Startup tracks smartphone users to see where the action is

You know what Google Inc. does on the Internet, right? Sifting through billions of destinations to help people find what they’re looking for? That’s what Sense Networks executives want to do for every city on the planet.
A New York-based startup that will emerge from stealth mode this week, Sense Networks hopes to glean location information from phones, taxi cabs and other on-the-move items to determine traffic patterns and uncover consumer trends in the physical world. The company plans to sell the information to hedge funds eager to discover the optimal site for a shopping center, say, or determine which new bar attracts the most affluent clubbers.
“Google indexes the Internet using machine learning – instead of looking at the content on a certain page, it looks at pages that link to that page” to determine how to categorize and rank the site, Sense Networks CEO and co-founder Greg Skibiski explained. “You look at where everyone came from to get to that page, then, less so, where they go when they leave.”
Here’s how it works, at least in theory: a consumer with a mobile device agrees to allow Sense Networks to access their location information – but not identity or other personal data – in exchange for a downloadable location application. The information becomes part of an anonymous database that allows users to see the tastes and behaviors of people who live or work near them, for instance, or who are active at the same times of day.
So while the platform, dubbed MacroSense, can’t tell the difference between a hip-hop club and a ballroom-dancing competition, it might deliver information on the first nightclub to a consumer living in a trendy urban area and details about the latter to someone who resides in a sleepy bedroom community. Unlike social networking sites, which feature user reviews and suggestions, Sense Networks users will see where fellow members are actually going, and when. And it’s that quantitative element that Skibiski hopes will be key to tapping some of the revenues social networking sites have yet to access.
“First you start with search on the bottom, then there’s local search, then above that you have where everybody is right now” with friend-finding applications, on top of that, you have ‘Where is everybody like me right now?’ You don’t just see hotspots (with Citysense), you also see where everybody like you is right now,” Skibiski said. “Sure, people use Facebook (for networking), Loopt for friend-finding, but it’s all filled with bias as well, which makes (them) useless for any sort of financial analysis.”
Sense Networks is targeting on-the-town types with a new application, Citysense, and has quietly launched service in San Francisco. The company will adopt a Facebook-type model, opening its platform in the hopes that outside developers will build to it, and its coming-out party is timed to coincide with the launch of the first third-party applications for Apple Inc.’s iPhone.
The startup is backed by players in the hedge-fund space and has already raised an undisclosed sum in a Series A financing round. And Sense Networks’ front office is formidable, technologically speaking: the other two co-founders are Tony Jebara, director of the Columbia University Machine Learning Laboratory, and Sandy Pentland, who 10 years ago was named by Newsweek as one of the 100 Americans most likely to shape the 21st century.
There’s little doubt that analytics regarding physical traffic could be prized stuff – indeed, retail centers in the United Kingdom have begun tracking the movements of cellphones in an effort to pin down the behavior of shoppers. But Sense Networks must build a sizable user base before it can offer information of any real value, so its success may hinge in forging alliances with successful developers of location-aware applications.
There are other major hurdles, too. While taxi cabs and other means can provide substantial insight, mobile will need to play a pivotal role. So users must be willing to allow Sense Networks to track their movements, likely in exchange for free or subsidized applications. The company has no plans to build an application for feature phones, and is only considering a Symbian version. And the most obvious distribution strategy for the company’s wireless play – viral marketing that allows users to invite friends to use the platform – is a dangerous play, given the thicket of privacy issues surrounding GPS-enabled applications.
Skibiski acknowledged the challenges, noting that the effort is “just a proof of concept.” But the new wave of GPS-enabled applications will result in a sea of valuable data, he said. Sense Networks just needs to figure out how to turn that information into dollars.
“All this location stuff is going to start generating tremendous amounts of data,” Skibiski maintained. Companies and their investors “are not going to know what to do with it, but they want to know that information about their customers. Our core business is machine learning, analytics, all this kind of stuff. And we can monetize it.”