In our case, the goalposts for what the product should be kept moving every 3 months, which was probably the largest single problem, it caused most of the friction imo. It didn't help that we were a hardware startup, so changes couldn't be made as easily. A solution to that would be to agree upon (and document) what the MVP is, and not change course until you've validated that the idea doesn't meet the market's needs. We had a single client/partner, their desires ended up driving a lot of changes late in development that ended up being incompatible with earlier engineering decisions, but the founder pushed them through. We ended up not being able to do environmental testing because of the time those changes took, and the first time it rained the cases for our electronics filled with water.

I'm also now a solo-ish founder and I only have technical/team management experience, so I'm learning too :)