The fast-food chain is turning to artificial intelligence and machine learning in the hopes of predicting what customers want before they decide.
McDonald's has a new plan to sell more Big Macs: act like Big Tech.
Over the last seven months, McDonald's has spent hundreds of millions of dollars to acquire technology companies that specialise in artificial intelligence and machine learning. And the fast-food chain has even established a new tech hub in the heart of Silicon Valley — the McD Tech Labs — where a team of engineers and data scientists is working on voice-recognition software.
The goal? To turn McDonald's, a chain better known for supersized portions than for supercomputers, into a saltier, greasier version of Amazon.
As fast-food sales decline across the increasingly competitive restaurant industry, McDonald's is looking for new ways to lure customers. On Tuesday, the chain said same-store sales in the United States were weaker than expected for the third quarter, sending shares lower.
But in the coming years, the company's machine learning technology could change how consumers decide what to eat — and, in a potentially ominous development for their waistlines, make them eat more.
So far, the technological advances can be experienced mostly at the chain's thousands of drive-thrus, where for years menu boards have displayed a familiar array of McDonald's favourites: Big Macs, Quarter Pounders, Chicken McNuggets.
Now, the chain has digital boards programmed to market that food more strategically, taking into account such factors as the time of day, the weather, the popularity of certain menu items and the length of the wait. On a hot afternoon, for example, the board might promote soda rather than coffee. At the conclusion of every transaction, screens now display a list of recommendations, nudging customers to order more.
At some drive-thrus, McDonald's has tested technology that can recognise license-plate numbers, allowing the company to tailor a list of suggested purchases to a customer's previous orders, as long as the person agrees to sign away the data.
"You just grow to expect that in other parts of your life. Why should it be different when you're ordering at McDonald's?" said Daniel Henry, the chain's chief information officer. "We don't think food should be any different than what you buy on Amazon."
As the evolution of the McDonald's drive-thru shows, the internet shopping experience, with its recommendation algorithms and personalisation, is increasingly shaping the world of brick-and-mortar retail, as restaurants, clothing stores, supermarkets and other businesses use new technology to collect consumer data and then deploy that information to encourage more spending.
At some stores, Bluetooth devices now track shoppers' movements, allowing companies to send texts and emails recommending products that customers lingered over but did not buy. And a number of retailers are experimenting with facial-recognition tools and other technologies — sometimes known as "offline cookies" — that allow businesses to gather information about customers even when they are away from their computers.
In the restaurant world, the increasingly popular food-delivery apps have produced a slew of customer data. But much of that information is controlled by third-party technology companies rather than by the restaurants themselves, underlining the importance of tech expertise as the industry grows more competitive.
"A lot of the restaurant chains, the larger ones that have the cash and the clout and the depth, are really turning into quasi-technology companies," said Michael Atkinson, who runs Orderscape, a company that provides voice-ordering technology. "All of them have that ambition."
In recent years, Domino's Pizza has distinguished itself as a technology leader in the slow-moving world of pizza (it's hard to disrupt a crust recipe), aiming to capture the growing food-delivery market with streamlined phone and online ordering systems, data-collection techniques and even self-driving cars.
Like the new McD Tech Labs in California, Domino's also has a tech headquarters: the "innovation garage" in Ann Arbor, Michigan, where teams of employees drawn from departments across the company work on specific projects under one roof — an approach borrowed from Silicon Valley.
"That's 60 years' worth of legacy corporate structure that we have blown up by moving into this building," said Dennis Maloney, the company's chief digital officer. "Domino's started off as a pizza company that sells online, and we've managed to transform ourselves into an e-commerce company that sells pizza."
So far, however, Domino's has stopped short of the latest McDonald's play: acquiring entire tech startups. (Pizza Hut, however, recently acquired a company that produces online ordering software.)
In March, McDonald's spent more than US$300 million ($468.6m) to buy Dynamic Yield, the Tel Aviv-based company that developed the artificial intelligence tools now used at thousands of McDonald's drive-thrus.
