Oli Ostertag on PAR Technology, Coach AI, and Smarter Restaurant Operations

Russ Fordyce 🏆 (00:01.528) Welcome back to the winner's circle, everybody. I'm here today with Ollie from Par Technologies. They are at the forefront of making all of our lives a little bit better when we go to the quick serve restaurant. So Oli welcome to the award show and congratulations on the big innovation award. Oli Ostertag (00:17.186) Russ, pleasure. Great to great to be here and and definitely excited for for Parr to pick this one up. Russ Fordyce 🏆 (00:22.53) Yeah, now a lot of us behind the scenes, not behind the scenes guys in quick serve restaurants or in restaurants in general haven't heard of par. So give us a little bit of that elevator pitch on par and kind of what you guys are up to. Oli Ostertag (00:35.818) Absolutely. So PAR is an enterprise technology company. We focus on restaurants, we focus on C stores, and we've really built out a platform where we believe that we are offering products that cover the key pillars of operations. We have over 150,000 unique sites. You know, we are transact, you know, processing over 12 billion transactions. We have, you know, 400 million loyalty members actually using our products. So we we've definitely progressed as as time went on and and built on a strong foundation, starting with point of sale and services and then adding engagement products, loyalty products, you know, payments products, et cetera, on top of that. Now for par AI really offers a unique opportunity to increase our our output, right? And really bring new features to the market that will drive extra performance in the in the places that we that we operate. So again, restaurants and and C stores, but really par is more of a multi-unit operator. technology. Russ Fordyce 🏆 (01:36.035) Yeah. And it's so you you guys are behind the scenes, you're the tech kind of the tech back end looking at all these different systems, some sometime providing these systems. it it seems to me that, you know, if you if you were going into kind of a mom and pop, not even mom and pop, because there's hardly any of those, right? They're generally multi-service, you know, most of these guys have two, three, five locations, if I'm correct. You go into something like that, it's still run kind of locally. It's not a it's not a big brand you've ever heard of. It's not a company you've ever heard of, but it's a smaller operation. And I imagine that those systems are are uber disconnected. And now with AI and and you know, and with a kind of a more integrated stack, it's it like it's like turning on all the lights to the operation. Oli Ostertag (02:22.764) Hundred percent. Look, I I think the the big difference for enterprise brands or enterpr you know, call it enterprise grade brands, is the complexity in which they operate, right? They're they're using multiple different different products, and all of those products need to be properly integrated. You know, honestly, AI is only as good as the data that it actually plugs into. We call that context equity. Actually, you know, having companies be partnered over longer periods of times with customers, understanding exactly how their operations functions, that that matters, right? That that understanding of of workflows, that is the difference between you know having a a vendor like par that is partnering across multiple different layers versus you know piecing together multiple different products that don't speak together. In the best possible way. Enterprise grade brands, it it is a cutthroat environment out there, right? So there's there's a lot of margin pressures. You see, you know, value wars, for instance, on the restaurant side, value wars are really just price wars. And AI allows these operators to have an edge, right? The the edge exists only if you do have cohesive data functioning where you can make the best out of that. So again, the larger you get, you know, the the complexity. continues to rise. PARs there to meet that complexity. Our bread and butter customers tend to be, you know, call it 50 plus sites. and we're excited to partner with the likes of a Burger King, you know, the likes of a Papa John's, the likes of a of a Dairy Queen on the restaurant side, and then the likes of, for instance, an E.G. Parker's, et cetera, on the Seesaw side. Russ Fordyce 🏆 (03:58.959) So give us a sense of kind of the the the products or the features that are coming out of this. I imagine, you know, you're you're getting all sorts of data from the operation. You're probably getting data from marketing engines, from marketing signals, from, you know, traffic to websites, mobile clicks, all that type of stuff. Walk us through kind of a typical kind of installation at a Burger King. Oli Ostertag (04:20.27) Absolutely. So look, the the important thing to remember with with restaurants is there are effectively two sides to their technology. You have the operator products, and that will include, you know, the point of sale, the payments, the back office, and then you have the engagement products, which is, for instance, the loyalty products, online ordering products. We are looking to use AI and bring out, you know, product enhancements