Personalization or Price Discrimination?

 
Food&Power-1-30-2020.jpg

by Open Markets Food & Power Reporter Claire Kelloway

Could you and your neighbor pay different prices for the same jar of peanut butter one day?

More and more grocery industry insiders say yes. Personalized pricing was a popular topic at the National Retail Federation’s annual convention in New York City this month. Grocery stores can leverage a combination of data analytics and customer identification and tracking tools to offer real-time individual pricing and promotions, both online and in-store. While the practice may still be in its infancy, some experts believe that personalized prices will become the standard in food retail and beyond. Individualized prices could make goods more affordable for some, but they also raise legal and ethical concerns about price gouging and discrimination.

Sellers have long offered different prices to different groups of people, whether through senior and student discounts or lower airline fares to people who have Saturday layovers. But until recently, retailers rarely had the means to offer different prices to different individuals based on personal characteristics. This could change with the help of tracking technology, data collection, and machine learning. Economists call this first-degree price discrimination, and, if achieved, it allows sellers to extract the maximum amount that each individual is willing to pay.

Grocery chains collect large swaths of data on customers’ location, past purchases, demographics, and even web-browsing history, to predict customer preferences and behaviors for targeted advertising, promotions, and prices. Target, for instance, assigns customers a Guest ID to track their demographics and interactions with the store. These profiles can predict things such as when a customer is pregnant (down to the rough due date) and analyze past behavior, to determine whether an expecting mother is more likely to buy baby supplies if she receives a coupon via e-mail or printed on her receipt.

Such promotions could become so baked into the shopping experience that the former CEO of Safeway, Steve Burd, said “there’s going to come a point where our shelf pricing is pretty irrelevant because we can be so personalized in what we offer people.” In fact, stable shelf prices could even cease to exist. In 2014, the U.K. grocery chain B&Q piloted electronic price displays that connected with shoppers’ cellphones to display different prices based on their loyalty program profiles, offering a glimpse of retail’s holy grail: individualized in-store prices.

Despite all the hype, implementing personalized prices has been a technical and logistical challenge for grocery stores, says Bill Bishop, chief architect of Brick Meets Click. “I think we’re still in very early days in personalization and personalized pricing,” he said. However, Bishop did say that “in 10 years, personalized pricing will be the standard, and it will replace in large measure mass pricing in the grocery business.”

As a starting point, some grocers are implementing individualized prices online. In a presentation for the National Retail Federation convention, Brian Crain, head of global business development at Precima, said his firm helps grocery chains offer “individualized pricing per shopper” online based on past shopping habits and potential value. Say a customer is a loyal shopper but rarely buys from a certain category, such as seafood. Precima could work within a retailer’s given budget to offer that shopper lower seafood prices. In an interview, Crain said that he sees more retailers talking about doing personalized pricing online, and while only a few have begun to implement the practice, he thinks this is “the way of the future.” Precima also offers to sell this personalized data to suppliers to help inform their promotions.

Such one-to-one pricing is still controversial. Studies suggest that most consumers think individualized prices are unfairunless they are able to participate in the pricing process (i.e., to bargain). At the same time, a study by IDC found that 63 percent of shoppers surveyed said “personalization is important” in grocery shopping.

Personalized prices could make goods cheaper and more accessible for those who need them. “Price discrimination can actually help spread out fixed costs … [and] make sure anyone who wants that product gets it,” explains Maurice Stucke, a University of Tennessee law professor and author of Virtual Competition: The Promise and Perils of the Algorithm-Driven Economy.

But retailers may also use data-analytics to squeeze or manipulate consumers in new, sophisticated ways. “Companies can use behavioral biases to profit at our expense,” Stucke says. “Once they compile data about individuals, they can be much more targeted in [getting consumers] to do things that they may not otherwise have wanted and get them to pay at the highest price that they’re willing to pay.”

This could be as benign as coupons for your favorite comfort foods when an algorithm identifies indicators of stress, or, more problematically, gouging consumers who have fewer options. The Wall Street Journal, for instance, reported that Staples charged online shoppers more for goods when shoppers were more than 20 miles from a competing office supply store. These areas with fewer competing stores also tended to be areas with lower median incomes.

Personalized prices may also intentionally or inadvertently discriminate on the basis of race, gender, or other protected classes, which is technically illegal. An investigation by ProPublica found that The Princeton Review charged more for college entry prep in ZIP codes with higher Asian American populations, even in areas with lower median incomes. Overall, ProPublica found that Asians were nearly twice as likely to get a higher price from The Princeton Review than non-Asians.

Both Staples and The Princeton Review claimed that their pricing differences reflected varying costs of running their businesses, but the end results show how relying on personal pricing algorithms can still unintentionally discriminate in racist and potentially illegal ways. Pricing based on factors such as ZIP codes can easily become a proxy for racial discrimination, for instance.

As personalized prices become more commonplace, it will be critical to monitor for unfair or discriminatory outcomes. For now, consumers in competitive markets can take their business elsewhere if they feel slighted or dissatisfied by their personal price offerings. But as the grocery industry consolidates and prices begin to change more frequently, consumers could lose the ability to avoid or even identify price discrimination.

Reporting contributed by Katherine Dill.

Correction, 2/7/2020 - An earlier version of this story did not specify that Steve Burd is the former CEO of Safeway.