Put Profit on the Menu: AI for Restaurant Operations

Meet the Panel

• Clinton Anderson, CEO of Fourth
• Jerry Phillips, VP of Technology, Whataburger
• Drew Brodnax, Director of Operational Finance, Brinker (Chili’s)

Summary

Clinton introduces Fourth iQ and how it tackles key challenges like rising costs and regulatory pressures by enhancing forecasting and labor efficiency. Jerry and Drew share real-world insights on how AI is driving efficiency and boosting customer satisfaction at Whataburger and Chili’s, along with practical advice for those looking to get started with AI.

Transcript

Clinton Anderson, CEO, Fourth
My name is Clinton Anderson, CEO at Forth. I’ve been at the company for five years and it’s been an amazing five years. I think what the restaurant industry has been through, starting with Covid, coming out of Covid, technology continues to change, our managers’ job, jobs getting more and more complicated, right? The world of compliance getting really tricky, especially if you’re operating in places like Philadelphia, New York, California, Oregon, increasingly even places like Denver, Chicago.

And profits are becoming harder than ever to to, to generate. And so in the milieu of all this, this change, we also have, AI. And that’s what we want to talk to you about today is AI. But rather than me pontificate on what I think is happening in the future, we invited two restaurant executives to join us.

And what’s cool about these two guys? First of all, I love them on the marquee. I put them their names on the marquee, and you’re like, we get standing room only, like, you guys are going to get repeat invitations to join us up here. What’s cool about these two guys and look in full cander, they are customers. I know them well, I like them and I think they the way they approach technology in restaurant operations is really insightful.

So what’s curious here is you’ve got two guys completely different restaurant operations, completely different jobs, yet trying to achieve the same goal. So let me introduce my guest today. We’ve got Drew Brodnax, who is the head of operational finance at Chili’s. You say “what’s operational finance?” His job is working with all the restaurant operators across all the geographies where Chili’s located to optimize store performance, and that can be technology oriented.

It can be process oriented. It’s like, how do we drive improved operation, improve performance? And curiously enough, as I talked to drew about how he thinks about the world. Technology plays a critical part. Sitting next to him is Jerry Philips. Jerry is vice president of technology at Whataburger. Okay, so he’s got more of the technology job, yet they think about the world in a very similar fashion.

Right. Jerry is using technology to transform Whataburger or any water burger fans here. Yeah, this Texas God blessed to have a lot of chili fans too, right? But Whataburger is deeply loved here in the state of Texas. But if you’ve ever been through the drive thru on a busy afternoon, you’re like, what are they doing in that damn store, right?

I’m willing to wait for a good burger. And if Jerry’s got doctor Pepper shakes on the menu, the seasonal items, I’ll wait a long time. But it’s super frustrating, right? Because you just know that look, technology’s not quite working the way it needs to be. And Jerry, for the last look, he’s been there for what? How many years.

Jerry Phillips, VP of Technology, Whataburger
Now? Ten years.

Clinton Anderson, CEO, Fourth
A decade. But I would say in the last three years, he and Joe, the chief technology officer, have been very aggressive in pushing a change agenda in an organization that was pretty happy with some really great success over the last 50 years. And so you’ve got two leaders coming together, different worlds, different restaurant concepts and operations, but similar thinking.

And so as I’ve been in meetings with them individually and we thought, who should we bring onto this forum? These guys seem perfect. So we’re really grateful to have you here today. And I’m going to talk for about maybe ten minutes teeing up this concept, and then we’re going to turn the time over. Then we’re going to facilitate conversation with them, asking questions on their experience with AI.

And then look, we’ll open it up to the group as well. If you’ve got questions or ideas, things you’re concerned about, things you’re interested in, we’ll let you participate as well. But I promise I’ll try and go through this next content pretty quickly. All right, look, at the end of the day, we love the restaurant industry. For many of you, the restaurant, just the atmosphere, the vibe, the the excitement of running a restaurant flows through your blood.

But we this isn’t a not for profit, right? We do this to make money. And I’ve worked in a before I became involved in technology. I was consultant for 20 years of my life and I worked in probably 20 different industries. And there’s only one industry where it’s harder to make money in restaurants. You might want to guess what that is — airlines.

