You have a location in Denver and one in Boulder. Buy a billboard outside of each town, but outside of Denver put up a PSA and outside of Boulder put a real ad.

You have a location in Denver and one in Boulder. Buy a billboard outside of each town, but outside of Denver put up a PSA and outside of Boulder put a real ad. Look at revenue growth between stores 1 year later and you can see the impact of a billboard. Oversimplified of course, and this doesn't work with 2 locations, but with 200 you can start to make more solid conclusions. (I use PSAs to test digital marketing so I'm sure its different, but you pretty much just use PSAs as a control group)