The machine learning tool analyzes millions of daily transactions to suggest prices to franchisees, free to follow or not. The result: two Big Macs with $1.20 each other 3 km apart.
McDonald's relies on a machine learning system to recommend prices tailored to each restaurant, the United States and some foreign markets. It is not a generic AI: the tool exploits existing transaction and price data, millions of points a day, taking into account local competition and customer sensitivity to prices. The group has been using such recommendations since at least 2019.
The last word is for the franchisees, who operate 90% of American restaurants and almost 95% of the global network: they can accept or ignore the suggestion. The resulting differences can be marked: in September 2026, two Fresno settlements about three kilometres away charged the Big Mac 5.69 and $6.89.
Prices per restaurant, not per guest
McDonald's insists on a distinction: the tool sets prices at the point-of-sale level, not customer-specific rates. The grade counts, while the FTC proposed guidelines on custom pricing in August 2026. The platform also includes warnings of compliance with competition law for franchisees.
Read the full article on the original source
Go to the source →