We're going to start with the next session. Next up, pleased to have Savneet Singh, President and CEO of PAR Technology Corporation. Savneet, thanks for joining us again. It's a pleasure to have you yet again at the conference.
Thanks for having me.
Before we get into details, I think let's level set for the room. Anyone who may be newer to the name, PAR has been on a 40-year-old journey to kind of reinvent themselves from a hardware company to a software company. Can you give us a snapshot of who PAR is today and what the platform does across restaurant, front of house, and back of house?
Sure. PAR is a platform to run enterprise restaurants and convenience stores. We are everything from loyalty and online ordering, all the way to point of sale and back office. Our platform really sort of starts with plant the flag with point of sale, and then we looked up, saw you, back of house, online ordering, and loyalty. We've been sort of deep in software for six, seven years now. Expanded rapidly. Today, we cover over 100,000 restaurants and 25,000 convenience stores.
Got it. On second quarter numbers, you reported last month you raised full year on both the top and bottom line. You haven't been giving guidance for a while, so I think that's notable. What stood out to you in the quarter? Anything you'd like the script to focus on?
We started guidance this year. This is our second time giving guidance. I think probably the biggest takeaway is we continue to significantly outperform our EBITDA targets. I hope that continues for some time. I think that there's a really strong operating leverage story here, alongside really durable growth. I think probably the biggest takeaway was that we are overearning. I don't think it's overearning. I think we're earning into a continual growth. I think every dollar of future revenue will come at a substantially higher margin than where we started. That is really a direct result of a platform strategy, whereas when we're now selling two, three, four products per customer, the incremental margin per customer is much, much higher than it was in the past.
I think there's a little bit of underappreciation of how much that is flowing through the bottom line. I think the second thing that's interesting to take away is we crossed over 20,000 sites that have our PAR Intelligence platform on it now. I think that's really exciting for us because we're selling to enterprises. It's not like a tool we can just flip on and say, surprise, you have it. We have to get approvals. We got to get it signed up. I think it gets us really excited about the potential future monetization of what we're doing there.
I want to come back to that in a second, but maybe talk for a second about the decision to give guidance. What was the thought process? What changed in terms of visibility? Secondly, how are you thinking about the philosophy around guidance go forward, about how you set the bar and evaluate over the course of the year?
I think guidance for us was probably twofold. One, when we took over the company, we were less than $10 million of ARR, and we used to get asked to give guidance, and I am like, how many Series A companies give guidance? We just happened to be public. I think it was always hard. I think when we were building up our platform, part through acquisition, part organically, it was very hard to give guidance that I think was instructive. I think we got to a point now where the business is mature enough where we felt comfortable giving guidance. I think the second part was we had a lot of visibility coming into this year, so we didn't want to come out with guidance and be wrong. We felt like we had a lot of visibility this year and thought it was a good time to kick things off.
Got it. That makes sense. Just philosophy going forward on how you set that bar.
I think we try to put a bar we can hit and hopefully exceed. The first two quarters, we beat pretty substantially. I think we'll eventually narrow it in . I don't think we want to beat by that much all the time, but I think our goal is to put numbers we can achieve and give reasonable expectations.
Got it. Let's talk a little about AI and how that's impacting the business. Can you talk through how AI is actually changing how PAR operates internally, and how you're kind of separating structural cost takeouts from what I think you called AI washing cost-saving programs?
I think we went about AI really department by department, job by job. I think it took a little bit longer, but I think it really worked in our favor. If you exclude the set of assets that we acquired in Q1 of this year, our organic OpEx is down something like $15 million or $20 million. A substantial portion of that came from cost takeouts leveraging AI. I think an example is probably helpful. Our finance department today is down 25%, maybe 30% from where it started this year. Our goal is to take it down to 40% by Q1 next year. We've been able to take out a huge portion of our finance costs because we've been able to leverage AI in areas that we were manually doing stuff.
A couple of good examples are things like internal audit, where an agent can pull things that look fishy or needs to reconcile stuff. Areas of billing where we were doing a ton of stuff manually, collections. Really aggressively put AI to work, and it's dramatically helped us cut costs. Same thing within HR. All the queries that you send to HR, we've now made those agentic. Within our operations, our hardware team now has something called Hey PAR, where literally any question they have about a product, whether it's from 1985 or 1995 or now, they're able to query it. That allows us to cut the team down, but also get rid of all this old software we had. We got rid of Smartsheet with some of our demand planning because we were able to use a tool we built on Cloud.
