With every new "must-read" survey predicting the future of online advertising, collective eyeballs roll from Madison Avenue to Silicon Valley. By now, we all know that the industry is evolving; the paradigm shift has moved consumer acquisition strategies away from purchasing content as a proxy to reach consumers to buying very targeted audiences who have explicitly shown an intent or affinity (yes, we all get this).
The general consensus is that the companies best positioned to be able to do this at scale are ad networks. The real question is in the execution. How do we fulfill the empty promise left behind by the first generation of behavioral targeting platforms and deliver advertiser ROI at scale?
To answer this question, we need to look at where first-generation networks went awry. The first-generation behavioral targeting platforms were focused primarily on building proprietary data-mining tools. A lot of these companies engineered monolithic platforms that could consume billions of data points in a uniform fashion guided by a unique taxonomy. The issue is that data mining competency is inherently a very different business practice from successful campaign execution. The reasons companies have had difficulty delivering precise audiences at scale are three-fold:
1. There's no silver bullet. No single source or type of data can be all things to all clients.
2. Marginal utility. It is cost-prohibitive to scale data mining practices due to high infrastructure and human resource costs and low utilization rates of mined data.
3. Pricing. The value of data is difficult to calculate, which makes the pricing of data complicated.
Therefore, the question is this: How can the next generation of networks deliver specific audiences at scale in order to drive ROI?
The answer is as much about supply chain management as it is about effective media sales. As the industry evolves and fragmentation accelerates, the approach networks must take to maximize advertiser ROI hinges on value creation by way of aggregation rather than proprietary data mining. Networks must understand the value associated with each segment they are purchasing. By understanding how each data source and type impacts a campaign, networks can drive scalable ROI for advertisers by improving efficiency and minimizing waste. Proper pricing decisions must be made via effective bid management across multiple data providers in order to scale.
Networks should take a page from their SEM counterparts. They need to avoid the "win at all costs" mentality when purchasing data from third-party vendors and focus on identifying the optimal combination of data and inventory that drives advertiser performance.
Michael Katz is president and founder of interCLICK.
On Twitter? Follow iMedia Connection at @iMediaTweet.
