Okay. Welcome, everyone, to day three, the final day of Citi’s Global TMT Conference. We, of course, saved the best for last. We have the team from Roper here, Neil and Jason. Welcome.
Thanks.
Thanks for having us.
Maybe just to start, Roper’s a company that has undergone several evolutions in its history. If you could just walk us through the high-level journey for anyone kind of newer to the story, and then where the business is today, and how you think about the setting up for the next phase.
Thanks for having us, and thanks for the opportunity to just go through a little bit of the transformation of the enterprise. Most may not know this, but we grew up as an industrial business. Jason joined 20 years ago. I joined 15. We're 75%-80% industrial pumps and valves then. We closed that chapter early in 2022 and divested all those businesses. About 40% of our enterprise, revenue-wise, we had divested those businesses and went full towards vertical market software and technology products that were not cyclical. The history, though, between sort of the old and the new is there's still a lot of common components. In the industrial days to the current portfolio, it's been we like leaders in small markets, where there's strong product market fit, high gross margins, so we compete on customer intimacy.
A tremendous amount of asset efficiency, so high cash flow generative assets. It had been consistent between the two chapters. Once we divested the industrial businesses, then it really became about growth. We have been on this growth transformation, about the leadership, growth mindset, the assets in a portfolio, the enablement of growth. We can get into that later. I am sure you will have some questions. Then obviously, a couple of years ago, it is another transformation to become AI native in everything we do. It has been the one constant for the last decade or so in Roper has been this pace of change. We have become very comfortable with it. There again, there is this steady through line of the portfolio. Leaders, small markets, customer intimacy, high gross margin, high gross retention, compete on intimacy.
The enterprise today is about $8.5 billion of revenue, 40% EBITDA margins, 30-ish percent free cash flow margins. We then take all that cash flow that the 29 business units generate. Then we bring it to the center, and we deploy it on the next best idea, whether it is a new acquisition, tuck-in, or platform, or the buyback. Taken together, it is a mid-teens cash flow per share compounding, sort of low range of outcome enterprise. We go to work every day to sort of turn that into a high teens. We are not there yet consistently, but we are working to make the double rate at every four years versus every five.
Great. I think that was a great overview. When I think about the Roper model, decentralization is one of the first things that comes to mind. It has been one of the core defining characteristics of the business. Yet, as you kind of alluded to, there has been a lot of work under the hood of developing these kind of shared infrastructure that these different portfolio businesses can lean on. You already kind of alluded to some of these pieces, but maybe you could talk about that journey of the development under the hood, kind of how you still think about this decentralized approach versus how you are kind of supporting these businesses from the center.
Sure. First, just the context for why decentralization. As I have said now twice, the portfolio construct is we are the largest player in a small market. Which means, by definition, all of our competitors are smaller. It is a common thing they all have, which means probably they are able to execute at a fast, nimble pace. To match the pace of the competitive environment in which we compete, we need to be highly nimble and fast in execution. A huge byproduct of being decentralized is we have massive amounts of accountability, like ownership, entrepreneurial-based accountability. We match the portfolio construct, the competitive environment, with the organizational structure. We are deeply committed to it. About 20,000 people in the company, and there is about 130-ish at the corporate center, to give you a sense of the scale of this.
Now, at the same time, we have 29 businesses that while the end market's very different, the business model's remarkably similar, so there's tons of pattern recognition on what the best practices are in terms of how to build a business that is very effective at improving its organic growth rate over the arc of time. How to do strategy well. How do you run a talent offense? How do you run a structured operating environment? How do you build it on principles of continuous improvement? How do you become AI native in everything you do? How do you drive faster product velocity, commercial excellence? This growth system is something that we sort of put together over the arc of time, and it's very much the combination of the ands.
How do you lean into this decentralized, local accountability, highly autonomous structure, but then avail all these sort of best practices of the portfolio to these leaders so they can be most effective in the way they operate their business?
Got it. One of the pieces here you referenced was talent offense. I think that's maybe, I would argue, one of the more underappreciated parts of the Roper journey over the last few years. If you could just talk about maybe some of the more specific impactful recent hires as you see it, and then more generally, kind of the infrastructure of identifying, and developing, and placing leadership.