But congratulations, you picked the second hardest industry on planet Earth to make money. But people who love airlines that people love restaurants love airlines and restaurants. So I get it, but we’ve got this increasing challenge, right? Because look, profitability is hard for a bunch of reasons, right? Rising food costs. Right. Rising labor costs, shortage of labor. More turnover than we’ve ever seen before.

Not just in our hourly workers, but also in the management ranks. Our managers lives are more difficult. They’re dealing with not only rising food costs, food shortages in some categories. They’re facing, with this world of increased compliance, how many here have have operations where compliance is a real issue, right. If you’re coming in from states, Oregon, California, cities like Chicago, Philadelphia, New York, and even if those of us who operate in places like Florida, Alabama, Texas, where it’s a lot easier from a compliance perspective, that world is probably coming our way.

And so I would argue that it’s never been more difficult to achieve top end operating margins. The good news is, despite these challenges, if I think about what role AI plays and we’re all wondering what’s AI going to mean to us in our our daily lives, but in our professional lives, AI’s going to make our job easier to make money, and that’s what gets us excited.

AI is already at play at restaurants, right? So you’re probably thinking, where do we go with AI? Should I think about it? Should I start with it? Companies are already using AI, right? So we developed an AI forecasting tool that is in place at these customers. And the AI forecasting tool has increased forecast accuracy by over 20%. And that’s just one example of one use case where AI can make a difference.

And it’s end is having a profitable impact. Right. So Noodles to save over $4 million in labor right. Chili’s is saving over 600 labor hours a week nationally. Why does that matter? Well, to a certain degree, that’s money. But maybe more importantly, it’s freeing up managers time to be out front, training the team, ensuring an amazing guest experience.

Pizza Hut right? Cranking out almost $250,000 a quarter. And that’s at just the stores if implemented thus far. By the time they roll out across their entire US landscape, then impact will be better. Thai Leisure Group, of customer of ours who you haven’t heard of because they’re operating out the UK. Interesting story here. Sure, they’ve they’ve seen a 1% decrease in labor costs, but what they’ve done is redeploy that labor to actually drive increased sales.

We always think about labor being reduction of cost, but if you get the labor in the right places and you’ve got kind of a constrained operating model in your restaurant, you can actually end up with increased sales. So we think that it’s happening now and we should all be asking the question, which is why these guys are here, but how do you get started and where do you see the use cases?

All right. Let me just I said I won’t pontificate for long, but I will pontificate just a couple of minutes.

Where we think AI is going is it starts with it starts with exposing your data. Right. And Jerry and I had a conversation about this yesterday. It’s so important to be able to expose your data to the AI engine. And that AI engine should feed a set of dashboard analytics in real time. That then gets to the to to the holy grail.

Okay, so those three things are kind of fundamental. If you don’t have clean access to your data, that’s the fuel that runs the engine. The dashboard and insights are the AI engine kind of working. So you’re not looking at a graph, right. Or a chart and saying, okay, what does this mean? What do I do next? Instead? It’s kind of doing that analysis for you.

It’s like a 24 over 365 analytics engine. And if it’s in real time, then it’s not looking to say what happened last week. What do I do next week? It’s like, what’s happening now? What do I do now? And that’s getting to the so what? The real power of AI in our industry, which is when the AI engine is consuming data, analyzing situations and it’s all through correlation analysis.

It’s all about making correlations between what’s happening in this store, what’s happening in that store, and is that driving a difference in performance and trying to tease out what makes one store more profitable than another, what makes one crew more profitable than another? What decision resulted in better performance and then surfacing that. And when you get above store insights in a real time environment, you can then drive in-store, in-shift actions. Some of you are like, nodding your head. Others just saying “What does he mean by that?” Let me give you an example. I’m going to get back to this thing. Imagine we have Sam. Samantha has been working as a manager in a new store in Oregon for about a year. Right. She’s been in the system for about four years as a manager, but she’s been in the store for a year, and it’s a busy a really busy day.

And it looks she knows her phone and says, look, it’s not gonna be raining in Oregon this weekend. It’s gonna be warm. She’s got a big patio business, right? And she’s thinking, oh, I bet we’re going to be busy this Saturday. But she’s in the middle of the lunch. Lunch rush and she’s thinking, oh, I’ll need to remember that right now.

Imagine instead of her trying to remember, oh, I need to pick up some staff. And oh, by the way, I operate in, in, in Oregon, so I can’t just tell someone to pick up a staff unless they’re on the volunteer standby list. I’ve got some problems. And these are the decisions that face managers day in, day in, day out.

