McDonald’s is loving machine learning
It’s 30° C in the shade, I’m in my non-air-conditioned Peugeot 207, waiting in the drive-thru line. While I wipe the sweat off my forehead, I notice the illuminated menu display. Only three days ago, neither a McFlurry nor a Sundae were displayed here. «Whatever,» I think to myself, «I’d actually quite fancy an ice cream right now.» After my visit to McDonald’s, I couldn’t stop wondering about the digital drive-thru display. On the Tuesday of my first visit, 72 hours before my second one, the outside temperature was 16 degrees and it was raining. Did the menu display really adjust to the weather? Yes, it did – thanks to machine learning.
It's my turn at last. As soon as the voice from the loudspeaker asks me for my order, I give in and add two McFlurrys. After all, the heat is unbearable. Besides, they’re delicious. I gobble up my ice cream in seconds. Sure, I spent a bit more than I’d planned for, but it’s hardly noticeable.
The all-knowing employee
The same goes for the deal McDonald’s made: with almost six billion net revenues in 2018 only, this 300 million dollar acquisition doesn’t put a dent into the fast-food chain’s finances. Yet, this investment is of crucial importance for the Americans company – just as my McFlurrys are for me.
Machine learning trains a machine how to learn based on the recognition of certain patterns and laws in the data it’s fed with. This allows memorising individual behavioural patterns as well as linking them and drawing conclusions from them. It’s even possible to add missing information or to make forecasts for the future. With the help of algorithms, the machine learns a certain function (in supervised learning) or entire models (in unsupervised learning), which then make it possible to draw conclusions or make predictions about expected behaviour.
I’ll have big data with fries and a coke without ice, please
According to McDonald’s, the technology from Tel Aviv is initially used on electronic drive-thru menus. The software includes factors such as weather, time, local events, traffic jams in front of the restaurant and on nearby roads, currently popular products and even the menu you order. Based on this data, the displays can be adjusted and tailored to customers – as in the above example when the warm outside temperatures made the displays show ice cream. Other examples of how the collected data influences what’s displayed on the displays are, for instance, that quick-to-prepare menus are recommended if there’s a long queue on the drive-thru, or breakfast deals are displayed in morning rush hours.
McDonald’s has invested in such technologies before: last year, the fast-food chain tested this branch of Artificial Intelligence by Dynamic Yield in some of their restaurants. Daniel Henry, Vice President and CIO of McDonald's, plans to equip over 1,000 of their restaurants with the new technology from Israel. After this, the machine learning technology will gradually be rolled out to all 14,000 restaurants in the US. Expansion into other countries is also being considered. At a later point in time, self-ordering kiosks and the app will also be updated to feature this technology. According to Henry, expanding into the kitchen to increase efficiency isn’t ruled out, either.
«Technology is a critical element of our Velocity Growth Plan, enhancing the experience for our customers by providing greater convenience on their terms. With this acquisition, we're expanding both our ability to increase the role technology and data will play in our future and the speed with which we'll be able to implement our vision of creating more personalised experiences for our customers.» Steve Easterbrook, CEO of McDonald’s, about the acquisition of Dynamic Yield.
Dynamic Yield will continue to operate as an independent company – much to the relief of existing customers such as IKEA, Sephora and Urban Outfitters. Nevertheless, the Israelis will focus primarily on their new parent company, as the potential of their technology is greater there than anywhere else. As Henry states, McDonald's could imagine taking the technology one step closer towards a unique customer experience with licence plate recognition. With this, recommendations wouldn’t only adjust to the weather, but also to your personal wishes and needs.
Back to the future
McDonald’s leaves nothing to chance. The recent move of their headquarters to the lively neighbourhood of West Town, Chicago, is intended to attract young talents. McDonald's has been fully committed to technology ever since Easterbrook took over the company: he’s also behind the company's own global app and the partnership with Uber Eats. And he introduced electronic advertising and display boards. These aren’t outstanding innovations as such, but considering McDonald's is a fast-food chain, taking these steps is still somewhat brave. Personally, I think these expensive investments will pay off.
I believe that AI, machine learning and big data will continue to change and revolutionise the food service industry. You’ll soon no longer have to express your special requests with every order: any vegetarian, lactose-free, vegan or kosher preferences will no longer be communicated from you to the service staff. It will all be recorded digitally. Soon your order will be on its way to your table when you’ve only just entered a restaurant.
Will the food service industry treat data protection and my privacy with due respect? Most unlikely. Do I mind that they’re next in line to jump onto the Big Data bandwagon? No. I have no time to worry about it: my pizza Hawaii has been served and is waiting to be eaten. Who cares about data privacy when they’re hungry? Maslow's hierarchy of needs knows this, too.
And what’s in it for me? Personally, I’ve never liked pickled gherkins – especially when they’re warm. If my next Big Mac comes without gherkin, I know now that this is anything but coincidence.
Great news for me.