Nah because we care about the size of the effect too.

We can get a p<0.0001 finding of an otherwise terrible, annoying change that saves 1%.