I make the argument all the time. Even though I didn’t grow up in the restaurant industry. I’ve never worked in a restaurant, but I’ve come to understand the restaurant operations. I think being a store restaurant is arguably one of the toughest jobs in all of America. We ask our manager to do so many things and remember so many things, and no one’s great at everything, and we ask them to be good at almost everything.

But in this environment, imagine that your data is exposed, right? Operational data, weather data coming from third party sources, and the AI engine’s cranking and the AI engines are seeing something that popped up new. Hey, warm weather, sunny day in Oregon and they can look back historically and see what’s happened in past sunny days in Oregon and say, guess what?

In those in those situations, we typically see about a 25% increase in store traffic and 35% increase in ticket. And then without having to make Samantha think about this. Instead Sam gets Sam, gets a notification on her phone that says, hey forecast. Looking for warm, sunny weather this weekend? Please schedule two more staff members in front of house for Saturday at 3 p.m..

And by the way, I recommend Mike and Barry. They’re both on our volunteer standby list. You can ask them to pick up the shift without incurring any compliance issues with the Oregon Board of Labor. That’s one simple example. Imagine that applied across thousands of decisions that the manager makes each and every day and every week. And if we do this right, exposing the data to the right algorithms.

And by the way, the fuel’s the most important thing. The data is the fuel. Because remember these are learning systems. And as you feed more data through the systems, they get smarter, they get better. I was talking to Drew just this morning about it. You show me some data, which I’m really excited to talk about that highlighted that.

You can see the machine learning. It’s getting better each week. They’re they’re operating. And so as you expose the data to these these systems and they’re looking for the correlation analysis, the correlations get smarter and the recommendations get better. So. Think of AI as simply a tool that can be embedded in your systems. That simply allows you to make better decisions on a real time basis across the things that matter most.

Now if we were to take AI and bury Jerry, he’s a store manager. Bury him with a thousand things. He’s going to get overwhelmed. He’s not going to look at his phone. But if we can automate those things that are administrative in nature, so they happen automatically. The frees up time to Jerry to spend more time with guests, more time with the team, and simply prompt them to do the 2 or 3 next best things.

Right. We all remember. Well, some of you were quite young in here. Those of us who have gray hair or no hair, remember the days before smartphones, right? And before Outlook. And you have to remember all these things, right? Like, oh, I’ve got an appointment at 2:00, and you might even carry around like a little planner and. Right.

You had to know how to get from your house to the hotel or from the airport to the hotel looking at a map. And now we just completely trust our phones to do those things. I remember the day when I flew to Mexico for a meeting in downtown Monterrey, and I didn’t know what my flight was. I didn’t know where my hotel was, and I didn’t know where the meeting was, but I went got on the airplane because I knew it was all my fault.

Right? That concept would have been terrifying 25 years ago. So just like some of us are a little bit nervous about AI today, it’s going to become, it’s going to become so commonplace and such a useful tool. It’s going to become second nature to that next generation of managers and workers who are already in our stores today.

Let’s talk about people’s real experience with with AI. And the question I posed to both Drew and Jerry and we’ll start in that order is, first of all, like AI seems so daunting, like where do you start? And how do you guys think about the best use cases where this can be relevant to restaurant operations?

Drew Brodnax, Head of Operational Finance, Chili’s
Yeah. So, for us, so at Brinker, we believe storytelling is probably one of the most powerful things you can do with operators. So as a finance guy, you know, one of my challenges is how do I get our operators to trust, right. And so it’s not seen as some golden palace or some finance person, you know, trying to finance the right, trying to tell me what do so.

What you know, a big thing that we use at Brinker is storytelling. Well, how do you how do you use storytelling. So the first thing we did, you know, I, I like Clinton, I trust Clinton, but when he says, hey, we’ve got this, we’ve got this great tool is I’m going to give it to our aces. Right? So, you know, what we did is we we started with probably one of our highest performing areas, which is a cluster of about ten restaurants.

And we gave it to them. We said, kick the tires on it. talking really closely, you know, weekly with the director of operations and the GM and, and kind of first was like, hey, do we do we think we even have anything here before we get anyone else involved? And, kind of kicked it off in July of last year, which is actually really important timing because in August, at least for our business, we typically struggle towards the end of August and September because you have back to school.

