work - paid ourselves to prove out some things. Then we moved one of our devs (of 5) over to it full-time. As it started to take off, we began to wind down things on the consulting side while simultaneously moving more and more resources over to the new thing, until we were all over. It's always scary, and there's always a point where you just have to jump in, but this way made the risk feel much more manageable.