• by Steve Smith , Staff Writer @popeyesm, April 27, 2012

The emergence of the app platforms in digital media has opened up an entirely new vein of data to mine – several, in fact. With the help of smartphone-embedded GPS and tablet location awareness, an app maker can know exactly when and where someone is referencing his content or functionality. But that is only one piece of the contextual puzzle. To make richer decisions about what the use cases are for mobile media, we need to paint richer portraits of context. The possibilities for leveraging the location data layer are at least as exciting as behavioral targeting was back in the early days of the Web, when we started to see that all of those clicks told us volumes about what people were after.

The start-ups are piling in. Some like PlaceIQ are profiling locations not only by resources nearby but by activities that occurs within areas at any given moment. They are aimed toward advertising and targeting mobile ad inventory at specific places and times.

Another company called Placed is also profiling resources and activities in location but aiming their services purely as analytics for app developers . According to founder and CEO David Shim, the idea is to give a developer, whether it is a brand crafting a branded experience or a game maker, the richer context in which people engage their apps. “We tell them when someone opened their bar code scanner app and can tell them what other stores were nearby. We can tell a game maker what their users are near when they were playing their game. We are proving a context in the real world when people interact with your product.”

In some ways this is like the early years of behavioral targeting when companies like Tacoda and Revenue Science were first coming to understand the implications and uses of the new dimension of behavior (the context of clicking) they were assembling. At first the black boxes of BT were merely identifying as “auto intenders” people who had been to some page with car content on it within the last few weeks. Over time this aggregated data began to render new possibilities, like uncovering affinities no one could have suspected.. Who knew that someone whe liked XYZ topic also indexed so highly on viewing otherwise unrelated ABC content, or even were so much more likely to buy MNO widgets?

Likewise, mobility is offering us this location + behavior  set whose utility in only beginning to reveal itself. For instance, Shim says that when the company started building its  local profiling engine that works in a partner’s app, it could figure out whether people tended to use an app while moving or stationery, or even in a moving car. This knowledge can change everything when it comes to design. One game maker was able to see that its game was being played in cars. Yeah, I get chills at the idea of someone trying to play a mobile game while driving. But it gets the designer thinking about the viability of adding voice controls. Those working on a photo-sharing app can better understand the kinds of contexts in which people use the app, either to take pictures to share or view them  

For sales staffs, of course, location profiling of applications can be used to identify businesses that index strongly for being in the area when their app is open. In one app studied, for instance, people who opened in in Texas were indexing higher for being near a Walmart store while in California people opening the same app were tending to be closer to Target. As the database grows over time, the company can do benchmarking of apps to see how they index against other types for a given area, Shim says.

In my mind, this sort of data really gets interesting in the aggregate, when we can start looking for patterns in the way people use apps in combination, and how they use multiple apps in different places over time. What if we knew that three hours before people opened their Starbucks app, they tended to be at fast food restaurants? So you have fast food restaurant patrons overindexing as Starbucks patrons within a time parameter. You might be able to target people in a restaurant app with Starbucks or competitive ads four hours beforehand.

This is starting to feel like behavioral targeting in 3D.