across both dimensions. On the operator side, we've we've brought to market a product called Coach AI. What Coach AI does, it really allows operators to understand their data better and be really prompted in real time about how things are operating. You know, what what we found the biggest problem to be for a lot of these, you know, restaurant change is that the the performance gap between the best store and the worst store is around 3.6 X, right? That is a huge disparity. Our product, Coach AI, allows operators to really understand, hey, Where am I being wasteful with with with labor, with inventory, and be prompted in real time to optimize? Now, with AI, you will always see a step change in performance as that gets improved over time. So, you know, the the the starting modus, if you will, with some of these products is an ask and you will can you shall receive, right? You query the data, you get put outs, right, instantly, you know, language that just makes sense to to whoever is actually operating the restaurant, they will understand, they will see, you know, easy, easily digestible data that that tells them exactly where they should be focused. The next step change will be where this actually becomes more of a self-driving, right? Where the the product itself will make decisions based on whatever parameters you you put in front of it. So that that's the really neat thing about AI and and where that's headed. It will allow you to do, you know, again starting where you're still holding the the steering wheel, right? To where you can actually sit back and and let your your store operate like the best store. That's on the operator side. Now on the engagement side, you can of course always add, you know, the the same kind of tooling where you can ask and you shall receive with your data where it tells you exactly what what are the the campaigns that are running best, why are they running best, like what are the the key things that you should be taking away for your next campaigns. Oli Ostertag (06:42.146) But the next step there is again automated offers, where it is just collecting all that data and telling you exactly this is the campaign you should run. And at some point, actually just running it from a self-driving mode yourself, right? And that's really the way that we've tackled this problem. We're starting two different approaches, one on the operator side, one on the engagement side. Both drive ROI. We're definitely seeing that with the customers that are using our products, they're getting immediate results, but we definitely expect that to to only continue in one direction going forward as as these products get smarter and are trained on more on more context going forward. Russ Fordyce 🏆 (07:22.242) Well, it's it's really interesting. You you're the context of the AI engine is really kind of that overlay. And you know, having, you know, developed and sold systems my entire career, that was always kind of that missing part. And I always encourage the sales people and the and the deployment people to get involved when you deployed something, you know, go back and and show the customer, you know, how you sold it to them. You sold it on a specific problem, right? Generally speaking, or a set of problems. Go back and and and make those connections, show them how to use the product, how they're going to implement it. And so we spent a lot of time kind of doing that, making sure we're closing that kind of side of the gap. With these AI engines, it's really for where we are today, it really does close that gap. It really, you know, and you're talking about tomorrow kind of, you know, do going going a little bit beyond that with the agents. But you're really closing that gap today of connecting all the dots. Oli Ostertag (08:21.882) A hundred percent. You know, I I think the important thing as well is, you know, for instance, the National Restaurant Association show was just in Chicago. And, you know, if you are walking the floor, everybody is talking about AI, AI, AI, right? There's a real sense of AI fatigue within our within our customer base, right? When people are hearing AI, they feel skeptical. You know, there's been a bunch of studies, you know, from You know, for instance, from MIT from Deloitte, that the results driven from AI are still intangible right now for a lot of different customers, hence the the skepticism. But it's really just about pushing outcomes. So what when we're trying to engage with our customers, we try to talk less about AI and more about enhanced outcomes. And what you mentioned, absolutely, with with AI sitting on connected data, the outcomes will rise, right? So again. I I think you will see a step change in adoption that that is driven by enhanced performance. That's the way we'll always go. Our our industry, you know, whether whether you're talking restaurants, C SOR is simply too competitive for people not to adopt the the best products. And what I would imagine happens in the coming, you know, months and years ahead is just people really pushing forward with having better data integrity and data cohesion, right? Because that that's really what makes these AI products Format. Russ Fordyce 🏆 (09:44.483) Yeah, it it is it is a huge it it's still crap in, crap out, you know. It's as we as we used to say. And it I I think I think the CTOs and the CIOs and the business owners I talk to kind of get that. I don't know that everybody else does, you know, and they they see the the impact of it of, you know, a wrong name on a mailing address or something, or spelled weird or in capitals and you know, that's data integrity and and people don't Oli Ostertag (09:50.51) Mm. Hundred percent. Russ Fordyce 🏆 (10:13.346) realize how what an impact that has once it's kind of compounds in your databases and then in your business. But Oli Ostertag (10:19.66) Yeah. You know, for for us, we really think about this as playing both offense and defense, right? Hence again, par's focus on having both the engagement products as well as the operator products. On offense, you know, everybody wants to drive more revenue. You know, for brands themselves, when they have franchisees, obviously brands earn based on how many sales dollars there are. They get a percentage of that. But excess sales without fixing the underlying operations, like your margins are just continuing to shrink, which is why even when people at banner sales years raising the prices doesn't necessarily mean that the underlying franchisee is doing any better. And that's where playing defense is important by having products like, you know, a coach AI be be able to save you that 1%, 2% extra, which makes all the difference. Right. We're not talking about restaurants here having, you 30, 40, 50% margins. It's much smaller than that. These, these Individual points matter enormously. So playing offense, playing defense, that's why companies like Chili's have done so well. you know, and and you you see it when Chili's themselves are talking about their achievements, they're talking about both sides, right? Re really, it's about expanding the operations frontier. and you will hear them talking actually more about the operational excellence they have rather than the viral marketing campaigns. But combining that is unbelievably important. Russ Fordyce 🏆 (11:39.545) Yeah, and if I I don't know if you notice you you're in the business, but I I've noticed these types of of campaigns through my kids or my children. they're they're not kids anymore, but you know, they'll talk about the deals they get. Hell, I got a great deal today at Duncan or I got date great deal today at McDonald's. It was, you know, four dollars for two burgers and a fry and a coke. And you're like, I haven't paid that since nineteen eighty seven. What are you talking about? And so it it really is these kind of QSR wars. these value wars as you talk about but they are price wars. In some regards they're marketing wars, but in some t sometimes they're also I've got excess inventory of burgers and fries and I need to get it out of my store. Oli Ostertag (12:19.694) Yeah, you know, in in terms of that and and optimizing for that, this is exactly one of those use cases where AI will come to shine. This is what we're trying to tackle with, for instance, a coach AI, where you are prompted, hey, your inventory is expiring. And then being able, when you have that interconnected data, to understand here's how I fastest get this piece of inventory out in the most profitable way. Here's how I need to staff in order to do that. And here's a projected, you know, EPITA profit. view of what that campaign could actually lead to. If you don't have that, right, and value wars continue, and I and I really believe they they will continue just given the state of of the economy, you you will be in trouble. And the the operators that figure this out are are the ones that will thrive, right? The you know success really isn't a surprise when you look at the the the names of companies that do really well. They do well for a reason, right? Yet yes, their products are are considered High value, right? So you're always trying to balance that view of the quality of the product and also how cheaply you can you can offer that. But that combination is really complicated. So again, for par, our goal as an organization, and and really this is the mantra that that we live by, is food, people, nothing in between, meaning we are trying to provide a technology that is not getting in the way of the experience with the brand. And you are optimizing then for how. much you can you can sell, right, at the at the best possible price where people feel like they got a great value. The tech should really sink into the Russ Fordyce 🏆 (13:54.115) Yeah. Now I I want to get a little bit nerdy. we you know, that we've talked a lot about the technology in this layer, but it's really not a layer. You guys made kind of a deliberate choice to bake this AI inside of your kind of operator engine rather than kind of sitting out on top, kind of l looking at everything and maybe be able to bring stuff in or whatever. Why would you think that was the right move and you know, what may have been lost if you'd kind of gone the other way? Oli Ostertag (14:22.252) Yeah, you know, the the saying we've, you know, been been pushing, you you'll hear this from a lot of others, this idea of built-in versus bolted