McDonald’s is leaning on machine learning to recommend what a Big Mac should cost at each of its restaurants — and the results are already visible to anyone who opens the app. At one company-operated location in Fresno, California, the burger sells for $5.69. Two miles away, another corporate store charges $6.89 for the same sandwich: a 21% premium.
The gap is the most tangible output of a pricing engine that analyses millions of daily transactions across nearly 14,000 US restaurants to generate what the company calls “the optimal price” for every menu item, at every location. The story touches on franchising power, consumer trust and the growing regulatory scrutiny of automated pricing.
- Hyper-local pricing: The system estimates how much customers in a given neighbourhood will pay, and the spread between nearby stores has widened.
- Pressure behind the advice: McDonald’s calls the tool a recommendation, yet franchisees say they feel pushed, and the company tracks deviations from its suggestions.
- Regulatory risk: Franchisees can be treated as competitors under antitrust law, making a shared pricing engine a sensitive piece of software.
How the pricing engine works
Reporting that reviewed screenshots of the pricing portal captured in August and interviewed nine people with first-hand knowledge of the strategy describes a tool built on machine-learning algorithms that continuously parse daily transactions and generate location-specific recommendations, from Big Macs to discounted coffee for seniors.
The interface shows messages such as “Your restaurant is showing MEDIUM SENSITIVITY to Price,” based in part on “customer willingness to pay in your area.” The platform also pulls public menu prices from nearby competitors, including Wendy’s and Burger King. Both chains said they do not use AI to make pricing decisions.
The guardrails written around the model matter more than the algorithm itself. According to two former employees of Tiger Analytics, the consultancy that runs the tool, McDonald’s sets the rules: concentrate increases on items that have not been repriced for at least two years, raise a product only if at least 30% of stores have already done so, and keep ice cream and beverages out of summer increases.
Recommendation or quiet mandate?
McDonald’s insists franchisees are free to set their own prices and that the portal is “a tool, not a mandate.” Costs differ between restaurants, the company notes, and two locations a few miles apart can belong to distinct markets.
Five store owners say they were pressured to adopt the AI-generated recommendations. A franchisee document from June records deviations in detail, and from January operators were required to be “constructively engaging with McDonald’s approved Pricing Consultant and Tools” under revised business standards. Recommendations are issued at least three times a year, and CEO Chris Kempczinski reportedly told investors the metric now forms part of franchisee evaluations.
The tension is structural. McDonald’s collects a share of each restaurant’s sales rather than its profit, so higher turnover can benefit headquarters even when an individual operator is losing money. Some recommendations to cut prices have also caused friction between franchisees and corporate leadership.
Denials, precedents and consumer trust
The company has pushed back hard on the framing of the story, issuing a detailed rebuttal:
“AI does not set the price of a Big Mac or any other menu item. It does not change prices. And it does not determine what an individual customer pays. McDonald’s does not use dynamic pricing.”
McDonald’s says the tool focuses on restaurant-level market dynamics rather than an individual customer’s willingness to pay, and it dismissed claims that a standard business practice was being recast as controversial. That denial sits awkwardly beside franchisee accounts of a repeated, tracked and formally evaluated process — a discrepancy worth watching.
Regulators are already circling algorithmic pricing. Research on German fuel stations operating in a duopoly found margins were 28% higher after they adopted pricing software, and US authorities have scrutinised algorithmic rent-setting tools. Wendy’s drew a backlash in 2024 over plans to test dynamic pricing, while Instacart ended a limited AI pricing experiment in December after an analysis estimated a family could spend $1,200 more a year. Franchisees, sometimes treated as competitors under antitrust law, are warned to be careful when using the portal.
“It’s in the interest of the company to know where there’s demand and to get that feedback in real time,” Brooklyn resident Diane Bezucha told reporters. “But if that technology is used to raise prices simply because of more demand, that doesn’t really help me as a customer.”
Pricing is only one front in the company’s AI rollout. On September 23, McDonald’s showed investors Archy, a voice system that takes drive-thru orders in English and Spanish with more than 90% accuracy according to the company, as part of a wider platform called ArchIQ.
The real vulnerability is not technical. A brand built on predictable pricing cannot afford customers suspecting that their bill depends on what an unseen model thinks their neighbourhood will tolerate. The likeliest rival for any McDonald’s is another McDonald’s a few streets away, and a diner who feels overcharged will simply walk there. New York’s “click to cancel” law, in force since October 1, reflects the same instinct. Until pricing transparency catches up, the most practical defence remains the oldest one: compare prices before ordering.