Using geofences for profit, safety and fun
A geofence is a boundary of some significance. In today’s location-focused marketing arena the boundary is valuable because an individual on one side of the boundary is a potential customer and the one on the other side is probably not. But, where is that boundary? Is it a radius around a brick and mortar store’s centroid? Is it a buffer around a building footprint? Or, is it an existing boundary, one that a data company has all packaged up and ready to go?
Darrin Clement, CEO of Maponics
, gives credence to all of those options, but wonders why that last one is still not as popular as the others. His company is best known for its neighborhood boundary datasets, so it makes perfect sense to him to think of neighborhoods and other data layers as predefined geofences. His customers share that perspective as they tease out how to best use location data and geofencing tools to capture customers, keep citizens safe and develop location-based games.
I challenged Clement to provide an example of when a radius around a brick and mortar store would be acceptable as a geofence. He cited a competitive play. If a competitor, say Walgreens, wanted to lure customers from the CVS down the block, it might use a simple radius search to find those passing near the CVS. Individuals might be offered an incentive, 10% off on a prescription perhaps, at Walgreens instead.
New datasets serve as geofence boundaries
What Clement is finding is that marketers and others learning about geofencing want more detailed geographic information to use as boundaries. That’s led the company to develop several new datasets.
The Shopping Boundaries dataset
, launched in 2012 and updated quarterly, includes malls and city shopping districts. The mall boundaries include something very important: the footprint of the parking lots. Why is that important? Per Clement, “Once in the parking lot, marketers feel confident a visitor is there to visit more than one store.” The city shopping districts are harder to identify; they bound dense areas of shops and restaurants in the top 125 metro regions in North America. (Figure 1) Maponics has done the work to include both the formal names for these areas and the colloquial ones, those more likely to be familiar to locals.
Figure. 1 - Shopping boundaries in San Francisco, CA are among those in the Shopping Boundaries dataset.
The boundary of a college campus, a small compact one, is a simple polygon. But, many city and even suburban campuses have two, ten or even more remote “islands” separated from a main campus. Parties interested in this dataset did not want the complexity of too many polygons, so Maponics and the client agreed on no more than five “islands” per campus in the College Campus Boundaries dataset
. Interest in campus boundaries comes from marketers, safety and security demands (which police force has jurisdiction - school police? city police?) and game developers (different rewards are available for tasks on vs. off campus).
Destination and Venue Boundaries
The Destination and Venue Boundaries dataset
includes a variety of locations that have one thing in common: a large number of people are there together, for a short period of time, for the same reason. It includes 3,200 North American venues including amusement parks, sports stadiums, speedways, ski resorts, airports, train stations, and thanks to one customer who really wanted and paid for them, golf courses.
Figure 2 - Cowboys Stadium, Arlington, TX is in the Destination and Venue Boundaries dataset.
What about locations that host such gatherings, but are not really used for that purpose all the time? I was thinking about Boston’s Esplanade, where up to one million people gather each July 4th to hear the Boston Pops and watch fireworks. Clement agrees those are important gatherings, but offers that to date, his customers have yet to determine a way to use intermittent boundaries. He predicts that within three or four years marketers will want this data and know how to turn it into something valuable.
Tips for those just getting started
Clement offers a few tips for those dipping their toes into the geofencing waters:
Try to avoid any solution that involves installing hardware beacons.
Try to use existing data (even free ZIP Code boundaries) to get started, then move to more detailed and relevant boundaries as you learn more about the technology and your needs.
There’s a good chance that someone has created the geofencing boundaries you need; don’t spend the extra effort recreating them.