Constellation Software Annual Meeting 2026 Transcript

Read the full Constellation Software Annual Meeting 2026 Transcript with insights into AI, M&A, and capital allocation.

You are currently viewing a placeholder content from YouTube. To access the actual content, click the button below. Please note that doing so will share data with third-party providers.

More Information

Unblock content Accept required service and unblock content

Easily discover all the topics of this interview transcript by clicking on the table of contents:

Tilman’s key take aways – Constellation Software Annual Meeting 2026 – Transcript

Here are my 10 take aways from the CSU AGM. Around 450 people were there in person, with over 1,000 joining virtually.

1. AI can be an opportunity: Contrary to current market sentiment, Constellation seems bullish on AI. They don’t see it as an existential threat — they see it as an opportunity to serve customers better.
2. Building is only half the equation: Yes, AI makes it cheaper and faster to build and ship software. But you still have to sell it. That requires trust, relationships, and a track record — things most start-ups simply miss.
3. Constellation has those relationships: Decades of embedded customer relationships are a genuine moat. That’s not something you can replicate with a better model or a lower price point.
4. Decentralized AI strategy — by design: There is no top-down AI mandate at Constellation. Instead, they use “AI ambassadors” who move between their businesses, sharing ideas and best practices. Think of it like a beehive: distributed, self-organizing, resilient.
5. Shipping velocity has dramatically increased: What used to take months now takes weeks — or even days. This means faster iteration, more upgrades, and happier customers.
6. The core model still looks intact: Based on what was discussed at the meeting, there’s no sign the fundamental Constellation flywheel is broken. If anything, AI seems to be accelerating it.
7. Constellation is becoming more public: Historically, Constellation has been famously opaque. That’s changing. They are actively working to share more with investors and shareholders going forward.
8. The incentive system creates extraordinary alignment: Many employees were asked to buy shares. The panels at the meeting were filled with people who are now millionaires — or multimillionaires — because they did. That alignment is rare.
9. The AI fear in software may be overstated: Focusing only on the threat side misses the opportunity side — especially for companies with existing customer relationships and proven systems.
10. The AI pain is hitting investors harder than the businesses: Investors in software companies have seen losses of 50% or more in some positions. The underlying businesses tell a different story. At the operational level, AI can look more like a tailwind: new features ship faster, customers are actively willing to pay more for AI-enabled functionality, and legacy systems plus regulation mean the transition away from existing software takes years.

Introduction to the Constellation Software Annual Meeting 2026 – Transcript

[00:06:40] Larry: What a run. Uh what a a three bloody good decades as as Mark Leonard might say. I don’t know what your favorite figures are, most important takeaways. I think that final screen on the left showing capital resources and um no very limited um leverage. But um your your takeaways may vary. And incidentally, the number of countries in that panel is just the number of countries where acquisitions were made. Um Constellation operates in pretty nearly every country in the world and we’ll continue um that impressive trajectory. So that’s the history. Now to talk about the next steps, the coming decades, i’d like to ask you to join me in welcoming the president of Constellation Software, Mark Miller.

[00:07:40] Mark Miller: Thank you so much, Larry. Hey Larry, thank you so much for seeing the uh the event. And uh welcome to our shareholder meeting for uh our all of our our companies. So uh really looking forward to uh you telling you a few stories before we get started with the the uh the question and answer. And uh I have to tell you, I’m very privileged to have this role because I get to work with uh a lot of tremendous people, a lot here in the room and around the world, as we continue to build this business out, right? So uh and we really are the sum of our parts, and I want to want to sort of show you that and demonstrate to you how much we care.

I also want to thank Mark Leonard for his his guidance and in bringing us through these last three decades. Uh I’ve lived through them myself, i’ve been here for over 30 years and we’ve always been on a perpetual journey of learning. And every day you learn something new, and I’ve learned things, you know, just by listening to our business leaders, talking to my peers across Constellation, and I continue to do so today. And I think all of our leaders do here, which makes us a very unique company. So I want to thank everybody for that.

I want to explain to everybody, which is one of, you know, in talking to people over time who don’t know know a lot about Constellation, is we’re not this—we’re not a company that is a strong central control over what’s done. We really believe in empowering our business leaders who are close to our customers. There’s no CIO who decides what AI platform we use or what we should do, or whether we should turn right or left or do this or that. That’s done by our business leaders across the world, and that helps us a lot because it lets us bring new companies on board that want to be part of this ecosystem and learn from this environment.

But I want to make sure you understand this when you’re comparing us to other companies: we are very decentralized. We believe in this. We believe that our businesses all across across the world—and as you can see, there’s hundreds and hundreds of them—are the ones that should be making the decisions. They’re the ones that are running each of those businesses. They’re the ones that are listening to their customers and building their products. And in the advent of AI, you’ll see that that’s very important to understand. We’re not dictating what to do; we’re allowing our leaders to listen and learn from each other. And so there’s tons of CIOs across and tons of heads of professional service and heads of marketing and heads of sales. We’re there to coach and help, but you know, you tend to learn best from your peers than you do from your boss, and we’ve tried to create an organization that does that at scale, and it’s helped us grow this company over time.

But I just want to go back to some of the basics for some of the new our new shareholders or our new people who haven’t really heard us before, is we don’t believe in fix or flip. When we talk to a business when it joins us, we believe in buying and holding forever, and it changes the things you do. One of my favorite interview questions is um to some some new person who’s looking for a job is is what would you do differently if you owned a business for 5 years versus forever? And you do a lot of things differently, and we do a lot of things differently. And developing people and peer learning is very very very important to us.

People tell you scale matters. Scale can get you up and to the right really really fast. Well, we believe in decentralization, we believe in customer intimacy, we believe in listening to our customers and building products for them in the geography they’re in, in the vertical market they’re in, and we really believe that. And we’re willing to give up scale for customer intimacy, and we do that across the world. We’ve broken up many businesses into smaller businesses. So we don’t think scale matters. We believe decentralization, empowering people matter. Allows you to develop a lot more leaders, a lot more successors, allows it—just creates an environment that a lot more people can be included in. It’s a very inclusive environment.

So very important to me that some people will tell you culture matters. “Mark, what’s the culture at Constellation, how is it?” Well, I don’t believe in one corporate culture. We believe in a culture of cultures. Even the same even the a two companies in the same city in the same country can have two very different cultures, that’s okay. And that’s important because one of the advantages of joining Constellation is we’re not here to change your culture. We’ll measure you. Some of the events we ran at used to have people wear their numbers on their main their name tags. You need to be measured if we’re going to trust you with, you know, to run that your business all over the world. We need you to be measured, but you should determine what your culture is and how that works because that helps you address your customers better and solve problems that they really need to help them run their businesses.

And I’m not a big fan of parachuting in top talent. We’ve definitely done some, we’ve hired some amazing people who have helped us. We really believe in peer learning. We really want people to learn from each other and create an environment where they do that themselves, but we do help them get together and we’ll talk a little bit more about that. And you’re going to get a chance to talk some of our leaders on stage, we’ll talk about that as well. So this is what we’ve built, and I think we’ve done it very differently than most people would tell you to have done it. I don’t think if you look back at Constellation—I was part of the first acquisition uh in in 1995 and I could never have foreseen this is what we’re going to do. We learned it along the way. There’s no way you could say, you know, and we probably thought we’d have fewer larger businesses maybe back then, but we’ve learned that having small businesses—and our businesses are small, a lot of them have dozens of people with, you know, a few hundred customers, sometimes less—is is the way to go. And it turns out, and I’m going to talk a little bit more AI in a bit about AI in a bit, it’s actually positioned us very well to think about that because we are there with the customers helping them as they deal with changes in their businesses through time.

So um I wanted to um introduce Damian McKay because I that was a real opportunity for you to meet some of our vertical leaders across Constellation. We care a lot about verticals. Verticals really help us, you know, be closer to customers and understand what their needs are and sometimes, even though they’re decentralized, share best practices across the world. But uh I thought it’d be kind of fun for you to meet some of our our vertical leaders and I was going to ask uh Damian McKay to uh come on up, and Damian runs Vela, and introduce his panel. Thank you so much, Damian, i appreciate it. So thanks, there you go. Yeah, thank you, there you go. There’s a flicker, i don’t know you need it, you don’t need that do you? I don’t need it, yeah, that’s good.

Morning Presentation Panels (Constellation Software Annual Meeting 2026 Transcript)

Panel: Verticalization Strategy

[00:14:20] Damian McKay: I think we’re going to what what we’re going to do is uh have a quick chat about verticalization. Going to bring up a couple of uh a couple of people. It’s a pleasure to be with you here today. Uh it was great to to see that video and just see what has uh, you know, all those names popping up. I was I’m familiar with some but and and I know the story of others and just the great people we’ve bought into the business uh when we’ve added those those companies and uh you know how those things have compounded. And as we’ve got all those stars, it gives us the opportunity, not not the man you, not the uh uh something that we have to do, but where it makes sense is to align these uh in verticals. So we’re going to talk a little bit about verticalization, and uh we’ve got some of the, you know, some of probably the people who have demonstrated the best uh the best way to do that and grow that within within Constellation. So with that, I’ll I’ll ask uh Santina to introduce herself.

[00:15:21] Santina Allen: Hi everyone, i’m Santina Allen. I lead the healthcare group for Harris. Um I’ve been with Constellation 13 years. Thank you.

