Ad Age readers might have caught an article a couple of weeks ago about Chrysler and Organic’s predictive media spend modeling. Quite frankly, I’m surprised it didn’t generate more discussion. The article told the story of how Chrysler (who during the conference told the audience about its culture change toward accountability) challenged its agency, Organic, to come up with a predictive econometric model that could help optimize media spending for Jeep. Organic’s Steve Kerho and Chrysler’s Chuck Sullivan presented the case study in more detail for the Ad Age Digital Conference.
Hopefully, the presentation will make its way online via video soon. It was easily one of the most impressive of the conference. The model took into account a variety of marketing spend, online data, auto-industry and macroeconomic factors (too many to list here). It was presented in July 2008 – pre-financial Armageddon, mind you – and ended up predicting total U.S. 2008 Jeep sales within 1% of the actual sales generated. Generally automakers are happy to predict within a range of 10-15%, according to the Ad Age article. Incredible. It remains to be seen if this system can deliver a repeat performance in 2009, but the lesson learned here is that the data is out there and should be harnessed.
Another impressive item from the Chrysler presentation was that they created their own media attribution model with 11 total user touchpoints, including TV, search and display, and assigned a dollar value to each one. Included in this model is, “viewed a display ad.” As we’ve discussed several times on this blog, attribution should be looked at more closely by marketers who want to maximize results. The last-clicked-wins model doesn’t tell you anything. Count Chrysler as being on the leading edge.
Day 2 also included a presentation by Facebook COO, Sheryl Sandberg, called “How Many Friends Can You Have”? This presentation received a lot of coverage online, so I won’t do a full recap. The most interesting takeaways for me were the highlights of a few successful ad campaigns on Facebook. It was at this point during the presentation that I noticed none of the campaigns featured display advertising.
During the Q&A, someone asked her about the commonly held notion that advertising on social networks does not produce a good ROI. Here was her exact quote:
“Social media has had to do some evolution, some work to come up with the right ad products, and we find that we are really first on that path now,” she said. “Banner ads that interrupt your experience, or text ads, we don’t think work as well in this environment. It’s actually just in the last year that we were able to launch ads on our site that behave the way the rest of our site behaves.”
Bravo to Sheryl and Facebook for staying on the path of innovation. Banner ads on social networks do not maximize the potential of the medium, particularly on a site like Facebook. Here’s a further explanation of the direction that Facebook is heading from paidContent.org for their advertising opportunities.
Marketers should look at advertising on social networks with standard display ads with a very critical eye. If the #2 at Facebook says they don’t provide the optimal ROI, why buy them in the first place?