After a pair of lunchtime presentations by sponsors Doubleclick and Appnexus, the conference reconvened to discuss data and metrics, two areas of high-level discussion this morning, in greater depth. Anne Hunter, VP of Advertising Effectiveness Products at comScore discussed a report on performance metrics, noting that online advertising campaigns have long been measured by those metrics that are easiest to use. Following her presentation, a panel convened to discuss data and how advertising agencies are using it. Data is the new media, the new currency, and the new commodity, according to the emcee, and it was the goal of the panel to make data “passionate”.
Panel – Valuing the Data: How Can Agencies Evaluate and Economize on the New Data Layer?
Steve Ennen, Managing Director of Wharton Interactive Media Initiative (moderator)
Tousana Durgan, Associate Media Director at MEC
Loren Grossman, Global Chief Strategy Officer at RAPP
Anurag Harsh, SVP of Strategy and Business Development at Ziff Davis, Inc
Jeff Hirsch, CEO of AudienceScience
Michael Lampert, VP & Group Director of Media at Digitas
There is no limit to the amount of data an agency can use, according to Lambert, especially when creative a predictive model. The question is whether the cost of data is worth the incremental improvement. Grossman is in agreement, saying that agencies would prefer to use “all of” the data available to them. He is looking for new proxies to reduce variation due to time-series changes.
Durgan explained that there aren’t a lot of guidelines when it comes to which data to gather and use. Often there is a lot of duplication, and it’s hard to know which layer of data is better or more efficient. Grossman sees this as the biggest barrier to getting good results from existing data. Until there is a consistent, transparent method for rating the data, it’s often critical for an agency to source data on its own. He warned that there is no way to drive behavior without driving some form of brand engagement. One cannot live on data alone.
The problem, Hirsch believes, is at the top of the funnel, where awareness is generated. The method of filtering data in order to create the relationships that are necessary for engagement is unclear. Lampert reminded the audience that you need different success models throughout the funnel. Activity at the top of the funnel will affect the lower part of the funnel as engagement is deepened.
There are four fundamental aspects that define the quality of the data, according to Harsh:
- The data must be in depth
- The data must be in market
- The data must be in volume
- The data must be in time
You need to be able to collect data from multiple sources, and it must have some relationship to the buyer. In regard to volume, a small pool of cookies in insufficient. Often, it’s a case of garbage in, garbage out. Finally, the data has to be collected, scored, and used in real-time in order to be accurate.
Ennen asked whether there is a difference between data and metrics. Durgan noted that one has to measure the value of the data (data about data: pretty meta, huh?) Lambert is unsure whether the market is mature enough to know how much different points of data are worth. At some point, you have to look at data as what it is, a purchase. Hirsch notes that using data is not a simple process, as you can’t compare apples to apples. However, when one can really understand the data, then there are tremendous returns for the user.
Ennen asked who is responsible for educating whom, and who do you trust? Durgan explained that first party data is the most trustworthy. As far as third party data, the key is transparency. It’s best to find a provider with whom you’re comfortable and partner closely with them. Grossman recommends studying, learning, and then creating your own intellectual property around it.
Harsh commented that publishers likewise don’t know the value of data on their site, and they often risk theft of their data by widget providers. If publishers understand how to monetize their data, then they can create an additional source of revenue. Hirsch has undertaken such a project with publishers and reiterated the importance of transparency when working with partners.
Ennen called privacy “the biggest concern that no one’s concerned about.” Privacy isn’t a legal issue to Hirsch, but an ethical one. Self-regulation guidelines are critical to protect consumers and create trust. Grossman notes that reputation and brand issues will arise from privacy breaches long before legal issues arise. Durgan notes that agencies also have to respect client data in the same way.