In contrast, splitting the day helps to pivot more easily. I've had situations where I thought I wanted to work on a certain feature but when I talked about it, people didn't want that feature. So if I spent a week implementing it, it would have been a waste of my time.

Reverse your process. Talk about it to people first, then work on a feature they need (in aggregate).

I think I would prefer to code for the first half of the day, and market for the last half as my concentration wanes.

I have tried that in my work at my startup and failed. Alternating days did not work either. Jon's recommendation about alternating weeks is what I am going to try next.

Interesting, what went wrong when you tried that? I'm expecting to be in your shoes soon, and I currently have to balance a part time job with writing the software for my startup. I'd love to learn from your experience.

This works for larger companies and larger teams, too.

This is very interesting IMHO !