[00:15:28] David Nland: Yeah, my name is David Nland. Um I’m CEO of Lumen. Um started in 2013 building up a communications and media business.

[00:15:39] Bill Delaney: Bill, hi. Hi everyone, i’m Bill Delaney. I’m the chief executive officer of Medaxo. Uh I’ve been with Constellation for approaching 15 years.

[00:15:47] Damian McKay: Fantastic, great. Well, what we might might get started with uh u a couple of couple of questions each and we’ll we’ll sort of bounce those off. Tell me, uh uh David, in terms of the your customer intimacy, you know your customers really well, you’re well known in the industry. Uh how does that help you help your customers and then help acquisitions as they come in?

[00:16:20] David Nland: Yeah, well, you know, in communications and media, we have tier one customers and the big operators you would all know like the British Telecoms, the Rogers. And on the media side, the Disneys, the Discoveries—i think we just need a mic for you, sorry about that, i sabotaged my mic here, we’ll go again. Where did I where did I get to? Yeah, yeah, so we’ve got some big customers worldwide, big big brand names, and they have a lot of intelligence and so we learn a lot by speaking to them um by finding out what their problems are and obviously now supporting a portfolio of c of companies that sell into those customers. But also as they navigate their assets, their important assets, they start to share with us ideas about what else we could acquire to help them provide security with some of those assets that are in transient ownership at at this point. So you have to kind of earn the position to have that conversation, it takes some time, but the power of that relationship is unbelievable.

[00:17:24] Damian McKay: Fantastic. And Santina, could you share, you know, a little bit about your vertical and and how it, you know, how it shapes up?

[00:17:29] Santina Allen: Sure, happy to. Um in the healthcare space in North America, we span the ecosystem of the patients journey, but also the the information that surrounds that that journey. From acute hospital systems to physician practice systems to specialty solutions within the hospital, outside of the hospital um thinkies retailies local pharmacies independents uh radiology imaging um long-term care um all the way through to payer solutions in the health and shore tech space. Um these businesses um uh we own sometimes multiples in um in those areas. For instance, in the physician ambulatory space, we own a dozen different ambulatory systems that serve the the physi the private physician practices, that’s out of hundreds that are in our market. Um so um these uh these systems of record um uh, you know, provide us with that um different uh different views on that particular market. So those um systems that we have, some serve large physician practices um up to 50 physic 50 physicians per practice, uh sometimes very small, one to two physicians in different geographies as well as specialties around uh around uh the world.

Collaboration of those businesses gives us, um you know, opportunities to learn about the unique space, the demands, the regulatory environment. Um so physician office practices being one one area. The the payer space is another area that’s interesting for us. Um we own the um the leading platform in Medicare shopping enrollment quoting and and retention for the over seven, you know, over 60 million um members in Medicare in the US. Um that solution paired with um a solution that helps those um health plans develop the product management, the products that they develop and bring to market for um uh members, again we can collaborate uh collaboration uh leads to um uh advancements in AI within our within our group, cross-selling, so lots of different benefits from that ecosystem that we serve. Done a great job.

[00:19:46] Damian McKay: Fantastic. And and Bill, you’ve uh got a global vertical with assets and how maybe share a little bit about how you built that and what advantages that global nature can bring to your your customers?

[00:20:00] Bill Delaney: So I’ve had a different path. I was uh uh fortunate enough to inherit the uh assets that Constellation had been investing in right from the very start, including Trapeze, so no pressure. Um and in late uh 2019, the decision was taken to bring all those assets together under one parent overlay which we ultimately branded Medaxo. So Medaxo today is the operating group within Constellation that focuses on people mobility, and we cover every element of your journey, right? So you’ve all come here, you may have flown here, you may have come from another city by a train, you have traveled in a private vehicle, a train, a bus, a a tram, uh your kids are going to school on school buses, uh at times we have to carry you in ambulances. Whatever that mode of transport is to get you around your city and between cities, we provide the technologies that allow that to happen. And we sell those technologies to large government customers and and operators who work for them across the whole gamut of everything you can can think of: asset management, people management, scheduling, planning, um fairing, um you name it, we we do it in that that that ecosystem.

[00:21:22] Damian McKay: Awesome. And how do you bring that value sort of on a global scale? Because you’ve got you’ve got those tools and the different markets.

[00:21:28] Bill Delaney: Yeah, it’s it’s interesting, you know, we we learn a lot, right? So we’ve gone very deep for example in transit uh you know, we’ve got uh over 3,000 government customers in 37 different countries, u so our people who are at the coalface with those customers are part of the fabric of the industry and they’re recognized for that and so they know really deeply what’s going on and they’re consulted on that and and we feel part of that community as well and we contribute to it. We sponsor large events, um and if I take something like a big global, you know, uh expo, we had one last year for UITP in uh Hamburg. Massive event, the whole Hamburg expo site was taken up for the first time um it’s massive. Um we were a platinum sponsor, one of three platinum sponsors of the entire event. Medaxo could do that and that provided a halo effect for all the businesses that were there under their own brand in our Medaxo neighborhood. So that was a great value we gave them in terms of, you know, giving them a, you know, a a stage on which to, you know, show their credibility to their customers. But at the same time, we conducted 100 M&A meetings at that event, right? Because, you know, we were there to meet our competitors and we’d arranged to meet them ahead of time. So we we we use the Medaxo family and the Medaxo brand for both of those purposes.

[00:22:53] Damian McKay: Yeah, it’s very well, you know, very well known in the global market.

[00:22:55] Bill Delaney: It is now, right? It took a little bit of time to get there, and I know you’ve had the same journey with Lumen uh David, but uh you know, I think we’ve we’ve, you know, we’re getting bankers and and brokers, you know, bringing deals to us now, particularly as we’ve shown that we can do quite large deals.

[00:23:15] Damian McKay: Yeah, and David, i’d be interested in in your sort of the origin story of Lumen and uh and how you built that up.

[00:23:20] David Nland: Yeah, well, we worked our way up to diamond status, so it’s just one it’s one one above platinum. Um yeah, so it was humble beginnings because we didn’t have anything. I came in as a consultant, an industry domain consultant with a lot of software experience, and uh it was about finding that first business, making that first business a success, building a bit of brand, a bit of credibility, and then starting to leverage that into subsequent deals. So at the beginning, it was fairly modest, you know, it just building up a, you know, relatively small businesses at that time and just learning about the trade of, you know, value acquiring this asset class um and how to get to great returns and to build a compounding rhythm inside the portfolio so that, you know, that really builds momentum. It’s a it’s a rolling thunder and we’re still, you know, really at the beginning of our journey. So, you know, you get the chance to brand, you get the chance to sponsor, you get a chance to have really top level relationships with customers, and it just feels great to be in an opportunity to leverage that moving forward, especially with everything else that’s going on in the world and the opportunities that are going to present themselves.

[00:24:26] Bill Delaney: You know, the the other thing that happens is it’s a great thing for our our people to come. We we brought all 330 of our people together a couple of weeks ago under the Medaxo brand. Now they’re all there rooting for their own business, but when they come into a room and see that they’re part of something bigger, it it lifts the energy, right? And and apart from all the peer sharing that goes on, you know, it’s a and you guys do the same thing, right?

[00:24:48] David Nland: 100%.

[00:24:51] Damian McKay: Yeah, and I think the what what you’ve been able to do in your verticals and and, you know, particularly using Lumen as an example, something that uh, you know, you might not think could be as big as it got, uh now we’re looking at other startup verticals in sort of your image as well. So uh I think a lot of a lot of uh be interesting to see this this um presentation in in 10 uh in 10 years and see what other uh strong verticals we have. Santina, um when a new company joins your team, what sort of value do we bring to the employees and and their customers by being part of something bigger?

[00:25:24] Santina Allen: Yeah, so I think uh when a seller considers um the healthcare group as a place for the forever home for their business, um they do look at the ecosystem that we have, our understanding and appreciation of the challenges that uh that businesses small and large face within healthcare, the highly regulated industry um uh so I think they they um they appreciate that we understand them, they we understand their um their customers, whether it’s uh large health systems, physicians, um whether it’s payers, again the complexities. Each one of those buyers has a different persona, different uh um uh risk profile, if you will, um for you know, which vendors they they invest in. Um and so, again we bring that understanding and so when someone joins us um from the outside, they are surrounded by um peers who understand the the challenges they face.

[00:26:23] Bill Delaney: Before we created Medaxo, I was seeking to buy businesses under the Harris brand, right? And I would go and speak to a transit and they would sort of scratch their heads and wonder, well, what was the strategic value? And within 12 months of creating Medaxo, we had at least two cases where they saw the value then and transacted with us. So I think that that speaks to the value of that approach.

[00:26:40] Damian McKay: I think each of your verticals has got a reputation for people coming to you uh as well. If they’re in the market, they they they know that you’ve got a significant presence, and uh if they want to sell, they’ll be uh coming and talking to you. Bill, uh we’re going to hear a little bit about AI later, but it would be interesting to uh look at AI just from a vertical perspective. What uh additional optionality does a vertical or a group of businesses in the same segment give you when it comes to looking at uh leveraging AI in your business businesses?

