McDonald’s Embraces Artificial Intelligence for Fast Food
By now you’ve seen the big news about the world’s largest fast-food chain by revenue. It was a bit shocking that it took this long to happen, but McDonald’s (MCD) is finally embracing advanced technologies like artificial intelligence to boost sales. The guy credited with the turnaround at the 65-year-old restaurant behemoth since 2015, Steve Easterbrook, was behind the technological push that saw McDonald’s not only embrace digital innovation but acquire two AI companies this year.
This is big news for a few reasons. Before this year, it had been nearly two decades since the company made its last acquisition of a mediocre fast-food chain called Boston Chicken. Most of the technological innovation before Easterbrook involved somewhat questionable food science in the form of the McRib sandwich. Or extreme (and inadvertent) advances in food preservation technology, such as the McDonald’s hamburger and fries on permanent display with a live feed that shows them in still near-perfect (if a little pale) condition after the last restaurant closed in Iceland about 10 years ago to the day:
But one of the biggest reasons for why you should care that McDonald’s is making technology core to its business is that it is one of the loudest signals to date that AI truly is the new electricity that will power the global economy in the coming decades. It’s very similar to what we wrote last year about how Coca Cola (KO) is using big data and AI to grow its business, from AI-powered vending machines to social media monitoring to decide which brands or products it should focus on.
A Dividend Growth Aristocrat
Both McDonald’s and Coca Cola are among the 57 so-called dividend aristocrats, an elite group of companies that have increased their dividends to shareholders for at least 25 years in a row. As we’ve noted before, our own money is heavily tilted toward this kind of dividend growth investing (DGI), as a sure and steady way to grow your nest egg. In the case of McDonald’s, the company paid its first dividend in 1976 and has increased it every year since then.
In the company’s most recent third-quarter earnings call in October, Easterbrook reported an 8% dividend increase. If you’re living off dividends, that income growth well exceeds inflation which means you’re able to live a better quality of life. Such increases happen because the company is becoming pretty optimistic about the future, having seen 17 consecutive quarters of global comparable sales growth. (Despite this great performance, Mr. Easterbrook was recently shown the door because he dipped his pen in the company ink. We’re not sure how things are back in the U.K. where Mr. Easterbrook hails from, but here in ‘Murica, if you look at someone for more than 5 seconds the HR harpies can kick you to the curb. It’s Like 7th Grade on Steroids.)
McDonald’s Acquires AI Startups
An increasingly important part of the company’s velocity growth strategy is the “digital reshaping” of its interaction with its customers. In March, McDonald’s plunked down $300 million to buy an Israeli startup called Dynamic Yield, followed by another acquisition in September of Silicon Valley-based Apprente. Let’s take a closer look at both startups and what they mean to McDonald’s.
Founded in 2011, Dynamic Yield had raised about $105 million before McDonald’s gobbled it up earlier this year. The company is one of many AI startups that provide platforms claiming to personalize a customer’s journey from beginning to end by leveraging big data and machine learning for product and content recommendations. McDonald’s is applying the technology at the drive-thru where the digital menu customizes its offerings based on a variety of factors, from weather and traffic to trending product orders and previous order history. Dynamic Yield was mentioned by name 15 times during October’s earnings call, with Easterbrook riffing on just
how easy it is to fool customers how well the technology works:
The beauty of this is there is nothing that our customer has to adjust to. They almost don’t know that this experience is happening for them, as we’ve got dynamic digital menu boards, and effectively as they start to place their order, the menu boards respond to that ordering process and therefore are more likely to suggest items that customer will want and less likely to show items that customers less likely to want. And, of course, machine learning helps you improve that, particularly given the transaction levels we have across our business, we can learn pretty really very quickly.
The technology from Dynamic Yield has already been implemented at 9,500 U.S. drive-thrus and is expected to be nationwide by the end of the year. It appears to be part of a larger strategy by McDonald’s to completely automate the drive-thru experience, where an estimated 50% to 70% of fast-food sales are made.
That explains its second acquisition, Apprente, a two-year-old startup that had raised $4.8 million for its voice recognition and conversation technology. Yes, a voice-activated chatbot will soon be taking your Big Mac order to speed things up (and maybe shed a few employees along the way). McDonald’s currently has one of the slowest drive-thrus among its fast-food cohort, though car idling times throughout the quick-service restaurant industry have been climbing for years, slowing down by another 20 seconds in 2019, according to QSR Magazine.
