Big Panels and Big Data: A new paradigm for cross-media panels

Media currency, and the measurement systems that feed it, have always been more a function of what's possible, than what marketers have wanted.  The goal for marketers has always been to sell more product, or more service, or drive higher prices for their product or service.  Media currency, on the other hand, has always focused on delivery.  Marketers buy access to a group of people, defined by demographics, and pay their distribution channels based on that delivery.

The digital marketing revolution, now nearly 20 years old, brought with it a glimpse of a new paradigm - one where media could be bought and sold based on results, rather than simply delivery of an ad.  Huge Data Management Platforms (DMPs) incorporate digital ad exposure information, digital website visitation, and yes, even purchase data.  The goal is to create a continuous cycle, where media is targeted to individuals based on past purchase info, ad exposure, and cost.  Attribution models incorporate this information to feed demand-side platforms, where impressions are served to individuals based on a real-time bidding process.  If the model says the future revenue from the person is likely greater than the cost of the ad...then the bid is made.

The critical limitation of programmatic buying is that it's largely limited to digital data, both in terms of inputs and also outputs.  How about offline purchases?  And TV advertising exposure?  How about consumer attitudes which are known to be predictive of sales?  How do we get this data into the programmatic buying engine?

The answer is turning out to be cross-media panels.  The MediaPulse panel serves as a matching key between the offline world and the online world.  We are capturing a wealth of data valuable to marketers, and to attribution modeling teams.  We make that data available to DMPs via a one to one match between our panels and their cookies.  The DMP managers then propagate that information across the entire DMP, using various data fusion tools.

The result is that marketers are moving closer and closer to paying media companies based on an increasingly-accurate purchase likelihood of each individual, rather than delivery to a coarse demographic target that is only loosely correlated with purchase.  Media companies are responding with sophisticated supply-side platforms to help them more accurately price their inventory.

SymphonyAM is working on a variety of projects now with agencies, advertisers, and media companies to help close the loop between ad purchase and product purchase.  Fully specified attribution models will help marketers and their agencies buy the media that sells product...not just the media that delivers an audience.  The same data can help media companies price their inventory appropriately.  The net result will be a larger, more vibrant media market.