[00:27:18] Bill Delaney: Yeah, look, uh we’re getting really excited about this now. Um uh you know, we’ve got 41 businesses today in in in Medaxo across all different functions in every department of say transit operations around the world. Um so what this lets us do is is build a layer above that, that Medaxo can invest in. We still give the federated control to the underlying businesses, but we’re able to build a platform that uh allows a a transit authority to start asking questions and uh uh building actions and agents that ignore all the industry boundaries, ignore all the underlying business and product boundaries and and look at it from an ecosystem. And, you know, i think we’re in a position that none of our competitors are in, in that regard. So from a Medaxo perspective, that’s an investment that we’re making that each of our businesses will be able to to leverage, and uh you know, the the feedback we’re getting from the industry is, you know, they’re they’re sort of blown away by this approach. And uh uh you know, we’re looking forward to to to seeing that really build out in in the coming months.

[00:28:28] Damian McKay: And and Santina, any thoughts on the same topic around how AI brings optionality to your customers within your vertical?

[00:28:35] Santina Allen: Yeah, so within we we are the system of record for so many pieces of clinical data that and data is the key to, you know, if we understand the data, we understand the workflows, AI can only um we can we can bring extra value um through that. I think you know um ambient listening may be an example in the healthcare space. If you’ve not been in a physician’s office and had them ask you to turn on the ambient listening to listen passively to the conversation that that they’re having with you um between the patient and the clinician, but that data is data we’re capturing, right? And that data is valuable to um to to to uh improving improving the outcomes within the healthcare ecosystem. So great.

[00:29:09] Damian McKay: And David, when I’ve spoken to you in the past, you’ve got a great long-term vision for what you want to do with Lumen uh you know, five 10 years. Any could you share what you can around uh you know, what you’re building to? You you’ve built been purposely building building to where you are, interested in a little bit of that, but also where you think you can go, as much as you can share.

[00:29:33] David Nland: Yeah, yeah, well, you got to dream big, right? So when we were small, we said, “We’re going to make you’re going to be a billion dollar company.” And everyone thought I’d been at the bar for too long, but uh and uh you know, we’re within that within range of that objective. So you know, within range of that, we, you know, we set more long-term objectives that are equally, you know, mindblowing, um you know, projecting things out. So we really see there’s really three strategic acquirers in our space ultimately, and every other business in in this industry is in a transient ownership position, whether it’s a strategic private equity or even a founder run businesses. But the final destination needs to be one of the three of us, and we’re just the exciting, interesting alternative to a functional integration synergy-based model, and customers really like our model. So they’re going to steer these companies towards us over the next 5, 10, 15 years, and that really excites me about what size business we will ultimately build, and the adjacencies around our markets are very significant as well. So we’re very very excited about the future.

[00:30:36] Damian McKay: Yeah, and and Santina, for yourself, you’ve you’ve built something that’s uh really large, but the market is so much larger. How do you see, you know, that vision panning out over the next five plus years?

[00:30:50] Santina Allen: I think we’ll continue to to to be a great home um for businesses that uh that are critical to health outcomes in in North America. Um they’ll continue to um find us a a home that uh that provides an understanding of what they what they need to do and what they need to accomplish um but uh that peer-to-peer learning that that we’ve talked—so many of us have talked about—is the key to um unlocking that additional value.

[00:31:18] Damian McKay: Yeah, great. And and maybe Bill, I know you you’re known to be a great uh developer of people and talent. How you maybe could you could touch on some of the, you know, talent opportunities within the within the group that as you’re in a vertical and multiple businesses, independent businesses, but the ability to move around? Anything you’d want to share on how you develop talent, but also think of examples where people have uh been able to leverage something bigger in a within a vertical.

[00:31:44] Bill Delaney: Yeah, uh maybe just to follow on from that theme for a minute. Uh you know, we set an equally ambitious goal when we set up Medaxo in 2020, and we we decided that we wanted to touch a billion journeys every day uh by the end of the decade. Um so as we sit here today, yesterday, tomorrow, we touched 640 million, so that’s getting too close to a BHAG. So we we just reset for 20 2035, and our new goal is to get to two billion uh journeys every day, right? And that’s an amazing thing for our people. I say to our people, “You can go home and tell your kids and your grandchildren that, you know, we touch hundreds of millions of journeys every day.” And uh you know, i think that’s amazing and and that sort of speaks a little bit to the scale of what we do. And and to do that, the biggest thing that’s holding us back is the development of leaders.

So uh we created a—and and it’s not so much, I think we’ve got good programs to build uh business unit leaders within a business. Uh where we were struggling was that step up into group leader roles and portfolio leaders, you know, people becoming leaders of leaders. Uh so we uh brought in some help and we curated a leadership program around that. Uh it’s much more around the soft skills uh that people need to learn because you’ve got to learn how to influence lead through influence, and you need to take a really different strategic mindset. So, you know, we invest heavily in that.

[00:33:15] Damian McKay: I just have to challenge you there, Bill. I think uh you’re a known sandbagger. Two two uh two billion journeys, I think we could lift the bar there, but uh that’s maybe that’s maybe on their internal decks.

[00:33:27] Bill Delaney: Well, you know, I I’m also a fisherman, right? So the fish gets bigger every time I tell the story. So fantastic.

[00:33:38] Damian McKay: Any any other comments people like to make on just how you’ve been able to uh leverage talent or grow talent within a vertical?

[00:33:45] David Nland: Yeah, I mean so it’s it’s number one most important thing, right? Well, customers may be number one, but it’s joint number one is developing talent so we can scale. You just don’t scale unless you’ve got an ongoing, active talent management process and you’re harvesting talent. So you think about putting crops in a field, it takes a certain amount of time. So we got to throw a lot of world-class fertilizer on that field so we can get that talent that harvested quicker and quicker and quicker as we scale um and that’s really that’s really a core competency of a compounding, you know, compounding this asset class is just getting the talent where you need it to be one year ahead of when you need it, and that’s just the constant focus.

[00:34:20] Bill Delaney: The one the one thing I tell my people is the only thing that’s going to hold you back—because we’ve got such strong growth—is if you don’t have somebody ready to take over your role. And and that’s a really big mindset change for people, that when they’re recruiting people, they really need to bring in really strong contenders who are capable of stepping up into their role. And uh I think that’s been a really critical approach to—goes back to the enduring business, right? Again, we have to have the leaders who are at the table when we’re not there, the next round. Yeah.

[00:34:50] Damian McKay: Fantastic. Well, I mean, we we’ve got uh we’ve got the capital, we’ve got the—there’s so many opportunities out there we can go after. We do need to make sure we’ve got the talent, and then I think we don’t have to always organize in a vertical, but when we do it, it also adds some uh some extra compounding power in there as well. With that, i’d like to uh thank you for uh your contributions, not only obviously today, but what you each of you have built in your verticals and the, you know, the foundation you’ve set up. Again, really looking forward to seeing how that’s going to look over uh over the next decade. And with that, I’ll hand back to uh to Mark. Thank you very much. Thank you. Thank you.

[00:35:38] Mark Miller: So I don’t want to get in the way of AI so um so yeah, and we’re not changing the name of Constellation by the way, because you know, it it’s who you are, not who you say you are. What you do, not what you say you’re going to do. So uh I just wanted to get this uh next slide up, remind you of before I discuss AI of how we work. It isn’t, you know, myself picking up the phone and telling someone that this is how we’re going to do it. It’s letting each of these leaders in each of these businesses think about it. And also, when you’re thinking about in a forever time frame, you’re thinking about buying and holding forever. Remember I said that was not always a fun question to interview, you know, you’re—you want to think about these things because your customers and your employees are thinking like that. They want to know what you’re going to do to help them.

So we really structured our organization to manage change, whether it’s been through—well, we’ve gone through Y2K, you know, we’ve gone through mobile computing, we’ve gone through SAS. We’ve lived through multiple technology changes, and it isn’t because we’ve had a directive from Constellation headquarters as to what to do; our business leaders decided how to deal with that situation. So when SAS came, it wasn’t like, “Oh, everybody should be doing SAS because SAS is what’s what’s the what’s the most important trend and everybody needs to be on it, and you better do that or else.” We’re not doing it that way. We’re allowing our leaders to learn, but we’re creating uh an environment—see if I flipped, there we go—so where our leaders are learning.

I mean, Bill talked about an event in Copenhagen, there there are learning events going on continuously across Constellation, not just about AI. These—these are AI ones in particular, um and but we we do that across operating groups. Uh we do it locally in certain countries—we’ll do something in Brazil, we’ll do something in in Germany or in the UK, we’ll do it in—there’s one in Toronto going on right now up near the airport. Uh it’s called an AI accelerator that a few of our operating groups are at, and uh I think it’s a five-day event where they have a room for the developers, a room for the product managers, and a room for the business unit leader heads. And then they kind of all, you know, they kind of like last night was uh they called it Red Bull and pizza night because uh the idea was to present your product idea today, and so a few of our people are heading out to see how those product ideas look.

But that’s just the culture, and it’s a culture of learning and sharing information. It means you don’t have any single one bet at Constellation; you’ve got everybody doing it and thinking about how we deploy our shareholders’ capital at high returns. Uh very important to us, and I’m so fortunate to be part of an environment that that does this, and I think the leaders at uh across Constellation have have brought that uh into uh into—so I think I’ve sort of said this in other words, but we believe decentralization, our structure is our biggest advantage because we’re close to those customers. Some of our customers we’ve had for decades, decades needs, and our people have worked for us for decades. This is the great thing about being part of a buy-and-hold forever organization, is you’re not going to get flipped again. You’re not going to have three business cards in the next 10 years; you’re going to be at that same company, and we preserve the brand of that organization.