on. The the reason for that is again context equity. We we want to make sure that the that the products themselves are are learning from the data as they go through, right? The the whole point is that products like again a coach AI become better with time because they are seeing more scenarios, they're seeing, you know, what what works, what doesn't work, and are adapting. If you are layering on a a a piece of technology that does not understand the workflows properly, right? Does not have that depth of context and is not continually learning, you're you're gonna, to your point, get garbage in, garbage out, right? So that has really been been our view of of what is required to to make these products better over time. And you know, again, the the Upside from this, we we don't know really where the the journey ends. All we know is that as we're building in this functionality and the functionality becomes more powerful with time, you will see better results. And that that's really why this idea of built in. Russ Fordyce 🏆 (15:26.978) Yeah. Now you've tested, we've talked about some of the brands, some of the big bigger name brands that you guys have, and you're talking about these kind of enterprise installations going, you know, fifty plus stores, and that's a pretty large enterprise. what did you learn from them that you didn't think you'd you would have learned during kind of the whole design review? Oli Ostertag (15:47.522) You know, the the best customers you have are always the customers that you can build and partner with. you know, one of the the things we learned very early on, par was one of the first companies to actually put out, you know, an AI product there from like a commercialization perspective in a coach AI, was that we needed to be very sensitive with how it was built, that you don't have the problem of hallucinations, right? Like w AI can make mistakes. It is unbelievably important that the products we put in front of our restaurants do not make mistakes. That would cost them a lot of money. So from a refinement process, like working through that and understanding how sensitive we need to be to double, triple query the data and to make sure that you don't run into those problems. That's an example. you know, ultimately we also want to understand where our brand partners are going. What do they care about, you know, in 26 and heading into 27 and understand. exactly where we should be focused, what we should be building towards. That is always a an iterative process. For par we have over 200 enterprise grade customers, call it where where you do have you know dozens of sites. we we have the the the luxury that we sit across multiple different formats of stores and that that's really where this iterating process comes from. We we partner with the the the growth aspiring brands as well as the very big ones. So I think we do have that holistic view of the market and can optimize as a as a result. So again, you know, with with AI, it is certainly early in the AI journey. Our job is to make sure that we are learning what works, what doesn't work, and then double down where we are seeing, you know, things actually come out the the the way that we would hope. Russ Fordyce 🏆 (17:31.673) Well, it it seems like it seems like a perfect layer. It's that, you know, it is that human layer that we're kind of working on, right? It's it's it's just a limitless supply of humans that you're applying to something. And it it's so interesting to watch that kind of come to fruition because even, you know, in my own work, you know, all of a sudden I'm a designer, all of a sudden I'm a coder, all of a sudden, you know, you have all these new capabilities. And as you're talking about, if you have that data in those connected platforms, now you just really need that human effort. And kind of back to what we were talking about, if if you, you know, you were sold that that outcome. And you know, before it was up to you and your staff or consultants you hired to come get you that outcome. And now it's with AI, it's a lot easier. And we talked about agents and hinted at it a little bit. you know, you you're kind of moving kind of from a directional of a manager kind of capability or an analyst capability into really an operator kind of capability. Where does Coach AI kind of go next into that agent realm and how are you thinking about that line responsibly? Oli Ostertag (18:39.65) Yeah, for for sure. Look, again, what what we're trying to really achieve, and I mentioned a little bit earlier, it's just there is such a wide disparity between the best operating stores and the worst. So, you know, having an operator, you know, be be able to monitor agents, if you will, versus having to themselves analyze what is working, what is not. You are you are increasing the quality of each individual person. That's really the point of of AI. So where does coach, where does pargo go next? You know, we we see this as PAR looking to to bring to market, you know, new agent functionality on an ongoing basis. So you know we we talked about Coach. I coach AI is is ultimately really a a data analysis tool, right? And a a a scheduling tool, things of that nature. We talked on