How do we solve the offline purchase problem?
The Adtech conference a couple of weeks ago was pervaded by two smells. One was the sniff of a slowly decaying traditional advertising world, the other was the smell of adolescent exuberance from a bunch of precocious but brilliant teenagers.
Because of this the conference was brilliant.
What has stuck with me is the cleavage between traditional advertising and new generation performance marketing.
The problem with traditional advertising is that performance metrics are woolly. That’s not to say it doesn’t work. We still see TV adverts for soap powder because of course there is a correlation between offline ad spend and revenue.
What the traditional admen didn't want to address was what will happen when web offers the same reach as tv. They seemed to retort that ideas like a monkey playing the drums will save offline advertising. I am not so sure.
At one level this is an online offline debate, but on another level this is product centric.
You can sell some products, like cameras online, and so using performance metrics like cpas is easy.
But what sort of performance ads can you use if you're Cadburys or Pringles?
How can you prove an online ad makes more people buy chocolate? Its hard.
Its not just FMCG. For example, what about cars? You can have an incredibly compelling piece of advertising that makes someone more likely to buy a BMW.
But they are going to buy it in a dealer.
Why should it matter that offline ads are hard to measure?
When you can’t measure the effectivness of an ad it should worry admen. Advertisers can get measurement elsewhere. They will go elsewhere.
It also worries small publisher sites; what if you have created a niche car blog with the perfect environment to persuade people to buy cars? You can't currently get the 5% lead-gen fee. But by god that’d be nice.
As an investor I am fascinated by this broken piece of the chain. How do you solve the problem? Vouchers? More sophisticated lead gen intermediaries?
I'm looking forward to entrepreneurs telling me.