Exercise a Bit of Control
It’s inevitable. Just when you’re about to celebrate an ad campaign that drove sales up, some naysayer will ask, “How do we know that sales wouldn’t have gone up anyway, even if you hadn’t run the ads?”
Resist the impulse to call the naysayer a … well, a naysayer. The truth is, it’s a fair question. It is entirely possible that sales went up due to some factor besides, or despite, advertising. I have pointed out before that gas mask sales rose dramatically about this time ten years ago. I doubt that gas mask ads had much to do with it.
So, how can you know? Set aside a small but representative, random sample of your market and insulate them from your advertising. Track this “control” group’s purchasing behavior along with that of those who are exposed to the ads. If the control group buys less, you can be pretty sure that the advertising led the larger group to buy more. If both buy at the same rate, your ads probably had no effect at all. If the control group buys more, your ads may be driving sales down.
The second and third possibilities occur more often than many of us care to know. This may be one reason why so few ad agencies track ad performance, much less use control groups. It’s easier to keep a campaign sold when the client can’t tell whether or not it’s accomplishing anything.
—Steve Cuno