Fast Food Industry Embraces AI
McDonald’s isn’t alone in its rush to analyze and automate its operations with AI. It’s certainly not the leader. The unofficial title actually goes to Domino’s Pizza (DPZ). For example, a couple of years ago, the pizza chain introduced its own voice assistant tech in partnership with Nuance (NUAN), a leader in voice technology software. More recently, Domino’s rolled out a computer vision technology in its Australian and New Zealand stores that uses cameras to quality check every pizza to ensure the cheesy pies have been made correctly and with the right amount of ingredients. The company is also reportedly ready to start deploying delivery robots in Europe, as part of a deal with Starship Technologies. And, by the way, Domino’s has seen its value rise dramatically in the last five-plus years:
Good Times Restaurants (GTMI) is another player in the fast-food industry that is using voice assistant tech to
replace supplement its human labor force. Meanwhile, Yum China Holdings (YUMC), which owns KFC and Pizza Hut, is taking advantage of China’s cashless, mobile happy customers to harvest data about their buying behaviors from loyalty programs. The algorithm then customizes a menu for each diner based on preferences and local tastes, according to an article in Bloomberg. The news report said that Yum’s AI-powered menu boosted average per-order spending by 1% in just two months, which comes out to $840 million worth of fried chicken and pan pizzas each year. That’s no small fry.
AI Companies Catering to the Fast-Food Industry
Most of the technologies behind these fast food innovations are from startups or even (teeny tiny) small-cap companies that are trying to cater to the fast-food industry.
Take Denver-based Valyant AI, the startup behind the AI conversational platform at Good Times. Founded in 2017, Valyant has raised about $3.2 million to develop and deploy Holly, an automated system that takes customer orders and inserts those orders into point-of-sale systems. Valyant’s AI-based voice assistants are built using actual customer recordings in order to ensure accuracy and eliminate AI bias, according to the company, which claims, based on the Good Time pilot in Denver, that Holly has an average attempted upsell rate of 43%, which puts it in the top third of the industry.
And remember those cameras tracking pizzas over at Domino’s? Dragontail Systems (DTS) is a publicly traded Australian company that claims to use AI to automate the entire fast-food operation, from food preparation to delivery. For instance, the Algo platform helps predict customer flow to help organize food prep, while also optimizing food-delivery routes based on traffic and other factors. The company claims Algo can run 1.2 quintillion (1.2 followed by a lot of zeros) different dispatching options, even on outdated hardware.
Apps are certainly where it’s at these days, and McDonald’s has also invested in its mobile technology by putting about $3.4 million into a publicly traded New Zealand company called Plexure (PLX) in April. Plexure supports McDonald’s customer engagement app in 48 countries. In Japan, Plexure is testing an app feature for McDonald’s that initiates order preparation once a customer enters a certain “geo-fenced” area where they’ll be served. The investment by McDonald’s comes with a big caveat: Plexure can’t provide its technology to any of McDonald’s competitors, though a deal with White Castle was already in place (and, hey, it’s White Castle).
In the last few years, we’ve witnessed the increasing automation of retail, from automated checkout systems to shoplifting prevention, so we’re finally seeing the fast-food industry make similar moves in what is probably a slightly more complex environment for machines to maneuver. And while McDonald’s may never employ robots to flip burgers, there’s definitely motivation to embrace even more types of technology. For example, McDonald’s outright says that “delivery is another area where we’ve moved rapidly to capture changing consumer habits around service and convenience.” The company expects delivery orders to drive $4 billion or roughly 4% of its global sales, which is up from $1 billion just three years ago. Maybe a drone-delivery startup would be in the cards for McDonald’s now that they have a taste for tech startups?
Toward feeding that appetite, the fast-food company created the McD Tech Labs, which it hopes will give it a presence in Silicon Valley, partly fueled by the acquisition of Apprente. Easterbrook stressed the importance of the move by saying that being tops in voice technology is a “must-win area” for the company. Let’s hope McDonald’s can carry on that vision without Easterbrook, using technologies like machine learning to grow that dividend like our future quality of life depends on it – because it does.