“Data-Driven Thinking” is written by members of the media community and contains fresh ideas on the digital revolution in media.
Today’s column is written by Lauren Moores, vice president of analytics at Dstillery.
Earlier this week, I participated in the Five Boro Bike Tour, a 40-mile trek through the streets of New York. We rode along closed expressways, across bridges and through neighborhoods that Manhattanites and Brooklynites rarely explore.
At one point, we had a bird’s eye view of a huge billboard, and those of us pedaling at a “comfortable” pace could linger on its message. It made me think of the age-old media challenge of defining an audience, a particularly difficult problem with outdoor media.
Can that audience be measured effectively? Well, yes.
Location signals derived from our mobile devices are the key to bringing traditional media measurement into the digital age. The industry has started to achieve that with in-store measurement, where the power of location has already proven to be an effective barometer for both desktop and mobile campaigns, as demonstrated by Placed, NinthDecimal, Factual and PlaceIQ, among others. Now, a similar application of location signals has the power to provide digital insights and audience optimization to offline media channels such as out-of-home (OOH) and terrestrial radio.
Measuring Local Audiences
Let’s look at OOH. Picture the advertising you see every day on the train platform or on your route to work. Most of the time, the decision to place ads in these locations is based on history, local brand experts or ZIP-based demographics. By capturing digital data on people in proximity to the ad location, we can start to paint a much more nuanced and dynamic view of the relevant audiences.
The location signals can come from apps and advertising with opt-in location tracking or an opt-in sensor. The relevant behavioral data, which can come from a marketer’s own CRM, a publisher, an ad network or a programmatic supplier, can help characterize and define the audience near an OOH asset during a particular period of time. Using predictive analytics to match consumer devices, it’s possible to marry the behavioral and location data sets to give an unprecedented perspective on OOH audiences.
Specifically, the data associated with various devices allows us to map app, site and place behavior to an otherwise offline local channel. From these behaviors, it is possible to derive brand signals and optimize brand placements. Compare the characteristic local behaviors to a broader population, either regional or national, and you can map the propensities of each audience to certain brands or brand categories, finally freeing the industry of coarse demographic proxies.
As an example, if I were measuring that NYC billboard within a small geofence during the bike tour, I would have signals from a subpopulation of bikers who were carrying their smartphones. I could match these smartphones to other devices, such as tablets and desktops, and capture the behavioral data of that population. I might find that the audience overindexes for fitness and health brands and certain beverage brands, but underindexes for insurance products and premium autos. This goes beyond macro census insights to better-planned media placements.
Media Optimization
Once a brand category or brand is chosen for a particular location, I can extend insights into the audience propensity for the same brand across a DMA or region to allow for optimal OOH media placement in a broad area.
Going back to our bike tour example, let’s say that I wanted to measure the propensity of fitness brands across all of New York City, not just the one billboard in Brooklyn. I score and rank each OOH asset according to the characteristics from the original audience and map these propensities visually. My analyses show pockets of the same propensity in particular areas across the five boroughs, giving an optimal placement for fitness brands.
The same approach for measuring local audiences and optimizing media is also feasible for local radio. I can determine brand propensities by mapping broadcast areas to ZIP codes and mining the digital data associated with the devices in those ZIP codes. To understand the level propensities across an area, I can map these to predetermined fixed locations and provide local audience propensities for radio at a granular level not otherwise achievable.
The rich insights that come from combining digital behaviors with location allows marketers to much more powerfully define audiences in traditional media channels, such as OOH and radio. Further, the scale of the data allows for time of day and week analyses that could benefit brand messaging or content, particularly with digital OOH. And again, even though mobile media is still difficult to measure, the location signals derived from this media are gold. The power of location is not just an alchemist’s dream.
Follow Lauren Moores (@lolomoo), Dstillery (@dstillery) and AdExchanger (@adexchanger) on Twitter.