I think we've been constantly going function by function to see what we can potentially automate away.
Got it. Makes sense. Just on the product side, you mentioned the 20,000 live sites on PAR Intelligence. Maybe talk about that product, what it actually is, what are customers doing with it today, and how you think about roadmap for that product.
So it is really exciting for us. PAR Intelligence is a single dashboard for all of your PAR products and hopefully your third-party products as well. But what I think is unique about it is, today what we are doing is what I expect every company in the world will do in software, which is it is like ChatGPT for your restaurant. You can say, what are today's sales like? How did they compare last week, last year? Did that promotion work? What store is working? How is my inventory looking? Your ability to just run reports. That is actually really valuable today, and we are not charging for that. The idea is, we call it Assist.
We had a really good example where one of our customers shows a screenshot where his CEO had asked, can you tell me the percentage of transactions between 0 and 10, and 20 and 30? Normally he is like, listen, I would have to download some stuff, run some stuff, because he wanted to see how that trended over the last period of time. It is actually kind of complicated, b ecause you are like, okay, how did that bucket of 0 to 10 compare to last year or whatever. He literally just copied it, put it into our PAR Assist and sent it to his CEO within 10 minutes or whatever.
That we call Assist. The next step in our evolution is Advise, which is instead of you prompting it starts prompting you. It says, hey, Will, did you know that drive-thru sales are down, or sorry, the drive-thru's backed up 30 seconds? FYI. It's sort of giving you, hey, there's too much inventory. Hey, this labor schedule doesn't seem to be working for you. It's actually predicting stuff for you to go take actions. So more valuable new prompting there. Our next third iteration of the product, which is coming out early next year, is the ability to take actions. In that example about the drive-thru, hey, the drive-thru's backed up 30 seconds. Do you want to shut off DoorDash so that the store can funnel the drive-thru faster? You press the button and it takes the action.
The idea is that over time we just automate that. So, hey, anytime the drive-thru's backed up by 30 seconds, just shut off DoorDash so we can add staff there or whatever it may be. When we get to that last point where we're going to take actions to help you optimize your store, where you literally don't have to do anything but press a button, that's when we intend to start monetizing it. Our goal is to get 50,000 stores on Assist by the end of this year, where they have this beautiful experience to kind of pull the data, pull the information, get a bunch of PAR customers on sort of Advise, then we'll look to start monetizing it.
Talk about where you are currently and just when you get to that Advise state. How long does that take and what should we be thinking about for a timeline to monetization?
Our goal is by the end of this year to have 50,000 stores on Assist. This is our first version where they're using it pretty extensively. Through 2027, our goal is then to get to this last point of action. Hopefully we'll have real revenues in 2027 that we can tie back to this product. We think over time, this product becomes how you interact with us for everything. There's no more point of sale login here, back out. It's all just in this one place, and that's really the major vision that we're pushing towards.
We've seen similar products across our coverage, most of them in the SMB space. What's different about rolling out AI-enabled products into the enterprise ecosystem? You've got corporate approvals, cyber reviews, longer cycles. I would imagine, though, that the focus on finding efficiencies and running businesses more efficiently, that there's more of a focus on that in these, some of these bigger enterprise organizations.
It's wildly different. I think, when I look at the SMB, a lot of the SMB AI stuff is like, hey, I'm going to grow your sales 10% - 40%. Imagine going to McDonald's, say, I'm going to use AI to increase your sales 10%- 40%. You just sound silly. No one's going to go get a multi-thousand store restaurant chain and grow their sales by 40%.
That just gives you perspective, like how different the SMB is from the enterprise. Because SMB will say, I can grow my sales 25% by using AI. Whereas that is just not reasonable to assume in any large business. The other part of it is kind of what you talked about, which is, okay, so what can you figure out? In the enterprise where there is a massive amount of focus is on the back of house. How do I optimize my inventory, my labor, my schedule, food costs, all that? Then it is on the loyalty side, which is how do I segment it? So, how do I actually say, Will is a father of two, that every week this week comes home super exhausted and needs a Big Mac or whatever? How do I segment down to you as an individual?
