How Trump Win Explains Future Of TV Advertising
The second biggest story on Nov. 8 was the incredible failure of polling. Many were surprised by Trump’s triumph because the expectations going into election night were so different.
A political pundit once said that the only poll that matters is the one on Election Day, but that belies the true difference between polling and voting: one is attitudinal, while the other is behavioral. Attitudinal measures can be notoriously wrong and, quite simply, miss undercurrents in the marketplace.
For decades we have worked with attitudinal measures to justify advertising spend, especially in television, where behavioral measurement has been lacking. Studies and tools have focused almost exclusively on using attitude or “intent” to justify investment or success. The problem is that these studies and tools can be wrong. Intent to purchase is not a purchase, and one doesn’t always correlate with the other. But if it’s all you’ve got, then you run with it.
Fortunately, there are changes happening within television media measurement that will provide the behavioral component that we all need.
Predicting revenue through advanced analytics
Over the past 24 months a number of companies have been applying advanced analytics to both media mix modeling and individual channel optimization. These services solve the behavioral problem by applying a “top-down” solution using various forms of regression analysis, connecting investment to sales and revenue. It’s still early days for these solutions, but they are evolving fast and promise to deliver a far better predictor of revenue than a measurement of intent.
The challenge with these solutions is that they typically operate in a black box due to use of proprietary algorithms and IP issues. When brands receive the output of the studies, the first question is typically “Do we trust it?”
That’s a fair question when you don’t have other evidence to corroborate results. These tools will accelerate in usage as they become validated, which may involve running the models where channel mix and attribution are known quantities.
Delivering insights through advanced TV
The emergence of advanced TV data sets from cable set -op boxes and connected TVs holds great promise for delivering behavior-based media planning insights, as well as an ability to connect viewing to direct behavioral engagement.
By matching first-party CRM data to direct viewing behavior, you have the potential to short-circuit targeting exercises by simply understanding what your customers watch, independent of knowing who they are. Now you view indices for viewing based on real behavior — not answers to survey questions or a lot of assumptions about target makeup.
These same data sets, when coupled with household device graphs and tracking tags, can also provide reporting on engagement and interactions with brands. The analytics associated with them potentially show attribution to station, daypart, and program, and can even reveal optimal frequency models unique to a campaign.
The problem with these data sets is that they are too small for many advertisers to utilize, often only representing a few percentage points of total households. Large volumes of first-party data and a significant amount of advertising have to be run in order to get results that are meaningful, accurate, and granular.
Some advanced targeting solutions have been tied to black boxes of media inventory or require a media buy as part of the service. This practice, which is a component of today’s programmatic TV, will most likely fade away. Unlike digital, there is no need to bundle advanced targeting and measurement with a specific media buying platform along with TV.
Certainly the evolution of analytics, targeting, and attribution are all headed in the right direction. The strategy of focusing on behavior vs. intent will reveal far greater insights and accuracy than ever before.
There will always be a critical role for measuring brand health — but it needs to be complemented by behavioral measures that correlate investment to revenue. Otherwise, you may wake up one morning realizing the revenue you see going down today is in stark contrast with the measure of intent you saw going up yesterday.