the engagement side with with the ability to drive offers. But the next one will be things like pricing agents, fraud agents, things, things of that nature. There is a whole bevy of different things that that restaurant operators and and restaurant, you know, cor corporate entities that that are looking to maximize profitability are looking towards. We can provide agents that are covering all of these different, you know, pieces of of the pie, if you will. but but again, where I want to be very careful is I I don't See AI as being something that is just replacing humans. Really, what we are thinking is AI make people more efficient, right? Like what we are trying to drive towards is again an expansion of the operations frontier where stores are more profitable and making more, and humans are able to do more. If you're able to do that and there is more profitability, that's really where you see this work. You know, in in in a way where you're not replacing people, you are just making them more performant. They're driving more individual value. And and maybe one piece that that I again think is really important. You know, I I don't believe in a future like where where the interpersonal connection goes away for restaurants. When you talk about brand value and which brands do really well, it is the emotive element to the brand. It means something. You you go into certain stores, you get a feeling, when you engage with Oli Ostertag (20:58.306) With even a a drive-thru, you you you have a certain feeling based on how you interact with a brand. That is a human portion to to many extent. So AI can enhance that, but any brand that is looking to only go all in on AI and not have the interpersonal element anymore, then then I do believe hospitality goes away. Hospitality is is innately a a human thing. And yeah, that's the way we're we're approaching it, right? I I think in in in in every possible way you will have humans and AI coexist in in a in a productive, profitable way if applied in the in in the right in the right way. Russ Fordyce 🏆 (21:36.248) Yeah, you you just described it and in my in my brain I saw, you know, I saw a giant vending machine, right? It's it be it becomes very impersonal. It becomes, you know, a very generic experience, which is which is what not what brands are investing in. Oli Ostertag (21:48.962) Totally. I mean if if you look at certain, you know, if you if you look at certain innovations like kiosks, kiosks absolutely have their place. Kiosks work, right? Like it's proven you you're getting upsell enhancements via via kiosk, and and for a lot of people, there's a there's an enhanced sense of speed, right? You can you can cover more people. If you look at any store that has both In-person capabilities, you can go to the point of sale, talk to someone, and order, and kiosks, you're never seeing people only go to the kiosks. That there's a place for that, but it does not replace the interpersonal element there. There's a reason why, you know, for instance, Burger King's recent campaign where they say you rule and provide you with a crown, why that works, why people care? Because that hospitality. element is important, right? The emotive element. So there's always a a place for technological advancement. Absolutely. You will drive outsize performance, but you should never you should never put the brand on the on the back burner. And that is also why, you know, for par, we care about both sides. We care about the operations, but we also care deeply about the the marketing piece, the engagement piece. And again, if you do it the right way, you can enhance everything with with AI. Russ Fordyce 🏆 (23:12.632) Yeah, I I believe in that engagement piece too. I d I I'm not a big app guy, but boy, it does work on the younger generations. It's it's how they live their food life these days, it seems like. Now you spent a lot of time outside of the US. Par has a good international footprint, you've got a lot of international customers, operators. I I'm wondering how do they see I AI differently than kind of the US operators? Oli Ostertag (23:38.646) It it's a great question. generally speaking, I would say that the US market tends to be among the first movers. I I would say from a QSR perspective, especially, the most sophisticated markets I've seen tend to be the US and then, you know, pockets in Europe, and then separately Australia. so again, we we have an office out in Australia, we partner with a lot of brands out there. you'll you'll see a lot of similarities to AI uptake versus versus the US. Other markets maybe a little bit, you know, call it a little bit behind the curve or maybe waiting to see how things unfold. You will also see that, you know, restaurant technology, you have providers all over all over the world, but you do tend to see a big focus on the US market that innovations, the the big innovations, tend to land in the states first. It is also obviously the largest you know, call it enterprise or proven enterprise market in the world. And you know, PAR doesn't do as much in China. I know that China has a lot of innovations as well. Maybe we put that to to the side, but the US remains kind of the innovation hub where