So we really believe in long-term relationships that are trusted with our customers. And do we know the customer space? It isn’t a ChatGPT “go ask a question” type of a situation. Again, great product, it’s you got to know a lot in order to build products for customers. You’re dealing with trusted information. You heard about Santina, you’re dealing with patient data in cases like that. That’s valuable data, and it’s got to be managed very carefully in this environment. But our biggest advantage, if you’re going to be decentralized, you’ve got to believe in peer learning. Our best leaders are the best at learning from each other, and it doesn’t matter how many decades they’ve worked for us, they’re always open to a new idea and a new way to do something and listen to their customers.

So I really wanted to make sure you all understood this. It’s a question we’re asked about a lot, and it’s our structure is our biggest advantage. But in order to give you a little bit more color, we decided to invite a few of our our leaders who are running our businesses, who are the ones that are back in that all those little yellow square squares across the organization um and what they’re doing. So put a little bit of color of some real real examples. You heard a little bit from the vertical leaders, and uh I have to thank uh Jeff Bender for taking on this panel for for us, and uh I hope you enjoy meeting some of our leaders before we get into the general Q&A. So uh and uh Larry, if uh yeah you can thank you, Jeff. Uh Jeff run Harris so yeah.

Larry: Just a little improv here, uh as I always like to do at these annual meetings, I have for the past eight years anyway—

Panel: AI Transformation Journey

[00:41:04] Larry: Just a little improv here, uh as I always like to do at these annual meetings, I have for the past eight years anyway uh have a running tally of how many are in attendance. The room is filled up to to capacity um here in town, so about 450 sitting here live um and now nearly 800 uh joining us online as well. So thanks everybody, and I’ll give it over to Jeff.

[00:41:30] Jeff Bender: Thank you, Larry, for that uh that update, and and Mark. So as Mark mentioned, I’m Jeff Bender, i’m responsible for the Harris operating group. We’ve gone from software eating the world to a fear that AI is going to eat software. At CSI, we have a deep belief in our value creation and value capture framework. We live this framework every day by getting as close to our customers as possible. When we do this, we develop a depth of knowledge, understanding, and expertise that we then use to solve their most important problems. At CSI, we see AI as an opportunity—an opportunity to rethink and reimagine how we do what we do for our employees and for our customers. So as Mark mentioned, you’re going to have a chance now to hear from a number of our business leaders about their AI transformation journey. So let’s meet the panel.

[00:42:29] Greg Richards: Greg, I’m Greg Richards. I uh lead AssetWorks uh as Mark mentioned, we’re one of the older companies, we’ve actually had customers that signed up in the ’90s. Um we manage about 14 million assets globally um particularly fleets for state, federal, and local government, but increasingly now more large private fleets as well.

[00:42:54] Andrew Jones: Great, thank you. Uh I’m Andrew Jones, i’m the GM of ClickDimensions. We uh provide marketing automation, so we help our customers do outcome-based marketing through our Dynamics native platform, as well as uh providing marketing automation, we provide marketing expertise as well as services to help our customers um, you know, do great things.

[00:43:11] David Nland: David, yeah thanks Jeff. Um I uh sold my business to Constellation back in 2018 so I’ve been on the whole journey um I don’t work within that business anymore. Um I have the great fortune to look after a couple of businesses that we’ve bought in the past sort of three or four years. Um the first of those is Celcat, that’s uh spelled with a C, Celcat are based in the UK. Uh we timetable academic life for universities and we do that in the UK, uh we do it in France, South Africa, Australia, and we’ve been doing it for 40 years. And then the second business is Bullet Solutions, who we acquired in 2024, and they do very similar thing, they also timetable academic life and they serve Spanish and Portuguese-speaking territories.

[00:43:59] Jeff Bender: Let’s stay with you, David. So my understanding is you did something uh that most leaders would say is impossible. Tell us more about that.

[00:44:07] David Nland: Yeah, uh well I I uh experienced one of the Harris AI workshops uh back in March last year, and uh I went away from that with one question which was simply: how fast can we go? And uh what I did on day one was that I stopped all development across both businesses, Celcat and Bullet. We stopped for the whole month of April last year and they went on an int—intensive learning program to learn um what we could do with the technology. And I think if I look back and we look now at the outcome of that, it—it’s actually pretty straightforward.

Uh at the outset, we came back in May and we were going faster. If you think of a cycle time from concept uh through to production software um typically for us that was measured in months. Um when we came back in May, we’d already started measuring that in weeks, and as we went through last year, we turned that from weeks into days, and today our cycle time is measured in hours. Uh we can cycle from concept through to production uh in hours, and that means a roadmap that might have taken say a whole year—we plan out a whole year roadmap—that can be literally a week now. So this—the pace at which we produce software is no longer a problem. We have different challenges instead, but it’s fundamentally transformational in terms of what it means for our business.

[00:45:38] Jeff Bender: And how are your customers reacting to this when you talk to them?

[00:45:43] David Nland: Uh yeah yeah, well let me put that first in the context of of growth. Um so if I go back to the end of last year, our growth was running at about 12% per annum um today it’s running at 23% per annum. Um our vision is that we’ll double our revenues by 2029. Um I think I’ve also been told I’m sandbagging on that. Uh the uh the interesting thing there is that growth, it’s not driven by going faster. It’s driven by us asking a totally different question, and that is: if we can go 50 times faster, then what does that look like for our customers? So what does 50 times faster look like for our customers?

And for us, that meant we’ve literally rebuilt our product from scratch. We—we’ve in effect reimagined what a new category of product might be. If you think about our old product, it was a system of record, you know, you created, read, updated, deleted data. Um our new product is a system of action: it thinks, it predicts, it gives you insights um it does really clever stuff for you. So that creates a huge value for our customers, and and and with value comes new price points, doesn’t it? And with that, we we we are driving growth and we’re seeing the results of that already.

[00:47:00] Jeff Bender: Thank you. Uh Andrew, so most people start with product when they think of of what’s possible with with AI, that’s not where you chose to start. Why?

[00:47:08] Andrew Jones: Well, we well we did um so we went to one of the Constellation peer sessions uh back in May uh last year, so exactly a year ago, and was we were really educated about what the possibilities were with AI. So ClickDimensions, we’re in marketing, which has the opportunity as well as the the negative side of being one of the first areas that is affected by AI, it’s it’s really being aggressively attacked. But I think if you can get through it, then it it creates fantastic um companies. So we we wanted to look at our product in a very different way, you know, ClickDimensions is is a horizontal marketing automation platform and we wanted to create something which was vertical. I’m very much the marketing rather than the technology of marketing and technology, but I passionately believe that AI is a fantastic thing and I wanted to be heavily involved in it. My background is very much in business transformation, and I wanted to see if there was a way that we could incorporate AI to transform businesses.

You have a lot of chat agents out there, and a lot of those those chat agents, they are designed to identify problems but not resolve problems. Um we had an incumbent uh ticketing system that effectively did that, and and I challenged um myself and a handful of people to go away and see if we could find a way of improving that, actually change this chat mechanism to being a problem resolution system. So that was our challenge. Um the great thing about Constellation, as well as you see there’s a, you know, 1,500 companies, you have lots of people that you can have discussions with to actually find solutions. We spoke to a lot of very large technology companies uh we spoke to a number of contractors that we used, and we also spoke to a number of companies within Constellation, and we found a company in Constellation that actually had some experience of this who could actually help us drive this forward.

So we had conversations, we decided we we wanted to create a support agent that could resolve problems, but that was not where we started. Our first start point was uh creating a a very simplistic knowledge agent to be used by our support engineers to to deal with tier 2 tickets. So when they’re on the phone with a customer, they may have a question, rather than having to go away to a library to get information, they could put in a chat and get the information first of all, and that’s what we started with. Um what was key with that though is when you start to digest knowledge or your databases so it can be read by AI, it is read in a very very very different way than how a human is. It is, you know, it is binary, it’s ones and zeros. So therefore, you need to make sure that the agent can actually pick up those key words so it can actually answer the question properly. So we started off with the knowledge first of all.

And then secondly, we wanted to see if we could have a very fluid interaction, so we created up a a web agent as well, so it’s on our it’s on our website so you can go in there, you can ask questions, and those are the first two things that we did. And then we effectively consolidated together to create the support agent. Um it took us about four weeks to do it after doing going through those various steps. We launched it in January of 2026 and what we found was that um with our customers’ interactions with this chat agent, conversational agent, you know, rather than being a lookup agent, it was a problem resolution agent. It effectively resolved 82% of all our tier one tickets that went through the chat agent, which is fantastic.

But what’s the most important thing is it allows us to generate greater customer intimacy because first of all, it means that our humans, rather than having to waste time dealing with tier one tickets, they can spend time focused on doing tier two, tier three three tickets, which are the most um uh important thing. But also secondly, the time that it took for a customer to actually get a problem resolved when it’s very simplistic as well, rather than it taking potentially, you know, four hours because they go into this room, they ask a question, they get a they get probably four or five different things that they could potentially look at which could potentially resolve resolve the problem, rather than that happening, they could speak to the chat uh agent straight away in any language, 24/7. So that is a massive massive massive gamechanger. So thats that’s what that’s the that’s our step that we did, and then since then we’ve gone on and created a lot more agents throughout the organization, looking at specific business problems or workflow problems that we’ve had, so it’s been a it’s been a really incredible journey.

