CrowdOptic Taps Smartphones to Track the Crowd’s Attention

The idea behind CrowdOptic, a service that allows event organizers to figure out what a crowd is looking at by tapping their smartphones, came to CEO and co-founder Jon Fisher last year during a routine meeting. As he sat down to chat with an investor in one of his earlier companies, he caught sight of two “tombstones”—the glass sculptures that memorialize deals between companies. He noticed that his chair and the two tombstones formed the three points of a triangle, prompting him to think about triangulation, focus, and sightlines. “That’s really where the inspiration came from for triangulating between smartphone users to isolate action,” Fisher says. “And that really formed the basis of the core technology, the notion that we can record smartphone attributes in real time and compare them with other users in real time.”

With this geometrical insight in mind, Fisher (pictured above) and co-founders Jeff Broderick, Doug Van Blaricom, and Alex Malinovsky set out to build a service that would give event organizers a way to monitor a crowd’s focus at live events. It’s a kind of outdoor equivalent of the eye-tracking studies Web designers run to see what parts of a Web page are attracting attention.

CrowdOptic's software assesses where users' smartphones are pointed to determine where audience members' sight lines are converging.

It works like this: event organizers partner with San Francisco-based CrowdOptic on an event like a tennis match. Match attendees download an app—either CrowdOptic’s app or one specifically attached the event. As they take photos and videos, tweet, and share information, CrowdOptic can draw on data from the compass and accelerometer in each smartphone to determine where the person holding the phone is looking—at one player or another, at the scoreboard, or even at certain ads in the stadium.

“We’re simply tunneling in through an existing process,” Fisher says. “We’re not envisioning everyone in the stands having phones in the air. People are naturally taking pictures and video.” Taken together, the data can show event organizers where the crowd’s collective attention is focused. In an event like a tennis match, there might be two main clusters of focus, one on each player, whereas a baseball game could have far more. (There’s a cool animated version of the graphic above at CrowdOptic’s website.)

With this information, organizers can rethink their ad sales strategy. In online advertising, sales offices can use eye-tracking studies and other forms of analytics to inform their pricing structure and explain it to potential customers. In live events, it’s hard to say which ads are viewed most. It may be a safe bet that most people in a given stadium are going to see ads on the Jumbotron, but how many fans will see an ad in right field or above a stadium box? “If one asset is being viewed five times more, you can charge five times more. You might think banner behind home plate is the best asset, but what’s the third best asset? We’ll be able to prove that data,” Fisher says. “It’s a whole new advertising form.”

Organizers can also hyper-target advertising or discounts directly to users’ smart phones. If a fan at a racetrack is focusing his attention on a given car, the driver’s sponsor might want him to see ads for their products, as a kind of augmented-reality overlay. Or they can choose to build in added benefits for fans, like providing up to the minute stats when someone focuses his phone on a given player. “If you’re watching a baseball game, you can get all info about a batter on the Jumbotron. But if he’s in right field, you can get the whole Jumbotron experience right on your phone,” says Fisher. Organizers can also use the information to retool their TV broadcasts, giving viewers at home an experience closer to what they’d get in the stadium.

The technology has security applications as well. Though CrowdOptic can’t use its apps to pinpoint where a given person is sitting in a stadium or who that person is—they can determine only the sightline of a given phone—the system can detect changes in crowd behavior, alerting event organizers to some sort of distraction. If, for example, a fistfight broke out in section 305 and people turned to look at the scene and dropped their phones, that change in position could alert crowd organizers that something is going on. It may just be that the whole section changed positions to check out the Goodyear Blimp as it floated by, Fisher says, but pinpointing a sudden change can help security identify potential problems.

“If we can figure out that an anomaly has occurred, we can get the first responders there 20, 30, 45 seconds faster,” he says. The company has already signed a deal to provide new technology to security firm Andrews International. But, Fisher notes, the company intends the technologies primarily for sporting and other events, not for broader security applications

Though a lot of companies are focusing on mobile apps, Fisher hasn’t identified any direct competitors. A serial entrepreneur—he sold his last company, Bharosa, to Oracle in 2007—he says this is the first time in his career that he hasn’t been aware of another company with a similar focus. “This is brand new technology,” Fisher says. “When we were taken up by Oracle last time, we had six competitors. I take a deep breath when I say this, but this time there’s no one.”

CrowdOptic raised a million dollars in Series A funding in March, some of it from Fisher himself and other private investors. In October, the company raised another half-million in a round led by Bowman Capital and Band of Angels. The company currently has about 15 employees.

So far, one of CrowdOptic’s biggest challenges has been mitigating the erratic compass signals in a lot of smartphones–even some of the latest models are pretty inaccurate indoors. The company has developed algorithms to smooth out the problem, but it’s a continuing challenge. “It works and deploys successfully, but [we] have to continue to work on it,” says Fisher.

And while deploying the technology is pretty easy at an event like a tennis match, where two players are well separated in different spaces, it can get more difficult at an event like a boxing match, where the two boxers are basically hugging. A game like football or basketball where players are constantly moving and in contact can make it much harder to separate objects of focus. But Fisher expects to see better smartphones within the year, making the process easier.

It’s also challenging to convince large entities to take up brand new technologies in a tough economic climate, Fisher says. Big institutions who want to be early adopters can be hard to come by. But so far, Fisher is pleased with the progress. The company launched out of beta testing in July at the Bank of the West Classic tennis tournament at Stanford University, and recently landed a deal to produce mobile applications for Sonoma Valley’s Infineon Raceway, which holds events like the NASCAR Sprint Cup Series.

“People really love the novelty. Whenever you have that in the mobile space that’s really a prized commodity.”

Trending on Xconomy