Those are the two areas you see in the enterprise, whereas the SMB is all about, at the moment, customer acquisition and juicing your business pretty quickly. Structurally, it is also very different where when we go sell something to the enterprise, we have got to actually commit to CIO to use the product. We have got to get it tested with a bunch of the brands and the stores, all the cyber stuff. It is a pretty robust process. But once you go through that, it is a huge moat between you and the next person because how many times do you want to go through that again and again and again? So I think us being first is really important.
How does the AI strategy extend to the C-Store side?
The C-Store side, we are actually a little bit further ahead. We are in 15,000 C-Stores already, leveraging Parsys. So we are seeing tremendous use cases. As an example, one of our customers, our first customer, actually, the beta customer, was actually able to use Parsys to realize they had, I think, a $1-million ad buy being funded by one of their suppliers that they did not use. So when you think about a $1- million ad buy, what is your cost acquisition? We paid for the product for years that way. So they are using it to predominantly engage on the marketing side. They are trying to figure out how do I segment, how do I target? So we have got a lot of traction there. We are going to test out a ton of other stuff.
As an example, we are testing with a customer, how do they use voice to interact with Parsys? Instead of you typing it, can it be in the headset and you can say, hey, how are we looking in the kitchen or whatever, of a C-Store. The C-Store, we have had a lot more engagement and a lot faster.
Why is that? Why are they so much more willing to engage?
I think it is our standing in the category. We are the very dominant loyalty provider. We have created a ton of value back to our customers. When we bring them something, they do not really have a strong competitive alternative, and then we kind of have this relationship. Whereas a restaurant, we have done a great job, but there is competitive alternatives. They are being called on all the time. In C-Store, you do not really have a lot of nearly the innovation and/or VC dollars that have flown into it like restaurant has.
Got it. Let us pivot over to Bridg. You closed earlier on this acquisition. Maybe you could provide an overview of the deal and the acquisition and how it fits into the better together thesis.
We closed on Bridg in Q1 of this year. Bridg is an IDR solution. Essentially, it is a tool that allows you identify who a guest is, whether they are in your loyalty program or not. It sort of takes matching data and figures out, okay, that is Will, he lives in this ZIP code, he has this demographic data. Then obviously, over time, the goal is to sort of target those customers. It is an incredibly powerful tool that historically was sold to the largest Walmart down. It is the biggest of the biggest. Bridg was originally acquired for $350 million in 2021 by a company called Cardlytics. Post earn-out, Cardlytics paid almost $500 million for the business, then we bought it for like $20 million in Q1.
We thought we got a steal. The idea was, can we take this Bridg product and combine it with our loyalty solutions so that we go to our customers and say, hey, not only do we have all your loyalty customers and every transaction that are done and how they think about you, we have your non-loyal guests alongside that, so we can then find ways to bring those guests to be loyal or vice versa, but also figure out how much of spend is happening outside. So if Will is a loyal customer of XYZ restaurant, how much is he spending on ABC restaurant and everywhere else? It is a really fascinating tool for our customers to demo because they are like, oh, crap. I do not have any of this information. We have a ton of interest.
We signed our largest restaurant loyalty customer onto it just a couple months after we bought it. We have a pretty strong pipeline of new deals coming through. So Bridg is pretty exciting, and I think will be the crux of our AI monetization going forward in that we can create really unique outcomes with stuff they couldn't do before.
What is the sales process or upsell motion look like on that?
It's very much pushed in conjunction with those that have our loyalty product. When you are selling loyalty, it is so much about data. Now, we have an incredibly powerful data product they did not have before. It is also very unique and hard to get to. We sort of sell it within that core buyer persona.
Have you talked about pricing relative to the base, what used to be Punchh?
We have not yet. Pricing does vary pretty significantly, and we are going to narrow in on this. You should expect a doubling of ARR on a loyalty customer when we turn this on. It does vary depending on what you're buying.
Let's talk Papa Johns. This was a major win for this year. Can you talk just a little bit about how that came together? How competitive was the process? What role did the Burger King win as a reference customer play in achieving that outcome?
Papa Johns is a big deal for us for a couple reasons. One, it's our first pizza chain, so we've never done pizza before. Pizza is a little different than our traditional QSR business in that you've got modifiers, you've got this pizza going through the oven stuff that Domino's pioneered. It's heavily delivery, heavily third party, heavily promotional based. So there are core product differentiators versus quick service and fast casual. So it was a big deal for us to move into a new vertical, kind of like an adjacent vertical, then do it in super fast time.
