Admittedly, what startup Immersive Labs is about to launch sounds more than a little creepy: Face detection software that studies your face at a kiosk or brick-and-mortar store and immediately determines your age, gender, attention span and maybe even your emotions. It enables marketers to run real-time analytics on this demographic data, and even quickly change a kiosk or advertising display's content to match the needs (or wants) of its customers.
Called "Cara," this face detection platform is Immersive Labs' first product, one scheduled to ship around March. Founded in 2011, the New York City-based Immersive is currently testing Cara, an "engagement platform for audience analytics," with about 25 customers, the company's chief operating officer Steve Lubin told InformationWeek in a phone interview.
The Cara software is designed to work with any off-the-shelf webcam and PC, and it works with a variety of operating systems, including Android, Linux and Windows.
"It will work with any kind of basic hardware, nothing special is required in order to operate it. We provide the software and the analytic tools," Lubin said.
[ Apps may be the best solution to the data scientist shortage. See Big Data Apps: The Next Big Thing? ]
Cara's real-time software detects faces on the fly, and collects the kind of customer data that's usually hard to obtain in brick-and-mortar stores, such as foot traffic and demographics. It also can monitor store entrances and endcaps (merchandise displays at the end of aisles) to determine the number of people who come to a location, how long they linger and what draws their interest.
Naturally, a face-detection system that detects age, gender and even a shopper's attention span is bound to raise a few privacy eyebrows. But customers have nothing to fear, Immersive Labs claims.
"We are face detection, not face recognition. The only information we're collecting is gender, age, distance from camera, attention time, what your emotions might be, glances -- things like that," said Lubin. "We don't keep any video. We don't store any video. We're not sending any video. The only thing that we're sending back to the servers is demographic information."
In fact, the company is putting a lot of pre-launch effort into educating people on how its face-detection system works. "We're very much trying to stay on the right side of privacy. That's the core goal of the company," Lubin said.
Cara is a cloud-based system. Retailers use a Web browser to log into their accounts to see their analytics. "There are also APIs, where you can take the data yourself and integrate it into your own data products and services," said Lubin.
Cara also ties into camera-equipped advertising and content services, including in-store kiosks, digital signage, endcaps and mobile devices.
"Besides providing real-time data, we're also able to provide triggers that (allow) you to switch out and change content in real time," said Lubin. "If you have a tourism kiosk in a city, and the people looking at it are young adult males, you might want to provide them certain content or advertisements. We have tools so that you can switch out those advertisements in real time with your content management system."
Immersive Labs is one of several companies that offer big data analytics systems for brick-and-mortar retailers.
Silicon Valley startup RetailNext, for instance, uses an on-premises Linux server to collect and analyze data from a variety of in-store devices, including video cameras and point-of-sale systems.
And the LightHaus Visual Customer Intelligence (VCI) system analyzes video from in-store cameras to measure customer traffic and engagement. The system is designed to help retailers boost their sales conversion rate -- the percentage of store visitors who buy goods or services.
Predictive analysis is getting faster, more accurate and more accessible. Combined with big data, it's driving a new age of experiments. Also in the new, all-digital Advanced Analytics issue of InformationWeek: Are project management offices a waste of money? (Free registration required.)