[00:52:02] Jeff Bender: Now somebody told me that you uh name all of these agents. Can you give us some of the names and and explain to us like why like why did you go through that?

[00:52:12] Andrew Jones: Yes, we do, we do, and it’s it’s not, you know, kind of arrogance on my behalf. You know, when you when you think about an agent and what an agent can do, it’s very important you get a yard stick, yard stick of that possibility. A human, a work person, they have a role, they have responsibilities, they have areas of responsibilities, and they have um tasks that they need to do. So a human may be doing 12, 14, 16 different tasks. You cannot have an agent to replicate a human. An agent can take away one of those tasks. So if you want to have a an agent to be able to take all these um to these roles, you would need a lot of them, and rather than having support agent number one and support agent number two and support agent number three, we decided to name them. So that was the first thing, was just kind of the number of agents that you would need to have set up.

But also we wanted to in we wanted to get um adoption of the agents. Um in in my background, I’ve done offshoring um set up, you know, uh technology organizations in India, finance organizations in India, a video conferencing uh business in Malaysia. And when that happens, there’s always an element of question from the employees about stuff that has been offshored. So I wanted to make sure that that there was that accountability set up there with with people being able to to be able to see it and and and adopt the agent. So yes, we gave all the agents names, they all have a Teams account. So rather than you having to go to some web page or try to find it, you Teams them, you can you communicate with them exactly as you would with anyone else within ClickDimensions, and I think that’s the most important thing.

The other thing is as well is by giving the agents names, you also have a human who’s a manager of them because again, this is really vital when you set up agencies. If you set up an agent and you leave it to its own devices, it will go off and over a period of time, it will it will go on a tangent. You need to have a human manager who manages those agents all the time. So by giving them a name and managing it, that’s that’s the way we we we moved forward with it. And again, also they’re in our in our organizational chart as well, so we’ve got absolute transparency about where we have the agents and that’s really driven things forward.

Where we’ve now evolved is we’ve set up agents—it was a, you know, we just wanted to experiment set up agents. Where we’re now going is we’re taking the next step, which is creating an AI-aligned department, which means that rather than having humans with a sprinkling of agents to try to improve certain, um you know, tasks and responsibilities with it, we’ve actually taken a very different way. We’ve taken a step back and we said, “If we did this from a Agentic point of view, how would we set up as an agents, and where would we sprinkle in the humans?” And the whole objective there is to get the efficiencies, to give the the greatest experience to our customers and allowing our humans to spend the most time of our customers as well, to understand the problems, to move things forward. So that was that was the reason why, but it’s it’s it’s it’s working as—

[00:55:15] Jeff Bender: It’s a a great example of I think this concept of, you know, rethinking and reimagining what you do, not just incrementally uh improving it. Uh Greg, not not to leave utility uh uh the end again, uh someone told me that you you like to break things. So—

[00:55:28] Greg Richards: You’ve spoken to my wife. Well, we don’t break things indiscriminately, right? But you know, so as been mentioned before, um Constellation gives us this enormous opportunity to network with our peers, learn new things. I was at a seminar about 18 months ago, and you know, we we could see what was going on with AI, and and you know, obviously the concern is that there are three kids in a garage who are going to come disrupt our whole business, and my thought was, “Well, why don’t we go disrupt our whole business?” So uh you know, got some people together and I said, “Hey, we’re going to buy all the tools, we’re going to do AI, go forth and and innovate.” And I think this is the part where I’m supposed to tell you that I’m brilliant, and uh we delivered a ton of innovation. I’m not, and we didn’t.

But the great thing about being part of Constellation is this enormous resource body of of peers and mentors that you can talk to, and and one of the AI mentors in the organization said, “Greg, you got to break something, right?” Because it turns out for some subset of employees, AI is kind of an existential crisis. If you built your life around writing really elegant code, the the transition to managing agents to write the code is is a hard process, right? Um so as a leader, I said, “Okay, look, I’m going to go find three people in the organization who are early adopters, i’m going to pull them out of their day jobs, and I’m going to give them a really hard business problem.” And in this case, we really needed to modernize our UI, and I said, “Look, the current estimate using traditional processes is this will take about 18 months and probably require 10 people. I’m giving you three people, and by the way, our user conference is in three months.” And they did it. It worked really well.

Interesting, so as this was going on, we started to think: if if we can make our processes uh this much more efficient, what can we do with our customers? So we started having—and again, it’s been mentioned before, this deep domain expertise, right? Really understanding what our customers do and how they use our products, right? So we we spent a lot of time with customers saying, “What are the things that you do that are that are really difficult, to take a lot a long amount of time?” And I was meeting with one customer, and they actually set a stopwatch and they said, “Let us take you through the process for how we acquire, manage, and retire assets, right?” And then they went through all the screens, and they—and it took about eight minutes, uh and they said, “Now Greg, understand we do this hundreds and hundreds of times a day.”

So what we kept hearing from our customers was: take these really complex processes where I print a report, and I go to a whiteboard, and I bring a bunch of people into meetings, and then I go back to the system, and then I get another report which may take hours, weeks, months, right? Just automate that process, let me validate at the end what that looks like. Uh and another great thing about what CSI’s done is we had an opportunity to send one of our employees to an AI sabbatical, which is essentially where they got to go visit with other companies that frankly were doing a better job than we were, right? Uh so one of my employees went out, did the sabbatical, came back just really energized, i was able to build a core team around him, and as Mark mentioned, we’ve been doing these AI accelerators. There’s actually one going on at the airport right now, there was one in Denver back in uh in February, and we were able to send a team to say, “Okay, look, here’s what our customers have asked for in terms of radical improvement of processes and and really substantial improvement in how we do our jobs and how we be more efficient um by the way, it’s now February, the user conference is still at the end of March, i need a working prototype by that event.”

Uh so we we had we had the team go to this event, start off cold. We we brought in developers, product managers, and even somebody from sales. Uh four days at this event, and they emerged with a working prototype, and we took it to our customers. And what was amazing is again, because we really understood how our customers use use our products and what their pain points were uh when we took it to the user conference, uh the reaction was overwhelming. Uh in fact, the product lead sort of had groupies at the user conference that that kind of followed him around, right? Asking more questions. Uh and and probably the biggest proof point is when we presented this as a product to one of our customers, uh it effectively doubled their investment with us.

[00:59:33] David Nland: I think we were talking about this last night, weren’t we? One of the things that we’ve done very well is that we’ve carved out focus time for our teams, and this isn’t stuff that you can do at the same time as the day job, and we’ve all experienced that. And actually sticking the guys in the broom cupboard for a few weeks and saying, “Here’s a project, get on and do it, and don’t worry about everything else.” That that’s been the key to making this happen at pace. And what was your customer reaction—i just want to finish up—your the customer reaction when you when you doubled the amount that they spend with us per year?

[00:59:59] Greg Richards: You know, I I’m not sure I’m supposed to admit this in this meeting, but they agreed a little too fast, i think we underpriced it.

[01:00:08] Jeff Bender: I think you did for sure, yeah, I think you did. Now, so again, I just want to for the uh for those in the in the audience, this, you know, Mark’s talked about it, uh I think actually the verticalization panel talked about it, we’re talking about it. You know, Mark says that we’re different and we do things differently, when this peer learning uh and peer sharing and this customer intimacy, you know, sitting with a customer and being understand what this process is and how they do it, they only invite us in because of the value that we bring and the relationship that spans decades that we’ve built with them, right? The trust that we have with them. And these are, you can call them intangible, but they’re very strong um determinants, I think, of where we’re going to be able to continue to move with our solutions, in this case leveraging AI, but just generally in terms of how we interact with our with our customers.

So I wanted to—the last piece I wanted to sort of talk about was it’s called the human change element, right? The impact element. You know, AI is more about people change than it is just about technology. And and, you know, you you know, you talk about taking stopping people doing their day jobs, you know, picking people out, locking them in a closet, you know, setting them to all this this training. You know, as business unit leaders, like, how are you dealing with this reality of this of this probably one of the biggest changes that we’ve actually seen, you know, one might say ever?

[01:01:44] David Nland: Yeah, well, i think it’s architectural from the business perspective, you know, the problem of pace of software development isn’t the challenge anymore. Um for us, if we if we look at go-to-market, then um think about this: we we have 300 customers across the globe that use our software. Um we have a deep level of trust, a warm uh relationship with these customers. We can we don’t have to go and find them, we don’t have to persuade them, we can go and have a meeting with them anytime we want. Um and from a go-to-market perspective, that that’s the ultimate leverage uh that we have that our competitors don’t. We can walk into the room and have the conversation.

So our challenge today, and what I’m currently working on the team with, is that if if we can transform our customers at this uh to this degree and at this pace, how do we do that at scale? Um we’re not used to doing that. Um we’ve had to rebuild our sales team, so our sales team is over the past 12 months been rebuilt. Um we’re doing the same with our services team, we’re rebuilding that. But I’ve also given them the challenge um they currently take about six months to take a customer from order to go-live. Um the challenge they have today that they’re working on, we have a plan, is to reduce that, not from six months to six weeks, but to reduce it to six days. If I can get them from six months down to six days, then we can go at scale um out to market with those relationships that we have today. Um and that’s huge leverage, isn’t it? Those um those warm relationships that decades of conversations in many cases.