The deal "has changed the way the high-tech industry thinks about potential M&A," said Liad Agmon, a former Israeli intelligence official who co-founded Dynamic Yield. "We'll see more nontraditional tech companies buying tech companies as an accelerator for their digital efforts. It was genius on McDonald's side."
Already, the recommendation algorithms built into the drive-thru menu boards have generated larger orders, the McDonald's chief executive, Steve Easterbrook, said during an earnings call in July. (Henry, the chain's information executive, declined to reveal the size of the increase.) By the end of the year, the new system is expected to be in place at nearly every McDonald's drive-thru in the US.
In September, McDonald's purchased a second tech company, Apprente, a startup based in Mountain View, California, that develops voice-activated platforms that can process orders in multiple languages and accents. In recent months, McDonald's has tested voice recognition at some of its restaurants, seeking to replace the human workers who take orders with a faster system.
McDonald's insists that the rollout of the voice technology will not cost jobs. But at a time when it faces renewed protests from workers over low wages and sexual harassment, the chain's new focus on technology could intensify scrutiny of how it treats its workers and how they might be affected by automation. While McDonalds has reported impressive growth over the last couple of years, some employees at its restaurants make less than US$10 an hour.
"Try raising a family on that," said Adriana Alvarez, an employee in Cicero, Illinois, who has helped lead the high-profile campaign for a US$15 hourly wage at McDonald's. "The company should be able to balance tech and other investments and, in the process, ensure workers like me are safe on the job and have a seat at the table."
With unemployment at just 3.5 per cent in the United States, the fast-food industry is facing one of its worst labor shortages in decades. Rather than eliminate jobs, McDonald's claims that voice-recognition technology would allow franchise owners to reassign workers to understaffed areas of their restaurants. But across the industry, fast-food experts say, some chains may attempt to use voice tools and other technologies to replace workers.
"The labor shortage frankly has done more to push restaurants toward technology than almost anything else," said Jonathan Maze, the executive editor of Restaurant Business Magazine, a trade publication. "It enables you theoretically to be able to run your restaurant with fewer people."
At the McDonald's drive-thru on Fort Hamilton Parkway in Brooklyn, every order still must go through a human being: Last week, the voice on the other end of the speaker sounded perplexed when a reporter turned down the free soda that usually comes with a cheeseburger and fries.
But the rest of the drive-thru experience — with its digital screens and recommendation algorithms — does indeed feel a bit like shopping online.
"It's a great, efficient way to take people's money," said Marayah Jerry as she waited at the drive-thru to collect a Ranch Snack Wrap. "I'll come with an idea of what I want, and then I see the pictures, and I'm like, 'That looks good.'"
Another drive-thru customer, Dalila Ruiz, said she noticed the suggested add-ons at the bottom of the menu board but resisted the temptation to splurge. "I don't want to be so fat," Ruiz said.
Not all McDonald's customers are likely to show such discipline. Critics of artificial intelligence have long warned that the technology could lead to a dystopian future in which humans are subordinate to machines.
Before the robot apocalypse, however, AI might simply make us fatter.
"There are real, significant unintended consequences of something like this further driving unhealthy eating and more fast-food eating and obesity rates and diabetes rates going up," said Scott Kahan, a doctor who directs the National Center for Weight and Wellness, an obesity clinic in Washington, D.C. "These sorts of technologies are making it hard for people to just find some reasonable moderation."
There is plenty of precedent for companies like McDonald's finding creative ways to persuade Americans to consume more calories. But the marriage of a fast-food giant and an artificial-intelligence startup marks an unusual new chapter.
When Agmon, the co-founder of Dynamic Yield, announced the McDonald's acquisition in a company WhatsApp chat in March, his colleagues thought he was joking. "When you start working for a tech company," Agmon said, "you don't expect this."
Soon, however, the news began to sink in: The next day, 250 McDonald's hamburgers arrived at Dynamic Yield's headquarters in Tel Aviv, along with fries for the whole staff.
But this wasn't really a McDonald's crowd. By the time the staff finished hugging and congratulating each other, the burgers were cold.
Written by: David Yaffe-Bellany
Photographs by: Gabby Jones and Robert Beatty
© 2019 THE NEW YORK TIMES