people are are focused on bringing their products to market and proving that it works. Adoption cycles, things that are working in the US, tend to work abroad as well. You can look at things like drive-throughs, right? It takes maybe took a bit of time to bring it to different markets, but it did take off in in different markets. But then other products, maybe like a kiosk, took off in some select European markets first. So tends to be the US market that is the most sophisticated and then a follow the leader approach, but kind of context specific as well. Russ Fordyce 🏆 (25:19.684) So I in in my experience, a lot of the times you were talking to usually we were talking to senior leadership inside of a large kind of tech sale, right? You you you really were working at the upper echelons of the business. And I imagine your your salespeople are too, and your your you know, your coaches and consultants. A lot of those people have been burned especially bad by tech kind of rollouts before. And and I imagine that this AI kind of flurry that or hype cycle that we're in is really kind of scaring some of them. Obviously they see there's opportunity there. What does that first week with Coach AI have to look like so that they know kind of almost immediately that adoption's going to be very high on this kind of product? Oli Ostertag (26:06.508) Yeah, Russ, I think the most important thing is training and and partnering there. You know, for for Parr, we invest heavily into the services front. We have, you know, experts that that actually sit with our operators and make sure they understand exactly what this product looked like and that we offer recurring training. You know, obviously a lot of restaurant operators have the problem of of staffing. not being super consistent over time. People tend to to churn out pretty, pretty quickly out of some of these jobs. And that recurring basis is is incredibly important. Separately, we of course want to make sure that the product itself is easy to use. So with Coach AI, it really is natural language queries, right? And and then it actually provides digestible responses where you don't need to, you know, under understand business in order to understand how to manage your your profit and loss. I I think that's incredibly incredibly important for for us to continue partnering there and then learning what what is working and what is not working. Russ Fordyce 🏆 (27:04.024) Yeah. Now you're actually the builder or one of the builders of this product, kind of the I guess the design lead on this. If there's one principle that you kind of learned when you're building this coach AI that another product leader or AI leader in the in the different space is looking at, what would you tell them to kind of watch out for or lean into? Oli Ostertag (27:23.52) I I would say a as a really important principle, restaurants, enterprise restaurant brands are are not the place to try out the, you know, move fast and break things, right? That can have an unbelievably negative impact, right? Like we're again, i if you're talking about the amount of sales that are generated, but also just the the the margin pressures there, if you get something wrong, the the Impact to the operator is is brutal. So as a first step, make sure what you are providing is actually performant, right? Make make sure that it is truly tested and make sure that you are trialing this in scenarios where you have that time and space to tweak and to refine. That's that's a really important piece because AI, while unbelievably valuable and and Absolutely with with the ability to enhance your your products, it's not perfect. And you need to understand the the gaps. And as mentioned on our product, for instance, understanding how do you cut out the hallucinations that could result if you don't understand how this exactly works. The second thing is again usability. You want to make sure that these products are actually used on an ongoing basis. People see value on a daily basis. you know. What what I think a lot of technology, you know, certainly restaurant technology companies forget is that usage rates really, really do matter. You can sign a a corporate contract, you can get a mandate. If operators are not using this product, then there's no value, that product will be replaced over time. So again, usability, unbelievably important, and making sure that it it works, not just for one facet of that you know. that ecosystem. It it shouldn't just work for corporate, it should work for the the franchisees and the in-store operators, right? And making sure that that that is super clear, that you are addressing all three of your key constituents. Russ Fordyce 🏆 (29:18.928) Ollie, thank you so much for joining us. It's a super fascinating conversation. And obviously all the all the success is warranted. This is some pretty interesting stuff. you know, I think all of us benefit when all of these operators are leaner, meaner, and you know, and operating better. You know, we get we get better quality food and at a cheaper price generally. So that's good for us. So to thank you so much for joining us. Oli Ostertag (29:40.308) Absolutely. Russ, thanks so much for the the opportunity.