[01:03:21] Jeff Bender: No, for sure. Andrew, what’s your thought?

[01:03:25] Andrew Jones: Yeah, I I think that um from a from a slightly different angle, I mean AI technology, yes, it’s it’s it’s it has a a massive accelerator on that. I think when you’re talking about the the workplace, the culture, the design, the evolution of how it’s going to be, I very simplistically think that AI will take away and do all the boring stuff. It’s going to be all the mundane business processes that are absolutely necessary but are probably parts of the jobs that most people don’t really want to do. So what you’ll do is you will enable your staff to be able to evolve and develop and become, rather than a tier one support person, they become a tier two or tier three. So I think that’s definitely a an evolution that will take place.

But I think also, the way I see AI as well is it’s a great magnifier: if you know stuff, you will become better; if you don’t know stuff, you will be found out very quickly. And I think that’s the whole Constellation advantage as well, is we do have that customer in—intimacy, we do understand our customers and therefore, we can give them advice about how they could potentially improve as customers because we have deep knowledge about the vertical industries that we’re in. And therefore, I feel that our ability to double down on that, you know, generate more salespeople, you know, be more focused on the relationship side as well, is going to allow us to evolve. So AI will take away a lot of the mundane stuff, there’ll be more interesting roles and more opportunities for us to create more salespeople to be able to have greater conversations of our customers, ergo grows. I think I think that’s that’s that will be how how we evolve. It’ be it’ be a it’ be a a much more interesting environment to be in, maybe we go to a 4-day week, i don’t know, but that’s how—

[01:05:20] Jeff Bender: Greg actually had an employee ask that, he said, “Look, if I use AI and I’m twice as effective, can I work half days?” I said, “No.”

[01:05:31] Greg Richards: Um our answer is you can work any half any half day you want, whatever 12 hours suits you. Exactly right. Uh you know, so yeah, first and foremost, this is a change management problem, right? And and I think we as leaders need to understand that this generates some fear and trepidation in employees, um and and candidly, probably not everybody’s going to make the journey. I’ve had employees say, you know, “Hey, this isn’t for me, this isn’t what I want to do.” And and we can be empathetic and we can respect that, but but I think we we also have to be relentless um because it’s coming. I think one of the real advantages of Constellation is just all of these training opportunities and peer networking opportunities we’ve talked about. So my goal is to send as many people as they can to those so they can build the enthusiasm um and and learn new ways of doing business.

And I think we’re there with R&D, the question is how does it affect sales, what do you do in finance, you know, how is marketing different, right? Same kind of things happening at all these different events, um and and and what I keep telling my employees—and I think this is key to the change management piece because in in the back of a lot of people’s minds, there’s a belief, “I’m going to train the AI to replace me.”—and I’m very clear with my staff: look, this is not an economics problem, at least not yet. Um this is about delivering innovation to our customers faster. We don’t have a profitability problem, we’re a profitable, successful company. Um this is not about cost cutting, but we’ve got to be able to innovate faster. And has been, you know, this is the sort of a mantra with AI right now: um your job is not going to be replaced by AI, but you may be replaced by somebody that uses AI better than you do.

[01:07:02] Andrew Jones: Yeah, I’ve said the same thing.

[01:07:04] Jeff Bender: Yeah, for sure, yeah. Well, thank you very much uh I think Mark and I were talking uh a little while ago, and I think, you know, I think we’ve always had a tremendous sharing uh culture across Constellation. We all run uh events, we all invite each other to our various events uh but when it comes to AI, the the the quantum leap that we’ve now taken in terms of the the quantum u of sharing, not just inviting each other to each other’s events, but actually sharing learnings and discoveries from business to business to business is is at a level that uh that I’ve not seen before, which I I think is absolutely, you know, fantastic. It’s been awesome, has it?

[01:07:44] Greg Richards: And exciting, it really it really is, it really is.

[01:07:46] Jeff Bender: So what you’ve heard from our panel reinforces uh our belief in customer intimacy. Uh and this is really—I know I I I so I talked about our belief in this value creation and value capture concept—this is what it looks like in the hands of great leaders. So please join me in thanking uh Greg, Andrew, and uh David. Thank you, sir.

Shareholder Q&A Panel (Constellation Software Annual Meeting 2026 Transcript)

[01:45:25] Larry: May I please invite the panelists to just uh introduce themselves. Half of them were already on the stage this morning, but nevertheless it’d be nice to just uh by name and and uh enroll at Constellation. Mark, uh we could start with you, not that it’s not necessary at this point, but uh—

[01:45:39] Mark Miller: Mark Miller, oh yeah, president of Constellation Software. Yeah, thank you, Larry. And I just wanted to uh just uh quickly point out that uh we haven’t re—received any of the questions, the management team here, that uh a few Jamal got that were required required calculations because we didn’t want him to bring his calculator on stage, so he got a few of them. But we’re really going to make this uh, you know, dynamic and in what gets asked, so which is a tradition of Constellation. Thanks for the point, Mark.

[01:46:16] Jamal Baksh: Jamal Baksh, CFO of Constellation.

[01:46:18] Bernie Anzarouth: Bernie Anzarouth, CIO Constellation.

[01:46:20] John H. Beltz: John Beltz, board chair of Constellation and board member of Topicus.

[01:46:23] Jeff Bender: Jeff Bender, responsible for the Harris operating group.

[01:46:26] Mike Dufton: Mike Dufton, responsible for the Volaris group.

[01:46:29] Damian McKay: Damian McKay, responsible for the Vela group.

[01:46:32] Robin van Poelje: Robin van Poelje, chairman, CEO Topicus.com.

[01:46:37] David Nland: David Nland, Lumen Group.

[01:46:40] Bill Delaney: Bill Delaney, Medaxo.

[01:46:42] Barry Symons: Barry Symons, Jonas Operating Group.

Question 1: AI Opportunities and Attrition Risks

[01:47:34] Will: Sure, yeah, thanks for the presentations, um they were great. Uh one of the big questions that we got asked was about AR AI opportunities and threats and examples. And I think we got a lot of examples of companies inside the Constellation group taking advantage of AI to accelerate things and deliver value to customers. There are a lot of questions about the converse. So Andrew from ClickDimensions, he mentioned that marketing software, for instance, is under attack early by AI-enabled competition. So we have a number of questions that ask: can you give other examples of CSI businesses that have experienced more attrition due to competition using AI?

[01:48:14] Mark Miller: We have, you know, we really like—I can toss over Jamal—but we really haven’t seen any AI-specific attrition uh you know, at any point. I mean, I think our businesses that experience high attrition already will probably be most under threat, um uh but luckily that isn’t a high percentage of our our revenues. I mean, Jamal, what what’s your what’s your thinking on that?

[01:48:38] Jamal Baksh: So yeah, I look at about 900 of the businesses, and I look at trending of organic growth and I follow up with the uh the operating group CFOs etc., if if if there’s anything that looks like an anomaly. And there are a couple things that might have been popped out, and both of them had nothing to do with AI, so so that’s my take. But yeah, does anybody else want to add? It’s a big topic. Jeff, maybe you want to talk a bit about it?

[01:49:01] Jeff Bender: Yeah, I I think the AI panel, like you said and Jamal said, I think there’s no specific evidence to date of of any uh significant or material attrition as it relates to to AI. But like I do believe it it’s coming, like I I think, you know, like we see competitors that are out there, you know, we hear of people, you know, doing things, so I I I believe it it’s coming, we just haven’t had it yet. You know, a lot of our verticals are, you know, fairly protected and and, you know, highly regulated or, you know, high compliance requirements. So I would say they’re, you know, they’re not always the fastest-moving verticals when it comes to making uh new decisions, and I think any AI-related solution causes them concern. Like it causes them to pause and understand exactly uh you know, what is it that they’re accepting into their—right, into their ecosystem of of systems. But I I do think, right, like it it it will start showing up, it just hasn’t—it hasn’t yet. And stay paranoid, right, like as always um—

[01:49:59] Will: Maybe one uh also on that panel um I know, you know, Greg touched upon how, you know, some customers were very excited to, you know, get their workflows automated and helped with with AI. But, you know, in any kind of vertical or anyu there’s going to be customers that are, you know, at the far end, at one end of the curve, and customers at the other other end of the curve, you know, maybe those that are resistant to change. And, you know, if you are kind of transforming your bus to um maybe optimize for those customers with the AI, how do you also manage to service the ones that are laggards, and how do you deal with that you know that kind of dynamic of customers in different stages?

[01:50:35] Mark Miller: Well, my—like from an AI perspective, again as I said as a developer, I I really because as we talked about having kind of relationships across across the the world across all of our customers, um I look particularly for those laggard customers, which are generally generally the the attrition rates are very low with the laggard customers. So even they’re very hard to capture as new customers if a competitor has one of those customers. So uh my hope would be we would use AI to expand our presence inside that customer by adding more functionality, which uh I would think as far as the uh as far as the, you know, the customers that are adopting it very quickly, i I don’t think we’ve seen a lot of that yet um other than hoping to add functionality.

