For years we’ve heard “this year is the year of mobile.” Premature as that was in 2012, it’s safe to say that marketers have quietly reached mass adoption. In a natural evolution, marketers and agencies have begun exploring how to leverage mobile data and technology programmatically to boost campaign efficiency and effectiveness. Yet, the biggest hurdle to flawless execution isn’t the complexity of the programmatic ecosystem; it’s an inability to apply the lessons learned from the more than five years of programmatic buying online.
Lesson 1: Context, Inventory Quality, and Creative Matter
Prior to the programmatic revolution, buying was largely limited to contextual targeting, or direct–to-publisher deals isolating pages or site sections relevant to a specific audience. This was less than an exact science and led to large media waste and issues of scale.
With the advent of programmatic buying, advertisers and agencies were excited by the ability to purchase audiences at scale across inventory sources. The context in which the impressions were served was less important than reaching a specific audience. Remnant inventory reigned supreme on the exchanges, as did the lowest common denominator standard .gif ad units.
We know that audience targeting works best when applied to contextually relevant, high-quality inventory and creative. This is true with programmatic mobile as well, where inventory comes from either mobile web browsers or mobile apps. Currently 80 percent of mobile usage takes place via apps, and in general these impressions are contextually more relevant and provide more engaging user experiences.
Lesson 2: Third-Party Data isn’t a Magic Bullet
The idea of buying pre-targeted audiences built from third-party data thrilled advertisers and their agencies back in 2010. They were told it would eliminate media waste by focusing only on users relevant to your brand. Auto marketers had scads of in-market auto intender segments to choose from, and GPG brands had soccer moms, and households with kids.
These segments provide scale and more focused audience targeting than run-of-site campaigns, but rarely perform at a superstar level and often do not warrant the data premium placed on the CPM. Often there is no telling how stale or reliable the data is, or whether it’s too granular to make a difference. And even in the best circumstances, third-party data is not exclusive and offers less of a competitive advantage than you might think. Your competition is buying the same information.
The same data shortfalls seen with online display campaigns also exist in programmatic mobile. Take location data, for example. Over half of all impressions available on mobile exchanges now come with some form of location data. Advertisers can leverage this data to proximity target, or audiences based on path analysis of consumer real world behaviors. For example, a device seen at an airport 4-6 times a month could be included in a business traveler segment.
The problem of relying solely on location data included in RTB impressions is that the accuracy can range from pin-point precision of GPS coordinates from location-aware mobile apps, to much less accurate signals like centroid data, which estimates location by triangulating signals from multiple cell towers, to zip codes pulled from user profiles. Just like online, data quality and exclusivity directly impacts campaign performance.
Lesson 3: Programmatic isn’t Always the Answer
Similar to online, there are times when programmatic buying makes sense and other times it doesn’t. Automated buying works best with always-on campaigns where targeting (audience and proximity), reach and spend remain fairly constant over extended periods of time. Setting up these campaigns once and then giving powerful algorithms the room to optimize over time can gain great efficiencies. This is not to say that all mobile can or should operate programmatically. Even more than five years into the revolution online, programmatic accounts for less than 50 percent media spend. Mobile will continue to have shorter maturation cycles, but it will be awhile before programmatic takes over.
So, as you plan your mobile strategy for the coming year, don’t forget to apply the programmatic lessons you’ve already learned: Context matters, data is no magic bullet, and we’re not ready for the autopilot to take over completely.