We have been working on this for a long time. For a decade, I mean, eight or nine years, we've sort of introduced the concept of a talent offense to our business units. We have a lot of admiration for a company called Stryker. It's a medical products business, like a remarkable business, and a lot of their success is built on their talent offense. We benchmarked that, understood from their CEO and their teams what they do, and brought a lot of the core principles to what we do. I will say that, like most things in business, this is not like a particularly novel idea, but the execution or the brilliance or the good things that come from it accrue because you're consistently applying the principles. With a great amount, well, a high degree of rigor.
And so we've done this, and all 29 businesses runs a talent offense. It's focused on employee engagement, employee selection, employee development. At the leader level and the employee base, there's management routines around this. I think every business, twice a year, four to six hours, is a deep talent review where they're making commitments across all these dimensions. We've seen the enterprise engagement score, as measured by Gallup, across the enterprise go from the mid-50s to the mid-70s over this period of time. So becoming truly top quartile in terms of employee engagement and sort of what you get, the discretionary effort. This has been a consistent thing. Now when we see these things happen, better selection that's objective, better engagement, better development of the high potential talent, you see a very clear corresponding increase in business results in the business.
You see employee attrition go down, you see product velocity go up, you see commercial excellence go up, and you ultimately see growth go up. This is very clear connection. But it takes time. There's no silver bullet. This just takes the grinding effort of building sort of true talent DNA in the business. On top of that, when we said in the second transition, our transformation here going from the industrial business to a more growth-oriented software business, we've had to re-sort of think what a growth-oriented leader looks like inside the business. We're very behavioral based. We hire leaders that are competitive, intensely competitive, curious learners that love to build, that can connect a long term to what every person in the organization does tomorrow.
We hire on these attributes, and we get this intense growth mindset at our leader level that then sort of sets into their organizations, and so you can have a growth system and all the best practice sharing, but if you don't have a growth-oriented leader, it's for naught. When we look at our portfolio of leaders, it's substantially more growth oriented today than six or seven years ago. We're seeing more internal promotions, from leader, from CEO to CEO. I think four of the last six or seven CEO promotions or hires have been internal. Then finally, on a talent perspective, we've supplemented from an AI perspective, a tremendous team of about 30 people that then is an accelerator team that's now working with our technical teams across our portfolio to sort of go faster. Yeah, talent is a huge focus for us.
Anything you want to add to that? No, covered a lot of ground?
No.
Okay.
Great. Maybe we can segue there into that AI centralized resources and the team that you've developed there, the accelerator team. Maybe if you could just talk about the components of the value that they're delivering to the portfolio businesses. How far does that extend, and what do you see as kind of the long-term vision?
Do you want to start?
Sure. I can start. We have hired around 30 folks. We started with two last October, so it has been, I would say, successful. The model is they are either going to be, it is kind of a continuum of providing consultative support all the way to working shoulder-to-shoulder with our technical teams, as Neil said. Really it is the businesses are serving up all their ideas around agentic workflow that could be agentified, and then our teams are helping to actually help build their teaching and then helping to build, and then the goal is for them, after they have teached and they have built, to then dismount and go to the next company. That is really the process and, I would say, so far it has gone really well.
You are always going to skin your knees a couple times with maybe some folks in the business that maybe do not work at the same speed as this team. What you are also seeing is there are others in the organizations that have raised their hand and say, "I want to be part of this." "I want to be part of this next wave of development." So they are working at sort of 10X in terms of agentic development of code, and that has also been an accelerator across the portfolio.
I think there's clearly an opportunity there on the product development side. I'm curious now, it seems like the main or an increasing focus is the commercialization side. Is there a component of that centralized resource today or maybe in the future that can help your businesses come up the curve on that piece of it?
Mm-hmm. The short answer is yes. Are we equipped to do it today? No. It's been about what we've added to the central sort of kit, if you will, in the last 12 to 18 months, is a true point of view on product velocity. Like organizationally, how do you drive faster product velocity? This is more than just using the models. It's like the whole organizational design approach, team structure, what they focus on, proximity to customer, so product velocity, and then obviously the enablement of the agentic use cases with the center team. We're going through a massive amount of commercial learnings right now.
