High-end retail stores are using facial recogntion software to identify VIP shoppers
When a young Indian-American woman walked into the funky L.A. jewelry boutique , store manager Lauren Twisselman thought she was just like any other customer. She didn't realize the woman was actress and writer .
"I hadn't watched The Office," Twisselman says. Kaling both wrote and appeared in the NBC hit.
This lack of recognition is precisely what the VIP-identification technology designed by is supposed to prevent.
The U.K.-based company already supplies similar software to security services to help identify terrorists and criminals. The ID technology works by analyzing footage of people's faces as they walk through a door, taking measurements to create a numerical code known as a "face template," and checking it against a database.
The company's VIP Identification software monitors data from real-time closed-circuit TVs or surveillance cameras, matching images against a business' VIP guest database. If it spots a match, an alert, mobile or otherwise, can be sent to retail or hospitality personnel.
In the retail setting, the database of customers' faces is comprised of celebrities and valued customers, according to London's . If a face is a match, the program sends an alert to staff via computer, iPad or smartphone, providing details like dress size, favorite buys or shopping history.
The software works even when people are wearing sunglasses, hats and scarves. Recent tests have found that facial hair, aging, or changes in weight or hair color do not affect the accuracy of the system.
The system is currently being tried out in unnamed designer boutiques and hotels in the UK, the US and the Far East. Previously, the company has designed software to help identify terrorists and criminals for security services.
'We're trialling the system in general retail, which would include hotels and anything where the public are walking in,' NEC IT Solutions vice president explained to the paper.
The technology which builds on NEC's existing NeoFace Watch security product, is being tested in a dozen undisclosed top stores and hotels in the U.S., the U.K., and the Far East. NEC hasn't responded to NPR's requests for an interview, so it hasn't addressed why the stores that are testing the software are staying quiet about it.
Chris de Silva, vice president of IT Solutions at NEC, told the Sunday Times that the company had addressed privacy concerns and found that most high-profile customers are "quite happy to have their information available because they want a quicker service, a better-tailored service or a more personally tailored service."
An article in the Economist points out how easy it would be to monitor/hack the system, “because most Wi-Fi devices broadcast a list of known networks, a monitoring system could, in theory, collect the list and match it against databases of known Wi-Fi networks, which are used as a rough and ready alternative to satellite positioning in built-up areas. Shoppers’ stored list of connections could thus reveal where they live or work, and possibly their identities.”
As Jathan Sadowski points out for Future Tense, “primarily underwritten by corporate money, much of which originates from the tech sector—Facebook and Google are listed.”
"I like to think that we treat all our customers like VIPs," Twisselman says.(therein lies the problem, how long before all customers entering a retail store are entered into a database)
http://www.npr.org/blogs/alltechconsidered/2013/07/21/203273764/high-end-stores-use-facial-recognition-tools-to-spot-vips
http://www.usaukonline.com/the-woman/22639-how-facial-recognition-technology-used-to-trap-terrorists-is-helping-shop-assistants-target-vip-customers.html
http://www.luxurydaily.com/facial-recognition-software-will-help-retailers-create-personalized-shopping-experience/
http://endthelie.com/2013/07/23/as-invasive-in-store-tracking-technology-becomes-more-common-companies-attempt-to-self-regulate/#axzz2Zy55tR7P
How retail stores track your in-store movements via smartphone –without you connecting to their WiFi:
The system called Euclid Zero provided by Euclid Analytics, can precisely track the movements of individual phones, even though they never actually connect to a Wi-Fi network. (For more information checkout the Stream20 website.)
In the US, retail performance consultancy Claris Solutions has joined forces with data analytics start-up Euclid Elements, to offer a solution that uses mobile sensors to map shopper traffic patterns in retail spaces.
The new ‘Claris-Euclid Store Experience Program’ uses these sensors to detect Wi-Fi signals on shoppers’ smartphones, to enable consultants to understand exactly how shoppers move through a physical store environment. Claris then provides anonymous and aggregated shopper data to help retailers make decisions about store layouts, merchandising, staffing and other aspects of the in-store experience.
How does it work?
The technique takes advantage of the fact that Wi-Fi wireless networking protocols are "promiscuous": the Wi-Fi adaptors in laptops, phones and base-stations reveal a lot of information about those devices as they attempt to negotiate connections with other devices nearby. Even before a device hooks onto a Wi-Fi network, it continuously spews identifying information over the air. Most devices send "probe requests", which are akin to a town crier shouting out the names of networks which the device has previously connected to, so that a nearby base-station that matches any of these requests can respond. The requests run unremittingly across all available frequencies until a connection is made. Even devices that are seemingly turned off, such as sleeping laptops, send out such probes, though at a slower rate. Place several Wi-Fi base-stations in a shop, then, and you can pick up these probe requests, trilaterate the positions of the devices sending them, and thus track the movements of individual shoppers, seeing which racks or displays they stop at, and what paths they follow through the store.
This is arguably just the latest development in the well-established field of "retail science", in which the movement of shoppers is tracked and analyzed. This was once done using video cameras, with footage examined by operators to determine where best to place new products or displays. Analysis of video is now heavily automated, and computers grind through the data to work out when stores are busiest, when queues are longest and how the positioning of products and promotional displays affects sales. The use of Wi-Fi tracking allows merchants to track individual shoppers more accurately than is possible with video, particularly in crowded stores. It also means returning customers can be identified without the need for facial recognition, because every Wi-Fi device has a unique, factory-set identifier that is broadcast in its probe requests.
Euclid's analytics package includes features like:
Customer loyalty/frequency, which measures things like daily deals, promotions, and whether or not customers returned based on those factors
New versus repeat visits, which measures how many times a customer visits a store and whether or not the customer has been to another store location
Window conversion rate, which measures how many people passed by a location and how many people came in
Dwell time, which measures how long customers shop
It is true that shoppers are on private property, and signs announce the use of tracking technologies. But improvements mean that Wi-Fi signals travel much farther than they did in the 1990s, so that people who merely walk past a store or look in a window may be picked up by internal tracking systems. More worryingly, because most Wi-Fi devices broadcast a list of known networks, a monitoring system can collect the list, associate it with the device's unique ID and match it against databases of known Wi-Fi networks, which are used as a rough and ready alternative to satellite positioning in built-up areas. Shoppers' stored list of connections could thus reveal where they live or work, and possibly their identities. Google faced worldwide scrutiny from regulators, and had to pay fines, after it emerged that its Streetview mapping vehicles had collected massive amounts of data broadcast publicly by Wi-Fi networks, computers and mobiles in many countries. Accordingly, Wi-Fi tracking firms now seem to be trying to get ahead of regulators. The day after the New York Times story appeared, Euclid and other firms announced a plan to partner with the Future of Privacy Forum to set rules about Wi-Fi tracking. In the meantime, if you are worried, there are two absolutely effective ways to prevent such tracking: turn off Wi-Fi and Bluetooth on your devices, or turn them off altogether.
http://www.rawstory.com/rs/2013/07/23/how-retail-stores-track-your-in-store-movements-via-smartphone-without-you-connecting-to-their-wifi/
http://techcrunch.com/2013/01/04/euclid-the-google-analytics-for-the-real-world-partners-with-aruba-aerohive-xirrus-others-to-make-customer-tracking-sensor-free/