Geo-fencing Part I: Don’t Mess with the Fence

By the Editors

What will be the next smartphone feature that will be differentiating and cool?  Seeking Alpha Blog

At the intersection of the physical and virtual world lies a mobile capability called geo-fencing. A context aware app that reminds you to pick up your laundry as soon as you approach the drycleaners on the way home from work? Brilliant? Scary?We asked our Director of Mobile to explain the basics of the phenomenon, address privacy concerns, the trouble with battery drain, key applications and why geo-fencing has finally arrived in 2013.

Q: Can you explain geo-fencing to the layman?

A: Geo-fencing is a technology that provides the ability to create a virtual fence around a geographic location in the real world. Smartphones that are location-enabled can detect when someone enters or exits these fences, which can be as large as a city block or as small as a retail store.

Q: How exactly does geo-fencing work?

A: In order to create a geofence, we first determine the latitude and longitude of a particular location, or set of locations. We then assign the radius of the geo-fences depending on the type of location. If we were to geo-fence a supermarket chain, for example, we might set a radius of 200 meters, but for coffee shops we would only set a radius of about 30 meters.

Once set, the location and radii of these geo-fences is communicated to the smartphone of those who have our mobile app installed. When and if the person then crosses into one of the fences, we record the date, time, and latitude/longitude of the event. In addition to logging this information, we can also choose to trigger a notification to the person, linking them to a survey within our app.

Q: How accurate is geo-fencing?

A: Geo-fencing is relatively (but not 100%) accurate. Though we can be very specific about defining the criteria for each geo-fence, there are certain technical limitations in the way that a smartphone determines your location, especially when the device is not actively being used (like when its in your purse or your pocket). Nonetheless, we are able to get a general understanding of someone’s location and can set an acceptable location accuracy threshold when attempting to understand when someone is within a geo-fence or not.

In a recent project, we set up geo-fences around a national fast food restaurant chain which triggered an alert asking if the person was dining there: over 80% of respondents said that they were at the location, with the remainder likely being at a store next door.