Steve Smith’s Eye on Digital Media—Context Ain’t What It Used To Be: Big Data’s Big Impact on Big Old Media

By Steve Smith
01/21/2013

About a decade ago, Dave Morgan taught me about targeting “people not pages” in digital media. Morgan now runs Simulmedia, which sells targeted TV advertising using data from set top boxes and online. Back then he was driving Tacoda, an early behavioral targeting company that tracked user actions online in order to target advertising to them based on their previously profiled interests, not just pages viewed.

It seemed to me at the time that "behavioral targeting" fundamentally challenged what media companies do–build contexts that both segment an audience into a likely target and create an environment where they are most receptive to a sponsor’s message. Jump a decade later and that challenge of data to traditional media models is unmistakable.

In just a few years, a large share of digital media spending has moved from site-specific contextual purchases and traditional ad networks to "demand side platforms" and programmatic buying through real-time bidding that acts on mountains of data.

Media buyers are finding their targets more efficiently on cheaper locations around the Web as data replaces media as the key element in gathering audiences. But it is even more challenging than that Morgan explains when I recently asked him to refresh his "people vs. pages" model. In a digital economy where distribution is plentiful, attention is scarce and data about site visitor may be at least as important as the description of the content they visit, "probably content and context will go from primacy to a secondary position in establishing media value."

The critical issue in a data-driven age is finding the target, the right person, with the right profile at the right time. One big data vision is that it gives marketers a 360-degree view of the customer because so many inputs (transactional, mobile, Web, etc.) paint a richer portrait of their lifestyle and purchase path. Following consumers across channels as they take this "journey" (the new nomenclature) requires focusing on the person not the place. "If context adds value, that will be a secondary consideration that will be determined in premium pricing," says Morgan. But in a data economy of so many different inputs, content and context now compete against search data, state of mind, physical location and histories of how that person previously behaved.

We all sense that a lush brand ad running in an equally lush and well known media brand surely has more impact than it would running on the blog of that latest pro-amateur insta-critic. But that intuition is no longer enough in a data-driven economy. "Content and context went from being the only game in town to being one of a dozen different games in town, and its value is going to be entirely empirically driven," says Morgan.

The point isn’t that traditional media doesn’t provide value with its clean-well-lit place. The point is that the industry needs better tools than the gut to make their case. "It is going to have to be proven and valued each and every time because we will have data available to know how well it works," Morgan says. Even more, publishers will need to prove to marketers how these programs improved sales.

The data-driven economy also introduced real-time reporting, constant iteration and variability. All marketing is coming to see, because of these new inputs, the effects that time of day, location, the mood and mode of the user and many other variables have on ad effectiveness. "One of the things that will be challenging for the media industry is the dynamism and unpredictability of value."

Just as challenging will be staffing-up for the age of big data. Ad agencies face this as they build up their digital ad trading desks with math majors and look for planners and buyers who understand what the algorithms driving these real-time systems are doing. The great print titles always had their coveted "lists" and databases they leveraged in many lucrative ways. But "big data" is another animal–with cross-channel, multiple inputs, real-time responsiveness, and attribution models that will strive to show just how many cases of lipstick that campaign sold. Having the people in place to speak this new language to marketers (and to one’s own staff) may be the critical first move publishers need to take.

Steve Smith (popeyesmith@comcast.net) is digital media editor for min/min’s b2b/minonline.com. He posts regularly on minonline and directs the min Webinars. Smith also co-chairs min’s Digital Summits and, as ceo of Roving Eyeball Inc., he consults for a number of publishers in the digital space.