So that’s that’s a really great time to see. Hey, do we think do we think we have something here? And you know, what we started hearing is, hey, this thing this thing is we can see that it’s detecting back to school. We can see that it’s it’s, it’s more accurate. 95% of of the GM said, hey, this is a go.

and this was our top performing area. So we did see we did see benefit with them. And then we said, okay, well, let’s go completely opposite direction now, let’s go to one of our lowest performing areas. That’s really when the film that right. You know, we didn’t we didn’t we didn’t tell them that. But that’s really where we saw the fireworks go off is, you know when you give it you know, we have coming out of Covid, our manager tenure is is lower than what it was.

So, you know, the average manager in our business, you know, isn’t as tenured. We have a lot of managers going through holidays for the first time going through back to school. And so when we gave it to some of the some of those lower tenure managers, to play around with it, we really saw the huge, huge benefit there.

And so then we said, hey, now let’s take this to the brand. And so we rolled it. We rolled it brand wide. But what we used is those to basically cluster of 20 restaurants, a couple of directors, we let them get on the phone with their peers and tell the story. The proof is in the pudding. Showed some results.

And then we got to a point where we actually started having, you know, other markets start asking for it. And so it was no longer seen as something that that corporate was giving me or from finance. It was seen as an operational advantage that my peers are using, saving the manager some time. And so we leaned in the storytelling, we kind of proved it out and then went to some areas, had some results.

And then we kind of we’ve been going for four months plus now and over 20% improvement. So it’s it’s really exciting stuff.

Clinton Anderson, CEO, Fourth
You know, I love how you referenced how it it makes the life of a great manager easier. Great managers are amazing, right? If you’ve been running a store for a long time, like they’re darn near AI, right? But it makes their lives easier. But where it gets super powerful, right is it takes that that underperforming manager, right. That new manager, that insecure manager and makes them very good.

Right. So takes a bad store. Bad manager makes them above average, takes an average store, makes them great right. And that’s and that’s where we see a lot of opportunity. Now Jerry, what was your experience. How did you think about it.

Jerry Phillips, VP of Technology, Whataburger
Very very similar. we did take a, we added to that a little bit. We wanted to make sure, our managers, of course, on a 24 by seven operation, they’re managing a team of 70 people. And so we were like, okay, we can’t just throw AI at them. First off, they don’t understand AI. They, they see the marketing.

They think that it’s, you know, what Microsoft says on TV, right? It’s going to make your emails easier. And these guys are like, I need help.

Clinton Anderson, CEO, Fourth
How does that relate to a restaurant?

Jerry Phillips, VP of Technology, Whataburger
Right, right. I was like, yeah, that’s not our job. Right. And so we said, the first place we’re really going to go focus is in that, that standardizing a sales forecast, which is the basis of your labor and your inventory. And we did a very similar approach. We took, the top markets and we took a few of our lower, performing restaurants.

And again, we didn’t tell them that you’re top and you’re not. But we said, here’s a tool. We walked them through the value that it showed. And then we just step back and we we let them actually start to see it. And then we’ve had some guidance that would come along with that, to sit on top of it and really give them the opportunity.

And it was very similar. What we saw was the good managers, the new managers that didn’t know any different. They actually accepted it a lot faster because they were like, oh, I don’t have to understand the 87 other things that I’ve got to look at right now to adjust. I don’t have to understand that weather forecast. I don’t have to understand, worry so much about it’s football season and there’s going to be a school bus of kids drop in here on a Friday night on the way to a football game.

because this tool knew that from looking back. And it’s not always saying, you know, the dates change, right? Friday is not always a Friday year after year, but it does take a good hard look at that data and understand this is football and this is weather. And how does that impact. And so what it did was it gave them a forecast that was anywhere from 10 to 12, 15% better.

And then it gave them the ability to nudge that where they needed it. And it goes back to what Clinton was saying, that it’s a great tool, to to at least get them 80% there and then they can do the last 10 to 20% on the tweaking.

Clinton Anderson, CEO, Fourth
Okay. So Jerry said something that’s a really important insight that I think is worth noting. And that is AI plus the human experience is what gives you the best outcome, right? Because we’ve experimented, you might imagine, across different ways of doing this. Right. Hey, let’s just go with the manager forecast. Let’s go with our customer’s previous forecast plus manager input.

