The issue here isn't so much over or under-engineering, but rather "Are we building the right thing?" or "Are we building the thing right?"
In a startup you don't know if you're building the right thing so trying to build it right is premature optimization. Since you have limited resources you really have to focus on ensuring that you're building the right thing...if you aren't it doesn't matter how well designed or built it is no one is going to use it.
Once you've validated you're building the right thing you can start focusing on building it right, but by they you probably know where the pain points are and where you need to spend the effort.
The trick with this though is that everyone needs to be aligned on that approach and be honest about the fact that corners may be being cut to get something working fast. Where I've seen this go badly is where a crappy initial version was built to get to market fast, but then no time was made available to address the defencies in the initial release.
In a startup you don't know if you're building the right thing so trying to build it right is premature optimization. Since you have limited resources you really have to focus on ensuring that you're building the right thing...if you aren't it doesn't matter how well designed or built it is no one is going to use it.
Once you've validated you're building the right thing you can start focusing on building it right, but by they you probably know where the pain points are and where you need to spend the effort.
The trick with this though is that everyone needs to be aligned on that approach and be honest about the fact that corners may be being cut to get something working fast. Where I've seen this go badly is where a crappy initial version was built to get to market fast, but then no time was made available to address the defencies in the initial release.