Think for example, Andrew was up talking about ClickDimensions, that’s a in a marketing uh space, which is a bit more horizontal than our average businesses, and when you’re in a it’s a very high attrition business by nature. So when you’re in a business like that, moving fast, and your customers moving fast, you got to move fast, too. Um it’s still going to be it’s it’s just a tougher business to to run, and I appreciate that Andrew is all over it, but most of our businesses really aren’t like that, right? So uh Bernie, i mean like thoughts on that?

[01:51:55] Bernie Anzarouth: Uh yeah, that’s correct. We have some businesses that are inherently uh have high attrition, and so um we deal with it uh as appropriate. But the bulk of our businesses are are are quite low, and um the spectrum of customers that that want to adopt AI versus those that don’t, I mean, we’ll take care of them equally. Um it’s just a matter of uh serving them with uh what they what they feel they need to run their businesses more efficiently.

[01:52:21] Mark Miller: Changing user interface is a big deal for a lot of our customers. Like they don’t want you to change their workflows, their user interface uh in many cases, right? Which, you know, means you could come along with a neat snazzy new solution, but if it’s working—it’s a small percentage of their operating cost to to for what what we do for them—they just want to make sure it works, it’s secure, their data is protected, and it it gets done what it needs to do, which uh is what most of our businesses deal with, right? I mean, does anybody want any any to Mike?

[01:52:56] Mike Dufton: Yeah, I would say I think it’s really important to understand that when we’re out talking to customers, we’re not out talking to customers about AI per se, right? Like most of the engagement with the client is, you’re talking to the business users and you’re still focused on: what is that business problem we’re trying to solve? So if AI ultimately is embedded in the solution, that’s not the opening conversation, because the pace of technology, particularly as it relates to AI, is moving very quickly, and we’re often engaging with clients that are not necessarily even familiar with AI. So I think when we’re out dialoguing, we begin the dialogue with: what is the business challenge, how do we think we can make you more efficient? So it’s not uh it’s not led with, “Hey, we want to come and talk to you about AI.” It’s like, “We want to talk to you about a business challenge you’re facing, does this seem of interest to you?” And as the conversation progresses, we would certainly share with the customer where we think AI could benefit them in that overall solution, but it isn’t a, “We have this cool new AI thing we have, do you want to buy it?” It’s much more of a longer-term conversation based on how we’ve been working with that client for for years or decades.

[01:54:10] Mark Miller: Yes, if I think about businesses across the board, you lose customers for sort of four four ways: you lose them because they go to business, which happens in some industries; they get acquired by a larger customer, and so hopefully it’s our customer they kind of, you know, the big fish eat the small fish sometimes, right? Um you could lose them for price where someone comes and says, “Hey, we’ll we’ll we’ll do the same thing you do for for what a Constellation company does for half the price.” We don’t tend to lose them for that. I mean, you know, Andrew might see that in a high attrition, high churn business. Um where you really lose customers is when there’s some functionality that that you your competitor has that you don’t have, and um I think, and as Mike Mike said, it’s it’s not AI per se, it’s that ability to do something that really matters to that customer where they’re actually going to go through the the pain of switching, and uh it’s a shame when our businesses haven’t foreseen that happening and they haven’t gone and put, you know, put ourselves in a position where it’s an obvious decision to stay with us as their vendor. Uh and that’s uh that’s yeah, that’s that’s at least how I sort of see it.

Question 2: Disruption from Scratch and Procurement Cycles

[01:55:25] Larry: Another um question that arises from this morning’s presentation, but was anticipated by someone who submitted this question earlier, concerns um efforts to disrupt one’s own business. David reported on doing this this morning, and so i’d love to hear from u volunteers on the panel an answer to whether any CSI business units have tried to disrupt themselves or disrupt another business by using AI to start from scratch—all new code, but using their industry and business workflow knowledge. Anyone want to comment that? Maybe Robin, you want to take that one, I think uh—

[01:56:02] Robin van Poelje: How yeah, so so I think it’s it’s easy set, you know, “I’m going to disrupt my own business.” And what you have, but people working in the business, in the existing existing business, I think that requires some change as well. So you might organize it outside of it, but there are of course businesses where you already had for a longer period of time the idea that uh we should step up our game and we should do more, and those kind of businesses might be in yeah an ideal situation to really try to disrupt yourself. And so we we try to analyze in our portfolio uh our strong businesses uh and also our weaker businesses. We try to win with our excellent businesses, but there are also business where we think, “Hey, we could disrupt ourselves and really try to make a quantum leapu uh into the future.” That’s what we try to do, but again, that’s not all over the place. It’s easier said uh than done, but we try selectively to do that. Yeah, anybody got an example where they’ve tried to displace themselves entirely in a in a material business inside of your your uh operating groups?

[01:57:11] Mike Dufton: We’re working on one uh so we’re definitely doing it. We haven’t done it yet, we’ve got the product out there, it’s probably got maybe 15 to 20 clients so far um still a long way to go uh but we’re actively doing it. So we’ve, you know, used AI a lot to rebuild the product and are going after our ourselves, um and so it’s in process, but it’s early days.

[01:57:35] Mark Miller: I think Jeff really pointed it out well in the uh in his panel about what people don’t understand is the customer relationship is very very important. Um having a product and selling that product are two different things. Like having to get—you could have the best product, and I and I talked about like it’d be great if you had something that the competitors didn’t do, but you know, selling hasn’t changed. You still got to get out there, you’ve got to convince the whoever you’re selling to, especially a new logo, new customer, that you are a good opportunity. It’s it’s a good opportunity for them to to buy your product, and that hasn’t changed. In fact, in some industries—i mean, Bill, I remember was talking about an event he did in Hamburg uh said one of the you got to do a keynote at that, and you said one of your messages was what, Bill? Like—

[01:58:24] Bill Delaney: Yeah, so um they’re asking us about the the pace of of of change from a innovation perspective, and, you know, we looked at the research and it was clear that innovation cycles were compressing, right? Um you from you know months to to weeks and soon to be days. Uh and, you know, I looked around the room and I said, “But what’s happening with with procurement cycles?” Now, we we deal with large government procurements that are, you know, it takes multiple years for a procurement even to come to tender, then they run the tender process, you do a multi-year project, and then you have 10 years of maintenance. Those procurement cycles in government are getting even longer.

So we’ve got this dichotomy of uh extended procurement cycles compressing innovation cycles u and, you know, there’s elements of those tenders that really people should have a look at. You know, there’s elements around product capability and functionality, and you—but then most of the scoring goes to your experience, your ability to contract on the heavy duty terms that they’re putting forward, and they need somebody to take liability, right? Governments like to shift risk, so you have to have capacity to take liability. That might be through accepting LDs if we don’t perform, posting bonds or or guarantees. You’ve got to have a scale and a capacity to do that, and and these are so many other elements of what our customers at least look for, Mark, when when they’re buying things, i think.

[01:59:58] Mark Miller: So, and what you know, we’re very fortunate at Constellation because we’ve learned to, you know, deploy our shareholders’ capital at high returns, and we we’ve acquired a lot of logos over you know, the three decades of Constellation, and uh those aren’t easily done because I, you know, I’ve done that when we were in the early days of uh our our company was a startup, and we started in the late ’80s. Getting customers is so hard, you got to work really hard on it, you know. Uh I think a really interesting industry is telecom though with David, because there’s such such a rapidly—you’d say telecom changes rapidly, right David? And you, you know, you’ve got a what are what are those customers like uh with this particular question?

[02:00:35] David Nland: Well, microphone’s working right? Yeah. So uh that well, they’re inherently conservative. Um they’ve been through a lot of technology lifecycles, they’re bruised and they’re fatigued, and um their pace of change is subject to regulation um subject to lots of uh big big complex platforms that are highly integrated. So when we talk to those customers about AI, you know, it’s enabling math functions, like we do a lot with fraud and revenue insurance, and how can we apply new math functions to determine faster? We have a lot of a agents already um we use a lot of machine learning in a number of our products um we have agents. How can those agents become smarter? So if you look at the telecom stack from the top, we we integrate with customers their customers, in the bottom it’s networks, it’s like delivering TV, it’s delivering um signaling security in the network. So the closer you go to the network, the more conservative things are in terms of making change.

But if we talk to them, they want the total cost of ownership to come down. So they want the operations, administration, the configuration management, maintenance of integrations, and new feature developments, particularly with math functions or agents. That’s what they want, and we’re we’re laser-focused on solving those problems um because their platforms just don’t go through—it takes, you know, we heard from for months today is we’re trying to get them from three-year cycles down to one-year cycles, down to maybe six-month cycles because of the regression testing, because of the regulatory testing, the security testing, you know. So if we can get them down to six months, that would be a great great achievement using AI, and we kind of look forward to that.

[02:02:22] Mark Miller: Capturing new logos in your space is really hard, right Dave? I mean, like big clients—

[02:02:26] David Nland: We mainly get new logos through acquisitions, right? It’s very hard to go into a large media or telco business and become a new vendor. They they want less vendors, they want consolidation, and if they if there’s something risky in their environment, they’re going to want us to do it. They’re not going to want some new name company that they don’t know, so they’ll always give us the first chances. So if we can move quick, we should be able to capture that business.