Now that we have products, agentic products that are hitting the market across, I don't know, 13 to 15, 12 to 15 of our 21 software companies, now we're starting to see the patterns emerge about how do you sell, how do you price, how do you deploy, how do you drive utilization? And we very much could envision in the growth system, if you will, a commercial excellence function or a capability. I don't want to overstate it. I mean, these are like one or two or three people. These are like culling best practices and then teaching them. It's not like they're doing it. But yeah, we could definitely see commercial excellence as part of the kit.
Got it. When you think about, there's been a few notable successes in the portfolio businesses in terms of development starting to come up the curve, and on the commercial side, seeing some notable impacts. When you think about how to broaden that across the portfolio, how that journey has gone, what kind of comes to mind in terms of maybe patterns of where you're seeing more and less success, and how to kind of bring the median company up more quickly?
Just to make sure we answer the exact question. On the commercial side or development side or both?
Maybe both.
Okay. On the development side, I mean, again, the good news is vertical market, our pattern is vertical markets, leaders in vertical markets and the surface area or the remit of software has just expanded massively because of AI. Obviously, AI is about what software can do versus just how it's delivered, which is on-prem mobility and SaaS. It's about what software can do, and it just can do more. And when you're close to these teams have to go do is incredible, like high ROI ideas. That's not the gating problem. The next gating problem is then how do we get to these products fast and furious, hence the velocity in the AI team. I think some of the patterns, by the way, is smaller teams are better than large, obviously. Closer to the customer is better than not.
Solving problems versus creating features is what the objectives should be. And Jason said it a few times, probably ultimately a smaller number of total development headcount, but the dollars might not be less because you're going to pay more for the 10X engineers, and they're going to obviously have the token cost that sort of companions with them. We're not viewing this as a productivity as a savings, but more how do you get more throughput out of the system. On commercialization, I think the early patterns are how to monetize and how to price. What's very clear across the portfolio from our customer conversations is, it's not going to be straight consumption-based. Customers have to have some sense of what the budget's going to be, what the spend's going to be.
The bulk of what we're doing is, I'll call it an old school cell phone or mobile plan, which is you buy a chunk of credits or usage capability against a tool, but it's going to be a fixed amount, whatever the number is, $50,000 a year for X amount of consumption. But it's going to be use it or lose it, so we can take it as recurring as opposed to reoccurring revenue. As you go through that, you need to buy the next bolus, you'll buy another sort of subscription tier on top of that, another 20 or 50 or other numbers. So, you have some sense of planning and predictability. You also get commitment from the customer. This is not about just trialing and testing.
It's like making a commitment that I'm going to change my workflows to do this automated task, and you're going to make some commitment to the dollars. So, we would expect, I think that's the pattern that's very clear. There'll be some exceptions to that on the tails of the portfolio, but that appears to be the primary monetization sort of methodology.
Interesting. You referenced kind of high ROI workflows. I think that's been a kind of a core theme debate at this conference in terms of how do we think about quantifying and capturing ROI from these AI products. When you think about the most tangible proof points, the highest impact use cases that you've seen
Yeah.
What are some things that come to mind?
I'll give you a couple. Three come to mind in this example of low dollar, high volume, I'll get a relatively high dollar, medium volume, and then very high dollar, low volume. The use cases are slightly different. We have a business that's in pharmacy automation for pharmacies that serve long-term care and skilled nursing facilities. These are not like the retail pharmacies. You wouldn't even see a sign on the door, but they're basically factories that are tens of thousands of prescriptions get pumped out on a daily basis. Every one of those prescriptions has to be verified by a pharmacist today, manually verified. The time and motion study on that, it costs $1-$2 of human labor to verify one of these prescriptions.
Our first agentic tool does that at a quality level that's higher than a human, and we charge about $0.30. It gives you a sense of the value capture and the value share with the customer. Very tangible hard ROI from the day you turn on the product. Similar concept, different end market in transportation. We are today, those that know us, we own a business called DAT. It is the organizing force network for the spot market in North America for trucking. We organize all the brokers and all the carriers in the marketplace, and we're the hitching post to where 750,000 loads are brokered or advertised every day, unique loads. We monetize that today on a subscription on both sides. To make a load, a brokerage happen, it takes about 10 phone calls between the broker and the carrier.
