Avoiding premature optimization involves looking at the whole situation, instead of the optimization.

* Working at a strategic level is often nebulous and requires some definition.

* Little changes are simple, concrete, and instantly rewarding (if only I could do X, this would work better).