What If Targeting Doesn't Work?

As regular readers know, I have a basement full of oddball ideas about advertising. One of them is that "media science" may be a lot of hooey.

What if all the "precision targeting" we do is mostly unnecessary complexity masquerading as knowledge?

What if there's only one important cut we need to make when planning media -- does the person participate in our category or not?

If we're selling golf balls, the only important targeting question we have to ask is, "Does this person play golf?" If we're selling wine the only important question to ask is, "Does she drink wine?" If we sell tires the only important question is, "Do they own a car?"

All the other stuff -- their education, their income, their weight, height, and serial number, their zip code and psychosexual predelictions, the websites they visited yesterday, and the number of chickens in their backyard -- may be interesting, but what if they don't do a damn thing to make our media buys more effective?

During my semi-brilliant advertising career I would never have suggested such a thing to a client. Clients don't like oddball ideas. They are resolutely devoted to believing what everybody else believes. And everyone else believes that leveraging data to create precision targeting is the future of advertising.

I guess it would be simple enough to either prove or disprove this theory.

I'd love to see an advertiser do a split run. In one market buy media based on the usual  demographics, psychographics, data-o-graphics, programmat-o-graphics, graph-o-graphics, and bullshit-o-graphics.

In another matched market run the same campaign but make the media buy based on one behavioral criterion -- does the person participate in our category or not?

I'd love to see the results.