In a memorable scene from Mad Men, Season 3, Conrad Hilton comes up with a print ad concept when he’s considering firing his agency. The folksy concept, complete with illustration: “When the country mouse comes to the city.” He wants free advice from Don Draper.
Don smokes, squints, and thinks. He says simply: “I don’t think anybody wants to think about a mouse in a hotel.”
Why would someone associate actual rodents in the hotel with a folksy cartoon character and a totally different message? It doesn’t matter. As the quintessential “ad man,” Don instinctively knows that in this category, rodent imagery is forbidden. It takes him two seconds to dismiss Hilton’s effort.
Don, of course, knew that advertising was often about how the “meta message” — how people respond to things the ad doesn’t literally say, but conveys — affects positioning and user response. Yes, it’s unscientific. But it’s important to know that insofar as you’re testing advertising “messages”, you don’t have the luxury of testing on one dimension only — the literal content of the copy and how “persuasive” it is. Anyone who has done multivariate landing page testing online knows that the persuasiveness of the actual copy is only one element. Even within the copy, there are aspects that position or suggest more than they convince or persuade.
That logic needs to be considered for search advertising too, even though these seem to occupy a very small space. And it works, if you practice.
If you didn’t catch my short piece in ClickZ about the metamessage in paid search, it’s here.