• by Adam Gitlin , September 25, 2012

Historically, the advertising industry has built services on top of categories of data…for example, competitive services on top of competitive data.  CRM services were built on top of customer profile stores.  Marketing mix on top of media spend data.  Even more recently, we have built “social” services on top of “social data” (quoted because of the enormous and undefined scope of data falling under the umbrella of that label).

However, this paradigm—the “data-centric” approach—no longer makes sense.  It limits our insights, and it suffocates our imagination.

It first restricts our ability to appropriately analyze data as we all know we should. For example, when examining campaign performance, an analyst would first sit down with the goal of answering “how did my campaign perform, and why?”  It would be difficult to argue that an analyst would not be better equipped to answer this question if they possessed access to not only ad-server data, but also buzz, word-of-mouth, website behavior, audience segmentation, engagement value, and customer LTV.  However, the traditional approach to building services—with respect to specific data sources—limits the analyst’s access to this data.  Marketers are reticent to provide media agencies with access to web analytics, but will gladly provide it to the agency responsible for web development.  A social agency may “own” the administration, communication, and optimization of a Facebook page, and fend off attempts from other agencies to access profile page visitor, post, and “like” data.  Outside of “territorial” considerations, there are often challenges of access—a media client may be tasked with reach and savings, and thus not have access to an LTV analysis managed by some other internal department, because LTV does not factor into the client’s remit.

In addition to restricting an analyst’s ability to answer questions, the “data-centric” approach also places artificial, unnecessary, and limiting barriers on the type of questions that can be asked.  This is perhaps the more dangerous, as it can quickly spiral into a self-fulfilling cycle of progressively less and less inquisitive and deep analysis.  Let’s take the same question of “how did my campaign perform, and why?”  If an analyst knows that she can only have access to ad-server data, she will be highly limited in the quality and sophistication of analysis.  Pretty soon, the analyst will stop asking questions such as “Which audience segments are driving performance?”, “Are we driving high or low value customers?”, “What are prospects doing on the website after the campaign drives the visit?”  Unmotivated and uninterested, the analyst will resign to “burn and churn” reporting, characterized as low-value analysis, having minimal impact on performance and buying time until she finds a more interesting role elsewhere.

The answer?  In the future, smart agencies and marketers alike will migrate away from a “data-centric” service approach, instead evolving to a “use-case-centric” approach.  This will require that marketers and agencies develop much more open data platforms, policies, and contracts to build more sophisticated, higher value, solutions.  This type of access will breed a variety of challenges around inter-agency relationships, but that is an entirely different post.

Adam Gitlin, US Director of Analytics, OMD