As a VP, Ed Lee leads the global menu strategy team at McDonald's. In this role he drives the growth of core categories and brands, creates future growth platforms and builds food marketing and development capabilities across McDonald's. Most recently, he also led the global strategy insights team where he leveraged consumer insights and business analytics to find new growth opportunities for McDonald's. Prior to McDonald's, Ed spent 12 years at Kellogg. Could you ground us in your perspective and tell us about your journey and roll at McDonald's and the teams that you lead. >> Absolutely. So I got my training in consumer insights in the CPG industry. And CPG is famous for understanding consumer and building capabilities and technologies to be able to do that. And I'm happy to be able to apply some of that at McDonald's. The industry has changed so much with this explosion of data that we've experienced in the last five years. It used to be good enough to just do some surveys or some focus groups, and now with all of the data and information that's available we really need to have new tools and processes to analyze all that data. And we're trying to bring those to bear at McDonald's. So when I came to McDonald's that was one of the things I wanted to try to do was to bring new tools, techniques, technologies, capabilities and talent into McDonald's to deal with this new world. >> So broadly speaking, what is the biggest value driver for McDonald's guests related to AI applications? Potentially, what are two or three outcomes that McDonald's customers can look forward to at McDonald's because of AI or machine learning models? >> Well the first one, Mary, is personalization. The more we understand about the customer, and we have a lot of customers come through. We do about 60 million per day. And the more we can leverage that big data to understand each individual customer, the more we'll be able to offer a personalized experience that's better for each individual. The second thing is around customer experience and customer satisfaction. If you go to McDonald's you might have been asked to fill out a receipt or a survey on a receipt and and we do about 60 million of those a year as well. And with all of that data, we're able to use AI to mine the information and find opportunities to improve the customer experience. And so if we can offer a more personalized experience that are more enjoyable for customers and we can understand where we're having some shortcomings and we can shore those up, we should be able to grow the business. >> Well, that's an explosion of data and that goes along with the explosion of evidence that AI applications across industries are happening. So I understand that something very unique is happening at McDonald's as you are applying both AI and robotics to improve the all important drive through experience. So could you potentially describe what and how the guest experience is benefiting from this type of innovation. >> Yeah, sure. The drive thru is very important to McDonald's and it went from about 50% of our sales to over 60% of our sales during the pandemic. It's been a very important channel for us and so we have a lot of customers that come through the drive thru. What we've done is a couple of years ago we acquired a company called dynamic yield whose experts at personalization. And they have been primarily doing work in the website area, and what we've been able to do is leverage that into the drive thru. So when you come into the drive thru, the machine will take a look at, okay, what's the weather out today? What have the last ten cars ordered? What do people usually order at two o'clock in the afternoon on a Tuesday? And can pull that together and offer real time suggestions. So as you start ordering, also, hey what goes with a double cheeseburger? And those suggestions will appear in real time on the menu board? And that has been a real success story for us in driving up sales and increasing customer satisfaction. >> And now with AI a lot of companies are using natural language processing. Is this being used at McDonald's and potentially which business decisions are being impacted by using natural language processing? >> Yeah, we are as a matter of fact using using NLP in a lot of different areas of the company. An example, just going back to this customer satisfaction with the receipts, we can mine the real text. If you go in there you say, okay, scale of one to five, how was your experience? But then type in comments. And if you have millions and millions and millions of comments, it's really hard to look through that and understand what's happening. But we can use AI to mine sentiment so we can understand how customers liked a particular, maybe new product that we've launched. We can also understand whether our new safety protocols that we put in for COVD are giving customers a good experience or not. And then based on that, we can do that in real time, we get that data in real time. We can do that in real time and provide very, very quick feedback to all the way down to individual restaurants. So that we can improve and leverage that to offer a better experience and grow sales. >> So AI is actually it's not only helping consumers with personalization, but helping the operations at each individual store where you're giving real time feedback through these different tools. So this leads me to a broader question, how is McDonald's approach to AI different than five years ago? >> Yeah, it's changed a lot. I would say, Mary, that five years ago it was very ad hoc. So it was a new capability, new technology, certain parts of the company we're able to leverage more than others, but it was very new. And then we started to experiment more and now I would say that AI is something we're going after with a real strategic intent from the top down. We've formed a data and analytics hub in the center, and the idea here is to