There's about $100-$200 of labor on the side of the broker to make that happen. We have the capability in-market today to do the full matching, the full transaction closing, the factoring of that transaction, the financial reconciliation of that transaction for $40-$60, depending on the transaction. Again, very hard ROI. The final one I would say is in our utility business. We have, they're talking about like a niche of a niche. We're the leader in tax accounting software for investor-owned utilities for their assets. It's a niche, within a niche, within a niche. There's a process that has to happen. All these utilities are investing all their capital, and they can only get a return on the capital in the rate base, once they get it perfect. I'm sorry?
In service.
In service, and they sort of perfect the asset. They get it in, so that's when they start getting a return on it. Today, to sort of put the assets in service and sort of curate the asset, it's a consulting project. So, that's why these utilities have billions of dollars of whip that they're not earning a return on. So, we have an agentic use case that's releasing right now that does that for the customers in lieu of a consulting service. And that's a button that'll be pressed a few times a year but has high value, as opposed to the button that's pressed thousands of times a day in the pharmacy automation. But you get a sense that a range of hard value that happens.
Interesting. To shift gears a little bit, last earnings call, you called out an expectation for maybe a potential warming of the M&A environment. It's obviously been slow this year. Maybe if you could just talk through your confidence there of maybe potentially seeing some activity towards the back end of this year and into next year. What are kind of the tangible points that you're seeing there?
Yeah, I think, well, that we've been talking about this for years, right? The DPI pressure continues. The hold periods continue. There's not many transactions that have happened this year in software, as we all know. There's credit cliffs that are coming in the next couple of years as well. So to be able to recap is going to be very difficult for the sponsors. So all those, that's sort of at the macro, I'd say, and the kind of what we're observing on the ground is just much more constructive conversations about value. We're getting a lot of access management meetings, sort of on one-off basis, getting a lot more access to information than we normally would. So I do think to keep the wheels moving in private equity, there does need to be some transactions. So we feel good. Again, just time's been in our benefit.
We're going to be super patient. On the bolt-ons, that's been much more liquid, much more active for us right now. We're putting a lot of deposits for platforms. We think that's probably going to be more of a 2027 timeline versus later this year. But, that's sort of what we're seeing on the ground.
Yeah. That's good.
On your approach to identifying and selecting potential targets, obviously, this is a very dynamic time in software. You guys have historically had more of a risk-averse mindset, trying to bat 1,000 as best you can. How has maybe that targeting changed at all? How do you think about incorporating a company's AI strategy? What are kind of the things you would look for to say, "We think this company is maybe better or worse positioned?
Yeah. We are definitely, the whole concept of a cash flow per share compounder is you need to have a low range of outcome, right? Our whole business model's built on a steady, consistent, low range of outcome, not highly cyclical sort of compounding methodology. Inside of the portfolio over the years, we believe that the best way to do that is you need each individual asset to have the same sort of low range of outcome, as opposed to have a bunch and assume the portfolio effect dampens it. We've always been, I call it, like a low range of outcome sort of investor. With AI, certainly it adds to the range of outcomes. We have to just assess that. But we've always been the ones that sort of try to evaluate existential risk.
If a company can go to a zero, conceptually a zero in the Monte Carlo, we're just out. Even though it could have like a 5X upside, it can go to zero, that's not us. That's the opposite of a low range of outcome. When it comes to AI, we have both an offensive and defensive moat framework that we built for our own portfolio. We apply it to every acquisition going forward. Increasingly, it's more objective versus subjective evaluation of the criteria. It's things like network effects, regulatory overlay, obviously the level and amount of proprietary data and the knowledge graphs of the various companies, the configuration density of the software, things like that we look at.
Regulatory.
Regulatory overlay, for sure. And certainly then the stability of the end customer. You can have a business that might score very well on all those dimensions, but if the customer's going to be fundamentally disintermediated by AI, then you don't want to be in sort of that end market. Right now, we apply the framework to our best ability as objectively as we can. We obviously have the pattern from our own portfolio we apply perspectively to these targets, and this only is going to get easier over time as the AI winners emerge and you see more bookings contribution, more revenue contribution from the agentic SKUs, and they're emerging as the winner, and they have the RL on their products. They're getting higher quality and more deterministic in their outcome. And those are going to be the clear and obvious ones to acquire.
But that's going to be a couple, three years from now, not right now. So we have to be perspective in the analysis on that. That's how we think about it.