Let’s go with just our Fourth AI forecasting tool by itself. And let’s go with the Fourth AI forecasting tool plus management input. And what we learned is AI plus the human touch, the human experience, is where you get the very best outcomes. Right. And so in a way, that’s one way to help people get less nervous about this.

This tool is not meant to replace you. It’s meant to make you more effective, more productive. Frees up your time from the administrative stuff I’m gonna call stuff that you hate and get on and spend more time on things that matter more to you. Okay. Awesome. okay. I promised, the folks who organized this event, they will not be overtly commercial.

So this is not a commercial point. But sometimes when I’ve been talking about AI in the industry, some of the restaurant companies like, yeah, we’re tackling that. We’re gonna do that ourselves. What are your thoughts on how companies should think about either doing it themselves, building their own AI tools versus partnership with other companies? And if and if you think that partnership is important, how did you guys think about, you know, where to look for that, that expertise?

Jerry Phillips, VP of Technology, Whataburger
Yeah, I think we’ve tried both. I’ll be honest. AI is the buzzword. I think everybody went out there and after it and they were like, let’s go, let’s go build our own, right? We can all tap into the same elements. We can can do this. What we learned really quickly is that’s extremely hard to do. Let’s not let’s not fool ourselves.

The marketing, there’s a lot of hype in that marketing. and it’s, it’s going to take you time, which means you’re always going to be behind. It’s going to be very expensive. and we run restaurants. We’re not technology shops. when you really think about it. Right? We support technology, we roll out technology, we provide technology that drives the business.

But, we quickly shifted and said, okay, what is this? Well, first off, AI is going to become embedded in every piece of software you have in your life starting now. Doesn’t matter if it’s on your iPhone, don’t matter if it’s on your Google phone, on your laptops, on your bank accounts. Everything is going to have AI on it.

And so we took a step back and said, where do we focus? Where’s the most significant part of our restaurant data? And where is that data sitting, and who has that? Who has that data? And that’s when we started talking to, Fourth and companies around what does that data look like and how can you take that? Again, it’s understanding your data flow.

It’s knowing where that data sits. It’s knowing who has access to that data. AI is, think of it as a, right now think of it as an eighth grader. It’s learning, it’s not going to be perfect. But the more data that you can get into it from real time and coming into the closest that you have to a centralized system, which in our case was our back office platform.

we decided that’s where we’re going to go focus and quickly moved out of the mindset of trying to build our own.

Clinton Anderson, CEO, Fourth
Yeah, it’s interesting. Right? Because for those of you who are less familiar with AI and I’m not a technologist, by the way. Right. So hopefully I can translate this because I’m not into the details of how exactly AI works. But the simplest way to describe it for for I think all of us, is the idea that there’s an engine.

That engine is a set of algorithms that’s based on data science, that are looking for correlations and then looking to put forward, you know, insights from seeing look alikes. I’ve seen this before. I think it’s this is what’s going to happen in the future. Now, the algorithms that get developed by technology companies, are very good generally, but they’re actually not the most important thing.

All right. The algorithms engine and there are learning system. Right. Which means they get better over time. Right. Which is mean the standard error on their recommendations based on what they thought was going to happen, looking at correlations versus what actually happened gets better over time. And what actually makes those algorithms really effective is the fuel. Where does that fuel come from?

Do you know?

Data. It’s the data is the data. So there’s two critical points for us to remember as we think about our own AI journey is how do you expose your data cleanly. Right. So that’s one thing you can start talking to your IT teams about is do we have clean data? And I know that’s something that that both these guys have talked about a lot in our meetings with them.

But the second part, too, is one of the reasons why I would make the argument. I recognize I have a vested interest. I’ll take it with a grain of salt. But the reason I think you should partner with a company who’s expert at this is that they are learning from a huge repository of data, right? So Chili’s has 1200 stores.

Drew Brodnax, Head of Operational Finance, Chili’s
Corporate 1600.

Clinton Anderson, CEO, Fourth
1600, total 1200 corporate stores. Whataburger is at 900.

Jerry Phillips, VP of Technology, Whataburger
900 corporate and almost 200 franchises.

Clinton Anderson, CEO, Fourth
Okay, 900 and another 200. So you guys are operating in excess of 1000 stores, right? The data that runs through our systems is about 110,000. Right. And so for example, one of the reasons why the forecasting tools work so well with our customers in the United States is because it’s been already operating for three years in Europe with our customers.