• Most companies overweight TV and print versus underweighting their online ad buys when you look at influence and interaction. Whole presentation could have been built around that sentence.
  • It’s good to hear the emphasis being placed on social media and the changes it represents from the old internet world. As Armano said, the more people that get this the better, especially at the C-suite level.
  • Behavior advertising doesn’t work in a Web 2.0 world. Really? His idea is that the content is not necessarily complimentary so it leads to some disconnects in the user experience. It has to, he says, include demographic targeting.

OK, let’s look at that a bit. Demographic targeting may, as he says, cost less because you’re not paying per click but by impression. That’s true. But if the ratio of cost-per-conversion is staying the same even if the overall dollars are shrinking, then is it really a more valuable solution?

The idea, and it has some merit, is that hitting customers with increasingly relevant ads for the same company will lead to their eventual relenting and signing up/buying the product/whatever the desired end result it. I have read things that say running ads on UGC actually helps those ads because some of the authority of the content is transferred to the ads. So if the ads that someone sees when they get to a blog or whatever are better because of that, that’s a good thing. It’s all about – or at least it should be all about – providing a better experience for the end reader/viewer.

So yes, online advertising needs to get a bigger slice of the pie. Kirk Skodis made similar points when discussing entertainment marketing spending. And that slice will increase as effectiveness increases.

Some more bullets:

  • The number of people sharing content wound up being surprisingly skewed toward females when they were expecting young men, the stereotypical online crowd.
  • The day is coming when online ad managers will be able to plan their campaigns based on a single set of data and then execute across channels. Not sure I completely agree with this but it’s worth considering.
  • Someone asked about the fact that targeting by demographics is just replicating the old media model since not everyone in a particular group shares characteristics with other members. But Johnson points out that while it’s not perfect it’s a decent predictor in the absence of other data.

This was very interesting, especially as it concerns targeting and data mining. Those are important issues that need to be discusses at length before advertisers dive in with a poorly thought out plan.