put in place a center of expertise with real capabilities, advanced knowledge and understanding. We've got some centers that we got through acquisition in both Silicon Valley and in Tel Aviv. And then out of the spokes of those hubs are practitioners in different parts of the business, different countries, different functions. And those practitioners are going through and saying, okay, well this year these are the different things I want to be able to do. And we're able to put the right attention, resources, focus in the highest priority areas of the company. So it's very much a normal business practice that we're taking with strategic intent, just like being faster in the drive through or having more delicious food. >> Okay, so we've talked about how AI is applied to specific business problems like the drive thru and in innovation. Well, innovations in the drive thru as well at McDonald's and types of AI technology that are being used like natural language processing. So now could you talk a bit about when do the business units pull in or connect with the data scientists, bring them into that customer business problem our guest business problem? So how do the technologists and the subject matter experts work together to innovate at McDonald's? You talked about the center of excellence, but maybe you could talk about those two groups. >> Yeah, sure, absolutely. So in my department, for example, which I run this global insights group. But you could substitute the global insights group in in France or the operations team in Germany for those. And typically what they'll do is they'll bring business problems. I've got this business problem and I'm either trying to grow sales or have better promotions or understand my business a little bit more. And so I want to be able to look at historical data and understand what the impact was of this new product. Or hey, I launched this new product, at the same time I launched delivery, at the same time I had this promotion going, how much of the sales are coming from each one? And they'll bring those forward and either somebody from the center of expertise or somebody from the business who has links into the center of expertise will be able to tackle that problem and leveraging AI tools. There's so much democratization happening. There's brand new tools and platforms that are available in the market that are helping us to be able to arm folks out into the business units with the kinds of things that they're going to need to run their own analyses and that's fantastic. What we don't want to have is one group that does all the work, what we want is one group of experts that helps the rest of the company to do their work. And so that's the kind of thing that we're trying to put in place now. >> So speaking about a tool, can you highlight potentially the key elements to building out good A/B testing tools and models, and what your team needs to consider when using advanced AI and machine learning methods with A/B testing? >> I'm glad you asked that, because A/B testing is a super important part of working with AI. And it is something that should just be natural, it should be something that just sort of is going along with putting an AI project in place. Because of the speed that you can obtain by using A/B testing very, very quickly, very real time. You can you can generate improvements on a daily basis or maybe even hourly basis. And many people don't know that when you go on the news website, all those headlines have been A/B tested to drive the most clicks possible. And so what we want to try to do is bring that into the physical space. And so we're doing that with our outdoor digital menu boards. You mentioned the drive thru earlier, when you drive through that big canvas of an outdoor digital menu board with those beautiful pictures of our food, you can do a lot with A/B testing there. In fonts, in size, in placement. We could run a test that says, okay at three o'clock in the afternoon we're going to give more real estate to coffee, less real estate to other items because people need that afternoon pick me up. So we can do all of these kinds of A/B testing and very quickly determine all the way down to a restaurant level what the design of that opportunity in the menu board should be. And in order to drive sales for the for the company. So it's a tremendous tool and and we use it rigorously in all of our different applications. >> So in your opinion, if the founder of McDonald's, Ray Crock, was alive today, what would be the one AI or machine learning innovation that would wow Ray Crock? I know he was very famous for his shake machine, but what would wow him now about AI and machine learning? >> Well, we still do have that shake machine at the headquarters. It's a revered gem. But one of the things that we're working on and testing right now actually is voice automation in the drive thru. Basically having an AI bot be able to interact with the customer and take the customer's orders as they go into the drive thru. And what that can do is if we connect that to the offers we're showing on the menu board or with the loyalty program. A person could drive up in their car, they can be greeted by name. Hello, Mary, thank you for coming to McDonald's, and hey, looks like you have 200 loyalty points left. Would you like to order a cheeseburger? These kinds of things offer a really interesting personalized experience. The other thing is that bot can speak multiple languages, or if we're running a Disney promotion, maybe it's in Darth Vader's voice. And so those are the kinds of things that we're able to do now. And I think that sort of innovation and technology probably would knock Ray's socks off if he had a chance to see it. >> I think so as well. And I want to thank you so much for your time today. It was great getting to know you and McDonald's use of AI on the business. Thank you. >> Okay, thank you very much as well.
