From Paul Feldwick's The Anatomy of Humbug:
"So theories of how advertising works, however implicit, have an impact on power relations within an agency, and between the agency and the client. A theory which is based on intuition and taste will privilege the creative department; a theory which is based on psychological insight may give power to the planning department; a theory based on measurement of message recall gives authority to the researchers and thus to the client."
I have always found the advertising industry to be fascinating. Primarily because "ad men" are practitioners of persuasion at a mass scale.
Philosophers, psychologists, and neuroscientists study human motivational structures in closed systems (out of academic necessity). But advertisers, they have to craft and deliver messages in a messy, chaotic, correlation-isn't-causation world.
And whether they are successful or not, well, who knows? Did that guy buy Trident because of the TV ad or the Facebook ad; or because his friend shared a stick of gum with him once, and the packaging looked nice; or because the gum was very conveniently located in the checkout aisle?
Consequently, the tension in the advertising industry between "creativity" and "measurement" is palpable.
Clients fire agencies for "ineffective ads," even if the data being used to determine that is incomplete and misleading.
Meanwhile, agency creatives fiercely defend their claim to intuition: "Persuasion is an art, not a science." And ineffective ads are thus made.
Perhaps art is simply the science of the things we don't know.
For things we don't know, I'd take a principles-driven tastemaker like Steve Jobs or Ira Glass any day. I'll step back and let you do your magic. You clearly get something the rest of us don't.
For things we do know, though, "art" isn't art if it isn't delivering the desired results.
In the case of advertising, the measurement geeks overmeasure and tortured artists overtorture. Maybe the same dynamic exists in every industry.