Are there any more quantitative metrics? I think the framework you laid out makes a lot of sense. Are there any more quantitative metrics that you're able to use today? I think, for example, in the public software space, a lot of people are looking for acceleration stories. Is that something that's.
Certainly.
Part of your framework as well?
Certainly. All this is about, we fundamentally believe, I think most people, or this is not a fact in dispute, that AI is TAM expanding for software. The debate is who's going to claim that TAM expansion. We have a strong thesis that incumbents have a high, high, high right to win for all the reasons we get into or we've been into in the past. But yeah, ultimately this has to manifest into accelerating growth, positive second derivative growth, for sure.
Maybe just more broadly on your capital allocation framework, it seemed like that's been shifting. Obviously, you had the buyback this year, reduced share count pretty meaningfully. If you could talk about positioning yourself, developing kind of a dry powder bank. How's that mindset changed?
Yeah, we're really pleased with the buyback over the last eight months. But back to the point of what I was talking about earlier in terms of the opportunity and the option value around M&A, we think it's a unique opportunity, so we are de-levering right now and preparing our balance sheet, in the belief that things will thaw in 2027. So we've got the proceeds from our Indicor minority investment that we've got the instrumentation piece with flow control, probably get in the fourth quarter, so that's about $1.6 billion or so, and then we'll have strong second half cash flow. So that'll prepare us and allow us to be able to do $5 billion or more of M&A in the next 12 months.
In terms of your decision framework between buybacks and M&A.
How is that?
Look, like I said, it's an option. If the M&A markets don't cooperate, we have the buyback. It's really just a time value of money sort of deferral of the buyback if the M&A markets aren't cooperative.
I think it's important that the folks here listening, we say this on repeat, but for us, capital allocation is simply just a math exercise. There's no emotion tied to this in any capacity. What did we say at the beginning? We're trying to build the best cash flow per share compounder. That is what we're attempting to build, we're working to build, we think about all day, every day. We're not trying to build the biggest fill-in-the-blank, transportation software business. That gives us tremendous optionality and tremendous discipline in just what the math says for the best long-term cash flow per share compounding. Now, we focus on the five- and seven-year comp, or about compounding, so not what's the best in the next 12 months, but what's best in the next five to seven years.
Mathematically, when we did the acquisitions of the church management and autism therapy acquisitions 18 months ago, or 12, 15, 18 months ago, this was a different valuation era. We paid 22 times the next 12 months for those deals. We underwrote those to be about nine or 10 times year five EBITDA. At the time, we could've bought our stock back at about 15 times.
Year five.
I'm sorry, year five. We could've bought ourselves back at year five at about 15 times. About a 50% risk premium to deploy the balance sheet towards acquisitions. Then valuation changed, our valuation changed, and we could all of a sudden buy ourself back at nine or 10 times year five. The private markets were still wanting to price on yesterday's valuation, so it made no sense. Boyd could've bought any company he wanted for 22 times trailing or forward, but it makes no sense because there's no risk premium attached to it. Now we think that math is going to tilt even further towards M&A, right?
If you can get a 20% to 40% discount on the private companies versus where we paid a year or 18 months ago, now we're going to underwrite to something like six, seven, eight times year five EBITDA, and you're back in the risk premium that makes sense, and the cash flow per share compounding accelerates from that point forward. It's just a math exercise. Any time that math changes, it changes the way we think about how we allocate the capital.
Interesting. Maybe I'll pause there if there's any questions in the audience for these gentlemen. We have a mic coming to you. One moment.
Just to pick up on what you said there about capital allocation being a math exercise. Just kind of wanted to know, and I heard your point about de-levering here into 2027, how does the balance sheet fit into that math exercise?
Yeah. We run an investment grade, but levered play, right? We attempt to be between 3x to 3.5x leverage over a long arc of time, sometimes above, sometimes below. It just provides a cost of capital advantage versus the people we compete with. We borrow at, whatever, 5.5% ± , and our competitors borrow at 10% ± in private equity. It's a cost of capital advantage for us, and we run the levered play, which helps on the equity return.
Great. Any.
A modestly levered play. We're always going to stay investment grade. Yeah.
