Building a consumer app in 2015: How developers can add context for better engagement

Location in mobile is still often defined by its WBSGAC (Walk By Store, Get A Coupon) heritage. But as our smart devices get smarter, and as predictive analytics gain momentum, location data could disrupt the status quo and deliver a more powerful, valuable, and personalized mobile experience.

Location-based marketing specifically is moving more aggressively toward the idea of employing location as a contextual device. Context here is created by the analysis of intermittent location signals of a device over a period of time to create broad patterns of geography-based behaviors; these patterns are then combined with business locations, event information, and demographic data to derive behavioral, geographic, demographic, and commercial attributes about the owner of a specific device.

The excitement of this approach is that the resultant data addresses the entirety of your audience, and it is created without keying off personally identifiable information, or “buying in” third-party data of potentially dubious origin.

The tradeoff is that this approach requires access to intermittent location — that is, location signals that are produced infrequently over weeks or months — and audience data is not available immediately off-the-shelf, but must instead be created over time. On balance, however, mobile publishers with access to a location signal now have access to a self-perpetuating and increasingly relevant user-data stream. As time passes, and more data is gathered and analyzed, the mobile publisher may see deeper behaviors and preferences, and has the opportunity to provide an increasingly personal experience to the user.

This enriched context is a leap beyond current thinking, where consumers are understood to be associated with only two locations: where they live, and where they are now. If you think this is hyperbole, take a look at the Real Time Bidding (RTB) specification used to standardize how ads are targeted to the appropriate device: both the user object and device object accommodate one location each: “home” and “current,” respectively — the idea of a person’s existing in a spatial or temporal continuum is simply not accommodated.

This example is not a criticism of the RT, which does its job very well, thank you very much, but is rather an illustration of how location is still employed very primitively today. With contextual data, however, developers can write smarter, more informed apps that deliver valuable content in advance or in retrospect — bringing anticipatory computing into the consumer experience.

Broadly speaking, this contextual data is used to complement and enhance product-specific learnings into consumer preferences. Here are a few examples of contextual personalization:

• Your Pandora app is loaded with reggae and Latin music stations. Based on your location — and maybe even your travel patterns (you travel from Chicago to Los Angeles once every quarter) — the app will send you a notification about your favorite performers and bands that are touring locally, or one that is performing in Los Angeles in the near future.
• Your weather app observes your trips to the beach or the slopes (or both) and recommends similar places for your next winter and summer vacations. Combine that with your frequent visits to Toys”R”Us or Gymboree, and the app might recommend a kid-friendly vacation destination.
• Your Waze app learns your commute and surfaces suggested changes in start or departure times based on historic traffic patterns and real-time data before you set out on the road.
• Uber or Lyft could observe airport travel patterns and suggest their services and promote savings over park-and-fly or competing taxi services for your next outgoing trip or return home.
• A messaging app, such as Facebook Messenger, can use behavioral patterns to disambiguate places you mention in texts, and link those exactly to the appropriate services such as OpenTable or Yelp.

Having learned and applied best practices around user engagement in 2014, app developers have an opportunity to push the envelope even further in 2015 with anticipatory computing.

The combination of behavioral, geographic and demographic data delivers insights that may empower app developers to surprise and delight their users. Mobile apps that are an integral part of our mobile experience — Lyft, Facebook Messenger, Spotify, and others — can further differentiate themselves from their competition and establish true brand loyalty by delivering a more personalized experience. This requires learning more about your users and their preferences, expanding the data that is collected, and employing it for an improved consumer experience.

If we do this correctly, our smart phones will get smarter, more personal, and significantly more interesting.

Tyler Bell is vice president of product at Factual.