Now Your Groceries See You, Too

Almost everything consumers now purchase is being recorded, measured and analyzed.

The question is whether any of the value being created by these transactions benefits only merchants or whether shoppers receive advantages, as well. Because the answer may determine to what degree this system will continue to grow without regulation. JL

Sidney Fussell reports in The Atlantic:

Demographic information is key to retail shopping. Retailers want to know what people are buying, segmenting shoppers by gender, age, and income and then targeting them. “Cooler Screens” do not use facial recognition. Shoppers aren’t identified. Cameras analyze faces to make inferences about shoppers’ age and gender. The camera takes their picture, which an AI system will analyze. The system can estimate if the person is a woman in her early 20s or a male in his late 50s, making note only of what shoppers picked up and basic information on their age and gender.

Walgreens is piloting a new line of “smart coolers”—fridges equipped with cameras that scan shoppers’ faces and make inferences on their age and gender. On January 14, the company announced its first trial at a store in Chicago in January, and plans to equip stores in New York and San Francisco with the tech.

Demographic information is key to retail shopping. Retailers want to know what people are buying, segmenting shoppers by gender, age, and income (to name a few characteristics) and then targeting them precisely. To that end, these smart coolers are a marvel.

If, for example, Pepsi launched an ad campaign targeting young women, it could use smart-cooler data to see if its campaign was working. These machines can draw all kinds of useful inferences: Maybe young men buy more Sprite if it’s displayed next to Mountain Dew. Maybe older women buy more ice cream on Thursday nights than any other day of the week. The tech also has “iris tracking” capabilities, meaning the company can collect data on which displayed items are the most looked at.

Crucially, the “Cooler Screens” system does not use facial recognition. Shoppers aren’t identified when the fridge cameras scan their face. Instead, the cameras analyze faces to make inferences about shoppers’ age and gender. First, the camera takes their picture, which an AI system will measure and analyze, say, the width of someone’s eyes, the distance between their lips and nose, and other micro measurements. From there, the system can estimate if the person who opened the door is, say, a woman in her early 20s or a male in his late 50s. It’s analysis, not recognition.

The distinction between the two is very important. In Illinois, facial recognition in public is outlawed under BIPA, the Biometric Privacy Act. For two years, Google and Facebook fought class-actions suits filed under the law, after plaintiffs claimed the companies obtained their facial data without their consent. Home-security cams with facial-recognition abilities, such as Nest or Amazon’s Ring, also have those features disabled in the state; even Google’s viral “art selfie” app is banned. The suit against Facebook was dismissed in January, but privacy advocates champion BIPA as a would-be template for a world where facial recognition is federally regulated.

Walgreens’s camera system makes note only of what shoppers picked up and basic information on their age and gender. Last year, a Canadian mall used cameras to track shoppers  and make inferences about which demographics prefer which stores. Shoppers’ identities weren’t collected or stored, but the mall ended the pilot after widespread backlash.

The smart cooler is just one of dozens of tracking technologies emerging in retail. At Amazon Go stores, for example—which do not have cashiers or self-checkout stations—sensors make note of shoppers’ purchases and charge them to their Amazon account; the resulting data are part of the feedback loop the company uses to target ads at customers, making it more money.