McDonald's (NYSE:MCD) recent acquisition of Israeli retail insight specialist Dynamic Yield gives it the ability to use artificial intelligence and machine learning to upsell items to customers at the drive-thru.

In the following segment from Industry Focus: Consumer Goods, host Jason Moser and Fool.com contributor Asit Sharma discuss how McDonald's may benefit in unexpected ways from the data Dynamic Yield will gather and crunch.

To catch full episodes of all The Motley Fool's free podcasts, check out our podcast center. A full transcript follows the video.

This video was recorded on April 2, 2019.

Jason Moser: We wanted to talk a little bit about McDonald's, the golden arches. They recently made an acquisition, a company called Dynamic Yield. If you're thinking that doesn't sound like a restaurant company, then you are not too far off. It doesn't really sound like a restaurant company. It's a bit more of a tech company. They use AI to deliver better customer experiences. Asit, how about breakĀ this deal down for our listeners?

Asit Sharma: Sure. Dynamic Yield is a small company based in Tel Aviv. It has some $80 million of venture capital already invested in it. It has clients. It's an operating company that delivers decision logic technology to retailers. What this is, essentially it's, as you mentioned, Jason, it's using artificial intelligence and also machine learning to provide insights.

McDonald's is going to use this technology to replace their vaunted human version of this technology, which is the upsell. Would you like fries with that burger? Which has been the go-to upsell technology for McDonald's, probably for the last three decades. This small company is going to help McDonald's churn data based on a number of factors. Where it will be focused initially is the drive-thru. As many people who have invested in McDonald's probably already know, the company drives the majority of its revenue from the drive-thru. We've seen McDonald's go through the cycle of replacing static menu boards with digital menu boards, where they have more control over what the customer sees. This is the next step in that evolution.

The company, Dynamic Yield, is going to use algorithms and machine learning to analyze not only past ordering patterns from a particular location -- this has been tested for the last year or so in a location in Miami. It's going to incorporate real-time data like weather, which will help it suggest add-on items to consumers who are coming through that drive-thru.

Jason, back to you. What are your initial thoughts on this out-of-the-box investment for McDonald's?

Moser: Back to the interview we had at the beginning of the show with Aaron, we talked a lot about robots working with humans in perfect harmony and complementing each other. It seems like more markets, more industries are going this way, figuring out ways to incorporate technology into their models to make the experience better.

I think about the Panera across the street here from Fool HQ. I'll swing by there sometimes to grab a little lunch. The kiosks there in the store are usually pretty intuitive in trying to upsell me something to go with a salad, whether I want a drink or a cookie for dessert or wherever. So I would imagine over time, McDonald's is certainly trying to rely less on people and more on technology. If you think about it, these stores probably at the end of the day could be run by computers, with just one person overseeing them. I think it makes sense.

I don't know. I'm not the biggest fast-food guy in the world. I don't ever really go to McDonald's. But I feel like maybe it would be worth going to one sometime just to see what the experience is like now. I think the last time I went to McDonalds was probably 20 or 30 years ago, maybe even longer.

Sharma: I think I've admitted this on this show a couple of years ago, but I never really kicked the McDonalds habit. I've become a much healthier person. I still succumb and sometimes on the sly -- I mean, I end up telling my wife, "Yeah, I took the kids to McDonalds after school." [laughs] "I didn't take them for that healthy snack." I'll still do that.

What's interesting to me about this deal -- if you get a chance, listeners, there's a great Wired article which summarizes this deal and provides some insights. Wired got a brief exclusive interview with McDonald's CEO, Steve Easterbrook. I want to read a quick quote. Easterbrook said, "How do you transition from mass marketing to mass personalization? To do that, you've really got to unlock the data within that ecosystem in a way that's useful to a customer." That really resonated with me as for what the potential of this might be, although it's going to take, let's admit, a few years before you might see something in an earnings report from McDonalds that says, "Because of this technology, we gained a few points of margin."

But what's interesting is machine learning applied to big data. I'm a fan of this. This is something that's mentioned in the Wired article. Sometimes the insights that artificial intelligence gleans from data is totally counterintuitive to how human insight works. I want to give a really brief example. If any of you play chess out there -- and even if you don't, I think it will be easy enough to follow -- Google has a program called DeepMind AlphaZero. This is a chess engine which basically taught itself how to play in about four hours, and then proceed to play a million games against itself. This natural learning, machine learning type engine then faced off against the reigning chess engine, which is a brute force engine which basically calculates how to play based on analyzing a lot of moves much faster than a human could calculate. So, on the one side, this natural learning, machine learning engine; on the other side, sheer, like, calculatingĀ 10, 20, 50 moves down. They pitted the two machines against each other. What was interesting is that AlphaZero, the program that Google developed, had some insights which were counterintuitive in an amazing way, even to grandmasters. For example, we have this principle in chess that you should protect your king at the beginning of the game. AlphaZero showed a willingness to leave its king out in the open if it meant it could move some other pieces more aggressively. Stuff that a human would just not do.

I think when you apply this to what McDonald's is doing, this won't happen overnight, it's not going to be something that totally upends their business model, but I think they're going to glean some small insights. Just dreaming up an example, which is completely random, it might suggest, "Offer a customer a hot fudge sundae on the coldest day of the year," something that a human manager wouldn't say. "Hey, offer people hot cocoa," is what the human manager would say. But we can expect some of these to happen, some of these insights to emerge, and McDonald's to capitalize on it now.