Any other questions out there? Okay, maybe we can just talk about the organic growth outlook. I think there had been this objective of returning the business back to kind of high single digit organic growth. There has been a few kind of stumbling blocks along the way that derailed that. Each sort of feels idiosyncratic, but when you sort of stack them all together, Deltek, Neptune, arguably DAT.
Yep.
If you could just talk through some of those, to what extent do you feel like those headwinds are behind us? Where do you still see maybe potential to linger or a wider range of outcomes?
Sure. So, yeah, with Deltek, they are heavily exposed to government contracting space. It has been, as you can imagine, last year it was idiosyncratic when you had DOGE, you had agencies that were sort of gutted, and then we had the prolonged shutdown. This year, starting to see appropriations flow. We are second and third sort of derivative demand on that because they got to get the money, and then they provision the licenses. Usually now it is more when they actually get the funds. It used to be a little, we used to not have the sort of lag. But we do think there should be some recovery, the timing of which is still sort of TBD. We saw some green shoots in the second quarter, but we are not ready to call a market yet. But when that does happen, we should be the beneficiary.
There is a lot of OBBB that is going to be flowing in for Deltek. The other sort of catalyst for Deltek, and this is more bookings, probably not in-period sort of revenues, they have announced end of support on their Costpoint solution for the first quarter of 2028. That will not all happen up through 2028. It will still continue. Customers will stay on-prem. But that is a catalyst, to move from on-prem to the cloud, which should move it over two times, from maintenance to subscription. That is sort of what is going on at Deltek. DAT, I think we have mentioned it, we have seen a freight recession. We had a freight recession for three or four years. It has been positive in the first half of this year. We hope that continues. It is more of a supply-driven rally, and so we are a little cautious, right? We are not seeing constructive demand.
But it is certainly, we like the trends in that business. At Neptune, it is sort of, I would say kind of doing okay, sort of bouncing along the bottom. They did perform better in the first half than we thought they would. We will see how. The market sort of is still settling from COVID. I would say it is not completely, we are not out of the woods yet. But it is in our guide, and we would expect that in the coming years, volumes will start to normalize back. The good news at Neptune is we have been able to sort of offset some of the market volume declines with the move to static from mechanical meters. That has been the story at Neptune.
Okay. I think that is great color. Maybe if we think about just bridging that organic growth to the free cash flow per share framework that you gave, mid-teens growth. Maybe you could walk us through the math to mid-teens growth, and then how do we bridge that up to the aspiration of high teens?
Yeah. I think it is simple, right? Getting that organic growth is a major component of that. So that is one component, you get a point or 2 of organic growth. Then I think capturing more value out of M&A, and that is really just some of these maturing leaders. Obviously, as Neil mentioned, paid a little bit more upfront for those. But as they start to compound over time, then you are capturing more value out of M&A. I would also say the bolt-ons, doing more bolt-ons. So we are doing a lot more bolt-ons than we did 3 or 4 years ago, and those are obviously very cash accretive, but they also improve the organic growth. So it sort of ties back to increasing the organic growth of our platform. So that is how you bridge the mid-teens to high teens.
Yeah. The baseline is, think you have, and not any like for this year or next year, more like the model, you have 6, 7% organic growth. You get a little bit of margin expansion, so sort of high single digit, sort of organic EBITDA growth. Then we have our baseline capital deployment, which then bridges you to mid-teens. Then as we get a little bit more organic growth and a little bit more return on the baseline capital deployment, it goes from mid-teens to high teens.
Perfect. Maybe just with our final minute, what gets you most excited about the position of the future and the journey ahead over the next few years?
I've never been more excited than I have in my career about that opportunity in front of us with AI. The resonance that our customers have with things, we're literally delivering a button that when you press it, magic happens. I've always been in the boring part of software, and all of a sudden software is not boring. Just the reaction, like I've said this a couple of times, but the guy who runs a local health system in Sarasota, we had lunch. He's like, "Hey, I hear you have a pretty cool new AI tool with a company called Strata. We'd love to learn more about it." That never happens. Then we have a conversation. They got into early developer group, and we just did a great, teach-in with 1,000 of Strata's customers a couple of weeks ago on launch day.
It's just really fun to sort of see the magic that's happening in the product base and the reaction from the customers.
Great. I think we'll leave it there. Thank you everyone for coming.
Thank you, guys. Take care.
Thank you, gentlemen, for being here.