How Kroger’s “Surveillance Pricing” and In‑store Tech Threaten Shopper Privacy
Google made billions off “surveillance capitalism,” where exhaust data was extracted from searches and sold as a predictor of behavior to advertisers. Kroger has discovered immense profit in “surveillance pricing,” where personal information is collected in order to determine the highest price that you are willing to pay for individual products. New tools like electronic shelf labels (ESLs) and facial recognition widen the reach of surveillance pricing.
Kroger maximizes profit during your trip around the store through “surveillance pricing,” then profits from “surveillance capitalism” by selling all the data collected about you to the highest bidder. Imagine you suspect you might be pregnant but aren’t sure, and you pick up a prenatal vitamin instead of your usual multivitamin. The store uses facial recognition to determine that you are a young woman of childbearing age and ads for pregnancy products start appearing in your email, social feeds, and on your smart TV that same day.
Every time you enter a grocery store you enter a system that tracks, analyzes, shares, and nudges your behavior. Retailers combine shopping history with data from brokers, web activity, and app use to infer age, gender, race, income, family status, health, and other lifestyle traits. Those profiles are used to push prices, personalized offers, and to sell “insights” to brands.
Surveillance pricing depends on huge data sets and algorithms that try to “determine the highest price” a shopper will tolerate. This is different from classic dynamic pricing that reacts to supply and demand; personalized or surveillance pricing uses individual-level signals to segment shoppers. The result is different customers paying different amounts for the same product based on inferred willingness to pay.
There are concrete examples that show how granular pricing can be. Target charged $100 more for a TV on its app when the consumer was in the vicinity of a Target store versus farther away, and some platforms have shown willingness to charge more to certain devices or browsers. Dynamic examples include surge-priced rides during storms and ticket platforms that have listed more than $22,000 for a Taylor Swift concert ticket in high-demand situations.
Even brick-and-mortar grocers are moving in this direction, and lawmakers have flagged Kroger for potential use of ESLs and facial recognition technology; the company has drawn attention from and the media. Kroger has said it has no plans to identify customers’ faces at digital displays, but the chain has used ESLs for years and expanded partnerships with digital screen vendors. Kroger insists the sensors do not collect or store shoppers’ data and that Kroger does not seek to identify customers through facial recognition technology—only to detect customers’ presence, dwell time, and door opens.
https://x.com/RepRashida/status/1846281168532263372/photo/1
ESLs replace paper tags with small screens that change prices in seconds, so a shopper can see one price on the shelf and another at checkout. Retailers argue ESLs lower costs, but the same tech can raise prices quickly by time, location, or customer signals. Walmart plans to install ESLs in 2,300 stores by 2026, showing how fast the hardware rollout is moving.
Vendors like Cooler Screens install digital cooler doors with embedded cameras and sensors, and some early deployments promised facial detection to infer age and gender for tailored ads. Kroger expanded partnerships with such vendors in 2023 while denying it seeks to identify customers for pricing. Whether or not explicit face ID is used, the hardware and data flows expand the capacity to profile shoppers at scale.
Retailers can infer highly sensitive traits from purchase patterns and third-party data, sometimes creating categories like “reluctant gamblers” or new parents who are seen as less price-sensitive. Those labels feed targeting models that determine which promotions or prices to offer whom. The practice can lead to customers unknowingly paying more while others get targeted discounts.
One everyday example shows how uneven discounts work: two similar shoppers buy different detergent brands based on history, and the store sends a discount to one to switch brands while withholding it from the buyer already loyal to the higher-margin product. That skews who pays more and who gets the bargain, and it cultivates a sense of price unfairness.
Kroger also monetizes this data beyond pricing by selling shopper “insights” to brands, including names, addresses, phone numbers, purchase histories, inferred health details, device data, and demographics. Those new advertising revenue streams create incentives to collect and trade as much personal data as possible. The commercial value of customer data means self-regulation is unlikely to stop the practice.
Policy responses are emerging: retailers have faced enforcement and settlements, including a $5 million settlement involving a major chain accused of deceptive pricing practices, and proposals in some states to ban surveillance pricing. Strong data minimization rules that limit collection, use, retention, and transfer to what is reasonably necessary for the product or service would reduce the raw material that fuels surveillance pricing.
Facial recognition adds distinct harms: false positives, disproportionate impacts on people of color, private watchlists, and opaque removal processes. The FTC has taken action against retailers whose use of biometric tech led to misidentification and unfair outcomes. Given the sensitivity and permanence of biometric data, legal limits on commercial facial recognition and robust notice and redress requirements are necessary.
Grocery chains cannot be relied on to police how their data and screens are used when profit is on the line. Binding rules on data minimization and clear limits on biometric uses would reduce the power imbalance between shoppers and retailers and curb the most invasive forms of surveillance pricing.