“Data Driven Thinking” is written by members of the media community and contains fresh ideas on the digital revolution in media.
Today’s column is written by Michael Lowenstern, Managing Director of Digital Advertising for R/GA.
Around last Christmastime, I was listening to one of my favorite podcasts, NPR’s Planet Money. The topic was – and I paraphrase – “Why Economists Hate Christmas.”
The economist being interviewed explained his argument: When buying something for yourself, you know what you want, you go out and find it, and you exchange money for it. The amount you exchange is generally equal to your perceived value of that item; presumably, you wouldn’t have bought it otherwise. But when buying a gift for someone else, you could exchange $50 for something that has no value to the person receiving the gift – like a toaster or a Power Ranger. That inherent value disequilibrium actually has an economic term: deadweight loss.
Like everyone else probably listening to that podcast, I started thinking about display advertising. It’s a natural leap.
As advertisers, we give people crap they don’t want all the time. We have all of this data to bear: first-party intent data, third-party demographic data, context, location – hundreds upon hundreds of data points. But that data is only used to inform the media components of the ad. Rarely, if ever, is the data used for the creative portion. So, despite our best efforts, we still show people advertising that has no value to them. Why? The ad location may be smart, but the content probably isn’t.
I wanted to prove this at a conference at which I recently spoke. Before the event, I went out with my 13-year-old daughter and bought 10 bags of candy. I asked her to collect a mix of what she thought qualified as “good stuff” and “god-awful stuff.” During the talk, I randomly handed out the candy to 10 participants, and immediately afterwards asked them to rate, on a scale of 1 to 10, how happy they were with what I gave them. Best potential score: 100. Our score: 48.
After rating their happiness, I gave them 60 seconds to trade their candy with each other, and conducted the poll again. The new happiness score: 72. By adding nothing more than choice – i.e. preference data, such as chocolate vs. jelly beans – we improved the perceived value of my targeting (candy lovers).
Simply put, when people are matched with stuff they like, everyone is happier. Can we do this with display advertising? Yes. We. Can. And by doing so, not only do we gain the benefit of happier viewers, but the cost of our media actually goes down too. The question is: How do we achieve this?
An increasing number of publishers and networks are being paid on a performance basis. In other words, they only get paid when a display ad is clicked. The best way to get those clicks is by showing a relevant ad that uses data on our target, but the data that informs the site placement itself only takes us halfway to relevance. The content of the ad should use the same target data for it to be wholly relevant. The content is what makes the ad personal. The content is what makes the ad memorable.
The closer we get to a marriage of media and creative through the partnered use of this rich data, the better our ads get, the happier people are to receive those ads, the more publishers get paid, the more brands succeed.
The moral of this story: Speak to each user like an individual, one who doesn’t just like candy but who specifically likes gummy bears and doesn’t like circus peanuts.