AI-powered personalised pricing could allow retailers to charge shoppers different amounts for identical goods, threatening the universal price tag that has shaped consumer shopping for more than a century, experts have warned.
Companies are developing tools capable of analysing personal data, including browsing habits, income, location and family circumstances, to estimate the maximum amount an individual may be willing to pay.
Lindsay Owens, chief executive of the Groundwork Collaborative think tank, said the technology risked overturning the principle of a posted price, which was introduced in the 1860s by American retailer John Wanamaker.
“The posted price has been paramount for consumers, and I do think the price tag is at risk of extinction,” Ms Owens said. “Predictability is really essential if you’re on a budget, and knowing how much something is going to cost from one week to the next is part of what the posted price offered.”
How AI personalised pricing could work
Retailers have long used dynamic pricing to respond to factors such as supply, demand, weather and the time of day. AI could take that practice further by allowing businesses to continually process large volumes of consumer data and set prices for particular shoppers.
Rather than pricing an item according to its value or a broadly applied market rate, companies could use consumer profiles to calculate what each customer is likely to pay. Ms Owens described the potential result as a form of AI-enabled haggling, with prices changing according to information largely hidden from shoppers.
“Technology is allowing companies to stop setting prices for products and start setting prices for you,” she said.
A trial by grocery-shopping platform Instacart prompted concern after research groups reported that its system had shown different customers different prices for the same goods from the same shop. The investigation said identical grocery baskets could cost some shoppers up to 23 per cent more, potentially amounting to about 1,200 dollars a year for a family of four.
Instacart disputed the findings, saying the 1,200-dollar difference reflected an “atypical” basket in which every item had the greatest possible price gap. It also said it did not use personal data, including protected characteristics such as race or gender, to set prices.
The company said in December that it would stop using AI to allow retailers to charge different customers different prices for identical items on its platform.
Ms Owens said other businesses had begun testing individualised pricing based on data such as online browsing, income, ZIP code and family structure. Research by Groundwork Collaborative also alleged that Walmart had applied for patents covering technology that could create detailed shopper profiles and set individual prices.
The group said electronic shelf labels could potentially allow one customer to be charged a different amount from another shopper standing in the same checkout queue. It also claimed one patent application indicated that prices could be adjusted according to items already in a customer’s basket, such as offering a different price for mayonnaise when tuna had been selected.
Walmart denied developing tools that would use customers’ personal data or other factors to set prices. Dan Bartlett, the company’s executive vice-president of corporate affairs, said Walmart would not use an individual’s income, shopping history, urgency or willingness to pay to determine a price.
“We do not engage in dynamic pricing, raising prices in response to a hot afternoon or an approaching snowstorm, for example,” he said.
Jay Zagorsky, a senior lecturer at Boston University’s Questrom School of Business, said retailers were unlikely to publicise personalised pricing because of the risk of a consumer backlash. He nevertheless said businesses appeared to be developing the tools needed to set individual prices.
“The days when the price you saw stayed fixed for a relatively long period of time appear to have passed,” Mr Zagorsky said.
