That is a fascinating line of reasoning. I wonder if there is some way to leverage historical data (for other brands, maybe?) to figure out potential local optima.

I original thought this was a nice "set it and forget it" marketing scheme, but your comment makes me think otherwise.