I think the core reason we won the deal is that Papa Johns, from a tech perspective, is run by a really innovative team that came out of Domino's, was the team that sort of did a lot of the really cool stuff that Domino's is known for. So they want to be viewed as tech forward. They want to be viewed as cutting edge. So they wanted a partner and a vendor that matched their values and their goals. So I think they took a huge leap of faith on us because we've never done what they're looking for them to do. So they're literally betting on us to do something we've never done before. But I think they felt it from our culture, they felt it from our roadmap, they saw the results.
I do think a huge part of it was the success we had at Burger King gave them massive confidence we could scale and handle the volumes and quantity of stores that they have today. I think that was a huge part of it.
So then I guess talk about pizza as a new vertical. You talked about how it's first major win in that category. Any changes to the product that will come from being in that vertical?
Absolutely. We've had to build a lot of that pizza functionality for Papa Johns. But what we've been lucky about is then we were able to go monetize that into those other customers. So in Q2, we announced we won a business called Pizza Factory. We won a third chain, a small chain after that. We've been making great progress. Pizza is kind of interesting in that there isn't an enterprise vendor that has serviced it before. Most of these pizza chains have actually custom-built software from a decade or sometimes decades ago that they've been trying to figure out how to work.
There's an incredible amount of excitement in the category because no one's ever gone after that category before. We think it could be a nice place for us to expand. We've got three customers on a standing start. A lot of that functionality, we can now use over and over again. I think that a lot of it too is just stuff where I suspect pizza's always been an innovator and far as, first to delivery, first to do DoorDash, first to Uber, there'll probably be stuff that comes out of that are relevant to the rest of our base over time.
During the Burger King process, there was a period of upsell where you end up winning additional products. Do you see similar opportunities here, and what can make sense over time?
Well, yes, and we already did. We sold them on two products at the time we won the deal. Again, I think that was because they saw what happened over there and said, okay, why don't we do that now? That was kind of exciting because we're able to launch with two products, not one, and not have to pause and restart. I think that they saw the value add there. It also sort of becoming the norm. In Q2, every single deal we did was multi-product.
Literally every deal. The prior four quarters or three or four quarters, it's sort of been 70%-80% of our deals were multi-product. I think we've been able to convince more and more customers to go that way. I absolutely think it's tied to the success of us working from these bigger, more complex brands.
Just remind us where you are in the implementation of that. I think you sized it in the past as kind of like mid-teens, millions of ARR once it's fully rolled out.
Papa Johns?
Yeah.
We'll hopefully kick off the very end of this year, but it's really majorly a 2027 event for us.
Got it. Do you think you can get it done in 2027?
That is the hope.
That is the hope. Got it. Similar question on Burger King. I know that one is farther along. Just where are you in the Burger King rollout? You have been running north of 400 stores a month. How should we think about the remaining contribution into next year?
I think we would be pretty much done by the end of this year. I don't think there'll be a small contribution for next year. Maybe some laggards, but for the most part, it'll be done. So 2027 will be a lot more about upselling the back office product and getting that going in the rest of the chain. So it's been super focused, super successful, ahead of schedule. We suspect no one's ever launched this many sites in a single year. So, it's gone really well.
That's great. I know there's been a lot of conversation around just the market opportunity of other brands, like the pizza industry that ran a lot of in-house software and are looking to modernize. We've been talking a lot about potential tier 1 RFPs out there. What's the level of confidence in PAR's positioning in some of those processes and any updates you can share?
We continue to make progress. We're careful not to give updates on stuff that hasn't announced or signed yet, given our customers are, it's kind of their private info. But, the volume of RFPs has increased pretty significantly in the last 12 months in our business. I think we are clearly in a replacement cycle of point of sale at least. So I think, given our win rates, we should continue to see hopefully above average growth, given that just the quantity of RFPs has increased. I believe the only department at our company that has grown headcount has been sales engineers.
Which are folks that do demos and architect around that. We're just seeing a lot of RFP volume there too. So I think we continue to make inroads into these big deals, and, as I said, our guidance doesn't assume any of it. So if it happens, it's just great upside for us to have.
What is the process of pitching and winning such a large customer like that look like relative to your standard sales process?
