MIT researchers published a study showing that algorithms don't always deliver the most optimal or consistent prices to consumers.
MIT’s Sloan School of Management is out with a new study on online pricing showing that prices set by algorithms might not be the lowest or best for shoppers. The researchers, who focused on the e-commerce business of major retailers in the United States, report that continual adjustments result in fluctuating prices across different times and geographies.
"The old saying about buyer beware remains," declared Professor Roberto Rigobon, one of the study leaders. "Consumers have only been making online purchases for the last 10 years and are still learning about these practices."
According to study co-author Professor Diego Aparicio, many consumers recognize algorithmic pricing for online items like ride shares and airline tickets, but don’t have that same knowledge level about the groceries they buy. He cited some examples. “Algorithmic pricing means that the price at Amazon Fresh for the same 12-pack of Coke is different in Miami than in New York. It also means that the price of Coke at Amazon Fresh and Walmart Grocery will be different, even for a 12-pack of Coke being delivered to the same zip code in Brooklyn,” he remarked, adding that price inconsistencies show up more online than in brick-and-mortar stores. “This is surprising because a consumer can easily check competitors' prices, which you would think would lead to greater transparency and convergence of prices. Yet, this only happens occasionally."