Instacart Ends 'Price Testing' Following Pressure From Consumer Reports Study

In an email to Privacy Daily Monday, Instacart said “all item price tests on our platform [will end] ... effective immediately.”

In a Monday blog post, the company said, “While these pricing tests were not dynamic pricing nor surveillance pricing, and were never based on personal, demographic, or user-level behavioral characteristics, we’ve listened carefully to customer feedback, and we understand these tests fell short of their expectations.”

It added, “At a time when families are working hard to stretch their grocery budgets, customers should never have to question the prices they see on Instacart.” This message was reiterated in the email from Instacart's spokesperson.

In the report earlier this month, CR and other groups said its data showed Instacart frequently charged customers different prices for the same grocery items. The company's blog Monday seemed to respond to that charge, saying, “Now, if two families are shopping for the same items, at the same time, from the same store location on Instacart, they see the same prices -- period.”

Consumer Reports isn't mentioned by name in the Instacart blog post. "Some" recent coverage of Instacart pricing "raised fair questions," Instacart said. "Much of it included misconceptions and misinformation about what we do -- and don’t do -- when it comes to prices."

Additionally, Instacart said its retail partners “will no longer be able to use Eversight technology to run item price tests" on the platform. Eversight's AI technology allows experimentation with prices for increased grocery sales (see 2512090044). Instacart acquired Eversight in 2022.

Instacart added, “Retail partners will continue to set their own prices on Instacart,” which may involve them varying “the price of items on a store-by-store basis” on the platform, but Instacart “will not support any item price testing services.”

CR, along with Groundwork Collaborative and More Perfect Union, conducted the study, published in early December, which found nearly 75% of grocery items were offered at multiple price points on Instacart (see 2512090044).

On Dec. 16, the groups called on the FTC and state attorneys general to investigate Instacart, citing a petition with 40,000 signatures and alleging Instacart violates the FTC Act by engaging in deceptive or unfair pricing practices (see 2512190045).

George Slover, senior counsel for competition policy at the Center for Democracy and Technology, said that there are “gradations” to dynamic pricing. Instacart alleges it was not using "very private and personalized data” in pricing decisions and instead was "offering two different prices to two different groups of people” to measure "if there was any difference in customers' willingness to purchase.”

If true, that practice is still "potentially problematic,” so it’s “a good thing" Instacart "pulled back,” he said in an interview. But a step further is using personal data to "build a very intricate and intimate personal profile” about consumers, which Slover calls “bespoke pricing,” as the price is "tailor-made" for the consumer.

Bespoke pricing is "Adam Smith's invisible hand being turned against consumers and picking their pocket,” the CDT counsel said. Though "it's not clear" how much this practice "has actually taken hold,” the "capability is there now to do it, and the temptation for sellers to use it is going to be irresistible,” even if it means going against the pricing practices of the past 150 years, Slover said.

Phil Radford, president and CEO of CR, said they “welcome Instacart’s decision to reverse course and end its secret, AI-driven pricing experiments on unsuspecting shoppers,” in an email to us. “At a time when everyday Americans are struggling with high prices, it is particularly egregious to see corporations secretly conducting individual experiments to charge different prices for the same products to get as much out of peoples’ pockets as possible.

"While it is encouraging to see Instacart end this practice, we urge policymakers across the country to take a closer look at these types of AI-pricing experiments on consumers,” he added.

A post from CR added that Sen. Ruben Gallego, D-Ariz., introduced the One Fair Price Act to prevent the use of personal data in setting prices, following CR's joint investigation into Instacart. Several states have also attempted to address the issue with legislation.

“Awareness is spreading” among legislators concerned about pricing and affordability about “the potential for [bespoke pricing] to occur," Slover said.

A New York state law requiring retailers to disclose when they are using algorithmic pricing -- called the Algorithmic Pricing Disclosure Act -- went into effect recently after a legal challenge from the National Retail Federation (see 2507070042). A federal court ruled Oct. 8 the law did not violate businesses' First Amendment rights (see 2510090015), though NRF appealed (see 2511030024).

California had also attempted to pass a surveillance pricing bill, titled the Fair Online Pricing Act, but SB-259 failed in September (see 2509150026), after drawing scrutiny from business groups and Republicans (see 2507150066). A similar measure, HB-1264, failed in Colorado as well (see 2510010064).

A bipartisan group of senators also called on the FTC to resume a surveillance pricing study that began under former Chair Lina Khan (see 2512180074).

"Surveillance pricing is fundamentally a mistake because of the opportunity costs," said Don Marti, a privacy expert for the vendor Aloodo. In an email to Privacy Daily, he wrote, "Any vendor or internal team who is good enough at surveillance pricing to win at it is also good enough to bid up their own price to capture any extra money that it brings in."

"So on average, it's going to be a wash -- and if you fail to correctly predict consumer work-arounds or state legislatures, surveillance pricing is going to cost more than it earns," he added. "The solid medium/long term alternative is to say no to surveillance pricing experiments, save money, and focus on quality and service improvements, R&D, or efficiency improvements -- which tend to pile up over time (think Chinese solar panels)." Marti expanded on the issue in an October blog post.