Man, the big ones are just very idiosyncratic. It is just impossible to sort of say if the one is like this or not. Some firms run a super strict RFP process where they bring in Accenture, Deloitte, or something like that and say, blind RFP response. Do not talk to us. Just work through the consultant. Other folks will sort of date you for a year or two and then sort of say, okay, let's do a quick abbreviated thing. Honestly with Burger King and Papa Johns, we were the last to the party, but we had won those very quickly because I think they had kind of gone deep on with other vendors and just were like, this is not going to work. Then we came in and could solve all their problems very quickly.
Unfortunately, it is not there, but again, it is a very small group of people that are in that, like very super large brands. Whereas our traditional process is, you obviously have account coverage across every restaurant chain in America, but by the time you get that RFP, it is like start to finish relatively smooth.
Just zooming out, I think there has been a lot of focus on, particularly early in the year, that AI may make companies manage more of their own software in-house. It is interesting you are saying that you are seeing a lot more RFPs coming to market, people looking to move to out-of-house vendors. What is the market like right now for some of these big opportunities? Are you seeing an acceleration in people looking to do kind of refreshes of their tech stack? How do you think AI plays into those decisions?
Well, I will say like, 99% of restaurant chains do not have their own tech stack now. It is not so much a move to external, it has always been external. There are a couple, McDonald's, Subway, Yum! Brands. There is a handful that have done stuff themselves, but everyone else has bought. I just think it is a continuation of that bought. I think even more so, what we are seeing is some of these firms that have built their own now look at AI as a way for them to get off their own. I think a lot of them are like, man, we are going to be stuck on our own clunky homegrown thing with AI. It is like, okay, now we can actually get off the thing we built that is flexible as resources.
I am not seeing any customer come to us and say, I want to go build a point-of-sale system. I do not think that is going to happen. If you think about it, almost all large restaurant brands are franchised. If I am that franchisor, I am not sure I want to build the payment engine, the transaction engine, the point-of-sale system to my franchisee. It is just going to be a massive point of friction for saving how much? Point of sale, you are paying $200 a month. It is like, is the ROI really worth the headache of giving your franchisees another thing to complain to you about.
I am not sure there is an ROI there given that. I have always said this, but to me, it is like the equivalent of us saying, I want to go build my own CRM system because I do not want to use whatever Gong or Salesforce or whatever we have developed. Why would I ever want to distract our team to build the best CRM system? You would only do that if you felt there was a massive competitive advantage. I do not think that exists in, or it is rarely, if you can build the best product for your store today, I doubt that lasts because these technology changes, integrations change. All of a sudden AI comes, and you are like, I got to rebuild it. I just don't think you want to be in that business.
The main driver of the movement of this aggressive sort of growth of RFPs, I think is really just down to the same exact stuff, which is, holy crap, our customers are more digital. We got to meet them where they are. We got to manage our operations more dynamically, because all of a sudden, I got to deliver, I got to have Uber Eats, got to have pickup, I got to be available on all these channels all these times. I got to manage labor. And I'm doing it from the exact same four walls that I did 20 years ago. And the only way to solve that is with technology.
Let's maybe pivot and talk about the competitive landscape. You sit behind or I guess between a lot of legacy enterprise vendors, Oracle, Micro, NCR Voyix, Global Payments. There's also a lot of SMB players who are trying to move up market. What's kind of stayed the same? What's changed over time on the competitive landscape?
The biggest players in the enterprise are NCR Voyix, Oracle, and Global Payments. That hasn't changed. I think they still have the most market share. Most of the customers we get are from those three or some other legacy provider. And I think that will continue for a long time. I don't think this is a priority business for any of them, and for the one that is a priority business, I think they just have too heavy of a product to fix to really be competitive. I just think it's hard. Even if they wanted to change, you're talking about products that are 15, 20 years old, millions of lines of code. That's really hard to do it. And even if you want to do it's like, the innovator's dilemma. How do you tell your customers that, tell your employees that? I think it's tough.
From the competitors from outside the enterprise, there's always been SMB people that have wanted to come up market. It's never really worked successfully. I think it's really hard to take an SMB product and make it enterprise. I don't know if there's a good example. I don't know if I've ever seen that. There is Intuit and there's SAP. There is an SMB product and usually an enterprise product.
