To ensure high quality data science project are produced by your team, incorporating code reviews in your workflow is a must!
I remember when I first started doing research, I always felt bad and stressed whenever I generated a negative results. My thought process was that I must have messed up something along the way and that I wasn't good enough.
All assumptions, especially those underlying the project creation, should be documented and fact checked!
I've never seen a research / data science project that went exactly as planned upfront. Projects either change incrementally, with big pivots or fails.