• by Kathy McCarthy , May 18, 2010

Access to digital data has brought a wealth of opportunity to marketers, allowing them to reach buyers and prospects anywhere online. Yet that abundance creates its own unique problems. Managing the vast amount of data associated with digital ad exposures and interactions -- especially as agency relationships change -- is a challenge that marketers must deal with in order to realize the full potential of digital media.

Marketers' approaches to data management can vary dramatically, depending on the type of data. Often, traditional media data have come from a single trusted source, i.e., Nielsen for TV GRPs. While managing it is far from simple, collecting and managing digital data add layers of complexity.

First, several different components sit under the umbrella of "digital," including Web site traffic, display and paid-search impressions and social media impressions. Moreover, data within each group can be broken down much further (display into rich vs. non-rich media, etc.)

Second, managing and analyzing all these data sources often require dedicated resources and expertise. This leads marketers to rely on digital agencies to host and store their data, which brings up another issue. Many marketers have multiple digital agencies, each focusing on different areas of expertise, such as SEO/SEM or social media, so their data often live in multiple locations.

While this reliance on digital agencies might seem logical, especially considering data complexities, marketers should be aware of the drawbacks in doing so and have a game plan to mitigate any of the risks.

The primary question that marketers need to ask is: "What happens to my data if my agency relationship ends?"

A rule of thumb: whoever owns the relationship with the third-party vendor supplying the digital information, hosts (or has exclusive access to) the data. For example, if one agency uses Atlas for ad serving and then a new agency uses DoubleClick, it is impossible to access all historical data in one place (and on an apples-to-apples basis).

Usually, a digital agency periodically supplies a marketer with aggregate or site-level data and dashboards. Most often, this information serves its purpose, providing a general understanding of digital activities and performance over time. In the event of an agency change, these reports and dashboards will likely be passed along to the new agency to give it a view into historical brand performance.

The problem comes when marketers want access to more granular or longitudinal data. This can be extremely useful in helping the new agency formulate creative, targeting or placement decisions, but is not likely to be in the aggregate reports and dashboards. Should the marketer want to analyze the impact of digital efforts alongside other vehicles, these data will not do the trick; that type of analysis requires examination of historical, weekly data.

Michele Madansky, senior digital consultant to MMA and former VP of research for Yahoo, has seen these challenges firsthand. "Unless specifically prompted, agencies typically provide the easiest data to access -- weekly total display impressions," Madansky said. "There are instances like the Yahoo home page which could include millions of ad impressions on a given day that are not ad-served. In addition, many agencies don't want to pay ad-serving fees for large ad network buys. These are two examples of missing impression data that could play havoc with your models."

So, what can marketers do to control the situation and mitigate risk? It starts with "owning" their data.

Marketers should absolutely rely on digital agencies to manage third-party ad serving and reporting. But wherever possible, they should also "own" the relationship with these ad-serving companies so it can remain intact and consistent in the event of an agency change.

One solution: assign an internal "data czar" who knows the disparate data sources and their associated nuances. For example, if marketers execute on sites that don't allow ad-serving (or don't want to pay ad-serving fees for large ad network buys) this could lead to many missing impressions.

Another major benefit of a marketer owning the data is related to data integration. If it is aligned with other online and offline marketing data, this allows for richer analytics. Also, marketers are no longer limited in their ability to analyze a single digital medium or campaign, which establishes a level playing field for all marketing activities.

The bottom line is -- marketers need uninterrupted access to disaggregated and granular data if they want to optimize performance; it allows them to look back in time, evaluating what worked and what didn't and adjusting plans accordingly. It's the difference between old-style guesswork marketing and 21st-century analytics that allows for continual evaluation and adjustment in today's fast-changing digital world.