I think it's really hard to make that the same product. You've got to almost completely change your stripes to go to that market. Specifically, we have incredible respect for Toast. I love those guys, I love those products, but I just don't think there's enough TAM, an opportunity to make it worth their time. At some point, I suspect economic rationalization comes in, and it's like, what are we doing here? I think there's a few startups here and there, but honestly, not a ton has changed since we probably last spoke last year.
Does having more people at least looking at the space, does that change any of the dynamics in competitive situations?
Not really. It's funny. It's been almost exactly four years since GPT 2.0 came out. We have not seen one new company in an RFP since that time. So whether it's loyalty, online ordering, point of sale, literally not one. We have not seen tremendous change on pricing or sales cycles. Only thing that's changed is the quantity of RFPs has increased.
Got it. One thing I wanted to talk about, payments was something the company has been investing in, I think more so for the table service part of the business, but probably applicable across the business. Any updates on where that product stands and how you're thinking about roadmap?
Payments is still less than 10% of our net revenues, but it's growing. When we sell to the enterprise, it's tough to always have payments because they've got their own deals with Fiserv or whoever. But where I'm super excited about payments actually is leveraging payments in a CDP system like we've talked about. How do I take that, get you on our payment rails, then I can take that data, put it into our CDP, and give you incredibly high fidelity information about Little Man so that I can target the hell out of you and make you a loyal customer. A lot of times I think about how do we not sell it as payments, but sell it as a robust CDP system. That's where I think we'll do more and more with these larger brands.
They look at it as, I'm not getting a payments product, I'm getting something else out of it. The place where I think we see pure payments growth is actually in online ordering. One of the exciting parts of our business is we've kind of built an online ordering product that's going after our base customers. This year, 2026, it's got a 50% win rate. That's nuts for us. Every quarter, we're winning five to six logos. About half of those come from the big legacy provider, and every single one of those deals has payments associated with it. That will be a nice driver because card-not-present transactions are very lucrative.
I think because all those deals are like, they go hand in hand, the ARPU is very high. I think we'll continue to have good, stable growth in payments, but I think the online side is actually where we're going to have the opportunity to do more than in the in-store. Can we build out this CDP idea? We will see where that is in a year from now.
Got it. Makes sense. Maybe a couple more financial-oriented questions. You have roughly doubled EBITDA year-over-year. As we talked about, you raised the guide. How are you thinking about margin expansion and the trajectory towards a kind of Rule of 40 model?
I think we have said our goal was to go to double EBITDA and double it again. Last year we did $24 million, $25 million. This year, I guess now we are guiding above that. We will hopefully do that again this year. Then, we think there is a potential path to do it again next year. A lot of that is coming from the stuff we talked about, continued growth and then culling OpEx and taking out OpEx. We are obviously ahead of our expectations on EBITDA, and I think a lot of that is some of this AI work just became more lucrative to us than we expected, faster than we expected.
I do not think it takes a gigantic leap of faith to see how we can get to a really robust EBITDA number there. Then I think it is just about growth and getting us closer and closer to that Rule of 40.
Then the other one I had was just on capital allocation. You've got a buyback authorization. You've done a lot of M&A historically. You've done work on the capital structure to term that out. How are you thinking about capital return? M&A versus reinvesting in the business?
I think we do not need to use cash on the balance sheet to reinvest in the business. The business is self-sustaining and operating leverage is growing, so we're sending out more cash than we need. I think capital allocation is a dynamic question. It's not static. I think today, I would suspect the best use of our best return on our capital is our own shares.
But that can change if a great unique acquisition comes up that we get, like Bridg. I would do Bridg instead of buy back our shares because we think Bridg will unlock a lot more over time. I think it's just where you are at the time and what generates the highest return. But we still don't want to be sitting on it because I think our shares would provide a really substantial return. So we'd probably look to do something there.
Got it. Just in the last minute or two here, how are you thinking about big picture, what is next for the company, specifically for PAR over the next 12 to 18 months?
For us, it's pretty simple. We want to drive the profitability we talked about and deliver that return to our shareholders and become a cash flow story. Then we want to prove the PAR Intelligence story. We've got to prove that AI is a growth accelerator for us. I think that if we can do that, I think we'll have a multiple rerating that hopefully reward the patience of what's going on now. But it's pretty simple. We've got to drive super high ROI on our AI spend, and then we've got to deliver on the cash flow numbers.
Got it. Well, Savneet, thanks for joining us today.
Thanks, Will.
Really enjoyed the conversation.