Wired has an article from their February edition called The Ad Scientists. It's a neat look into some folks who are developing sophisticated ad targeting and tracking systems for TV broadcast and cable advertisements.
The part I found most interesting, which relates to the recent discussions about ad agencies, featured the creators' comments about "Madison Avenue," ad agencies and the fear that performance-based metrics instills in traditional media marketers:
..."On the Internet, marketers love their dashboards, their control panels, the ability to see results and to make changes based on those results," Haberman says. "When you look at offline advertising research, it's like going to the morgue. They cut the guy open and tell you why he died. But that's worthless unless you can make a change. The real opportunity is in coordination and feedback."...
...there's no mistaking his revolutionary zeal as he rails against Madison Avenue. "We've seen every major agency in the United States, from New York to Chicago to Los Angeles," he rants. "The same people who make speeches at conferences saying, 'We need better measurement' - they don't want the data. They look at me like the guy who invented the report card."...
...In a clear sign of the disagreement between clients and Madison Avenue over the need for new ad metrics, most of the companies didn't tell their agencies about the test...
Incidentally, there's a meetup tonight for the Ad Club Seattle, but I just don't have the heart to try to talk search marketing with ad agency types. It feels like a lost cause.
"When you look at offline advertising research, it's like going to the morgue."
Its not just the research.
Given an equivalent amount of money in radio, newspaper or TV advertising, the internet blows the others away.
Each avenue is not exclusive however, and getting the combination where one supports the other is the key to good conversions. Example, TV to website.
The metrics are always going to be strongest and most believeable on the internet, because the process is interactive and there is more data to measure and optimise with.
It was nearly a decade ago when I worked for the #2 player in television ratings, and personally knew the lead engineer working on a plan to create small, wearable devices that would recognize television and radio, and also recognize the stores where you shopped. Tie that to receipt analysis at the big box stores and you have yourselves some darned fine performance metrics.
But the heebie jeebie factor was high (big brother to the Nth degree), as was the cost of development (and projected cost of acceptance of big box - they'd need to buy some hardware to make this thing fly), so it never really took off.
Darn shame.