[02:02:50] Bill Delaney: I just I just want to make uh the point that even though we have long procurement cycles, that does not give us the license to sit on our hands. So, you know, we are moving quickly to to use that time to to build out the value, increase the value that we provide. Yeah, okay.

Question 3: Subgroup Verticalization and Token Return on Investment

[02:03:07] Larry: Well, that’s sorry the panel, we’re uh having—I want to alert the audience that after we’ll ask the next question about this morning’s panel, we’d invite audience questions from this morning’s uh presentations too. So if you want to think about your question and um begin to line up, um we’ll take those questions after Will’s next question.

[02:03:27] Will: I have a quick one about the verticalization. How is uh CSI approaching subgroup verticalization when BUs in the same vertical are in different operating groups?

[02:03:38] Mark Miller: Well, we had uh yeah, i mean we’re we’re we’re discussing that on an ongoing basis and um you know, one of the reasons we we wanted to highlight verticals is in some cases we uh we think our verticals should work closer together. And, you know, but we’re doing evaluating that in a case-by-case basis and the whole team here is involved in that. So uh yeah, we’re going to continue to evolve our our verticals and get deeper where it makes sense, but ultimately our business units are the uh are what really drives our success. So uh we we we’ll cautiously think through that. So anyone else want to comment on that concept? He’s quiet. Jeff, take us—

[02:04:22] Jeff Bender: I think we I think we said enough on that for today, so um—

[02:04:27] Larry: We’d love to have a question um from the shareholders present um here today. And so why don’t you take it, the first one.

[02:04:35] Joseph Shapnik (Rainwater Equity): Sure, Joseph Shapnik from Rainwater Equity. Uh question on um the costs of investing in AI and how you guys are thinking about that. Obviously tokens are not cheap, and uh investment in in your your time, your team’s time and and all these efforts is expensive as well. How do you guys just think about getting returns on that over time? And then if appropriate, love to ask a question on M&A.

[02:05:04] Jeff Bender: It’s a great question. Jeff, do you want to take that or—yeah sure. I mean, obviously, you know, we’re measuring uh all of the investment that we’re making uh in AI. Obviously, uh you know, the token cost from the frontier models is is a large part of that cost, but we’re using all kinds of other AI tools um and changing them out on a regular basis as better ones come along. So I would say there there’s a lot more going on there than I would say we would typically see. So I think we’re monitoring it, we’re watching it right now. It’s not having any super, you know, negative margin impact on the on the overall businesses in terms of the way that or what we’re seeing. So I mean, I’m still concerned about what will happen with um the token costs. I do think, you know, that it’s if you see even what the vendors are currently doing, they’re constantly changing the plans or constantly, you know, offering you, you know, max plans that aren’t really max because right once you hit a certain limit, you got to either wait or or pay more. So I there’s going to be a lot of evolution there. So I think we just have to understand understand what it is, but but at the end of the day, you know, we can’t provide a solution to a customer that doesn’t have an ROI. So I think, you know, that that will be the the baseline of what will determine what we can or can’t do.

[02:06:11] Damian McKay: Damian, do you want to comment on it too? I mean, you had a meeting with a hyperscaler yesterday, I think. So you we’re measuring it, so you can’t move you can’t improve it if you can’t measure it. I don’t know if you can measure all of AI’s benefits at the moment. So we we understand there is a sort of an upskilling uh but we’re keeping a close eye on making sure that we uh looking at what impact it can bring for customers and how quick, you know, how quickly that can happen. Looking at the actual cost per business, how how much we’re spending, who’s spending what uh and then trying to uh look at the return from: is it helping us grow, how does it feed into organic growth and deliver, you know, those outcomes that customers need? Or if there’s productivity improvements, you know, how do we measure that as well?

So I think if we think about um, you know, at the at the rawest element to look at it, it could be, you know, net revenue per head. If if uh AI is uh producing, then we should be able to do more with the people we’ve got—uh more more value to customers and getting or winning more customers using AI uh that so thats a a base model. But we’re also uh evaluating a a few more metrics that we’re uh we’re going to use and and try to understand, but it’s uh it’s not all going to be captured at this point. It’s all—remember we measure because we’re decentralized, we measure everything at the business unit level, and we want to we’ll understand who is overspending on that. Which, you know, we’ve seen that happen with initiatives and organic growth before where invested a whole bunch of money building something new and returns in that were terrible. And and so we need to we’ll keep an eye on it and uh we’ll see who’s doing it well and who isn’t, and we’ll try to do our best to manage the costs using a little bit of central pressure on some of the large companies who are providing these tools when we can’t. But ultimately, the business units will own the uh own the investment, and we’re not sure how they do um—

[02:08:10] Larry: Thank you, Mark and Damian, for the question. Let’s uh turn to the middle aisle for a question number two, the first—

[02:08:16] Alex Captain (First Shareholder): Thank you. Alex Captain with the first shareholder I met yesterday when came in at the elevator. I wore my collared shirt for you, no tie, said no tie, so great great presentation today, and actually perfect follow-on uh to the previous question which was about AI costs. Uh and, you know, your response right now with respect to the metrics. Uh latest quarter, uh after you guys I think have had a lot of progress in some of the business units, as you talked about this morning, and driving some initiatives that are AI-based that uh generate some revenue, but the latest quarter is sort of consistent with the previous trend. What do you think uh is the timeline for actually starting to see at the aggregate CSI level metrics on the top line that are going to reflect the success that you have?

[02:09:03] Mark Miller: I love like I’d love to see more organic growth, i loved it to be yesterday independent of AI. But it’s really hard for me to predict that, Alex, like it’s really hard for us to to suggest when that’s going to happen. And uh uh we’re just um, you know, we haven’t seen a lot of loss of revenues from AI or any uh but I haven’t also seen a lot of new revenues from AI as well um so we’re just being um we’re just being uh we’re just trying to drive adoption of the usage of it and we’ll we’ll monitor it internally. But I’m uh I’m not signing up to uh a time when we can uh suggest there will be uh an impact on our revenues from a growth perspective. But believe me, the pressure will be on our business leaders to to do better and to continue to grow as much as they can uh at at at and obviously incrementally getting good returns on uh on the on the dollars we’re spending or euros or Swiss Francs on doing those investments.

[02:09:59] Alex Captain (First Shareholder): Maybe if I could ask just a quick follow-on on that. Um given that there were, for example, some costs related to tokens this quarter, that that was reflected in the maintenance uh cost line. Presumably there are some costs related to development. Are we covering our costs or more than covering our costs with respect to the AI investments right now in the P&L, putting aside what you’d expect going forward?

[02:10:22] Mark Miller: I haven’t looked at that.

[02:10:24] Jamal Baksh: Yeah, it wasn’t a material increase in that that COGS line that—

[02:10:29] Mark Miller: Yeah, I would believe that yeah, we’re probably outpacing revenue with cost today because I think we’re investing a lot, so yeah. But it’s not a material impact on margins, i think it’s a very minor impact, but yeah, I would believe it.

Question 7: Geographic Expansion and International Operations

[02:10:41] Larry: Thank you um Alex, and thanks Alex so much. Let’s take one more audience question in this segment before returning to Will. Sir—

[02:10:49] Ryan Floyd (Barka Capital): Thanks so much. My name’s Ryan Floyd from Barka Capital. Thanks everybody for putting on this nice event, it’s wonderful. We appreciate the candid transparency. My question is for Robin, nice to see you. Uh Topicus made an acquisition, or TSS did, in Indonesia uh which is really interesting, and they have operations elsewhere in East Asia. It would be great to hear about what that experience has been like, lessons learned from it. Uh it looks like uh public information, you’ve been looking at other opportunities in East Asia. It would be great to hear, to the extent you can speak about it, what the pipeline would look like uh but also just in general about what the lessons from that have been. Great question.

[02:11:34] Robin van Poelje: Yeah, so historically we were predominantly focused on Europe, but we slowly but surely move out of Europe as well. This is Indonesia as an example of it, and we we might disclose some more in in the near future. We do it selectively, we do it step-by-step. No other approach than in let’s say European or North American country countries um so that’s in general the philosophy. And what we learned, and that’s what we already knew within Constellation is, you know, if I have to fly from Amsterdam to to Helsinki, it’s probably three hours, or to Portugal it’s three hours, and to Jakarta it’s 14 hours. So it it consumes time, of course in the plane you can work and think, but that’s one thing, so you have to uh go over there.

And what we found out is that with certain things uh difference in culture, which we already knew because that’s in Europe the case as well, but there it’s even difference again. So the belief always have been in uh Constellation that vertical market software is vertical market software wherever you go. That’s still the case, but the way how you approach it, how you get best practices implemented, uh it differs and it also differs uh you know, the local management you have. I don’t think that’s always related to culture or the country uh sometimes uh you get lucky with an acquisition with great people and sometimes you don’t have that and you have to implement changes. And so in a sense, it’s not different than what we’ve been doing, it’s only some cultural aspects and traveling and all that kind of stuff. But we’re committed to to continue on that path, so it’s not just one thing, but of course we take the learnings into account, but I’ll expect us to do more there.

Question 5: Customer-Written Software Threats and Portfolio Tiering

[02:13:21] Larry: Thank you, Robin, and thank you very much, Ryan, appreciate that. Let’s return to the panel um pivot into operations. AI will obviously percolate some more, but Will has received some very interesting questions about operations.

[02:13:37] Will: Yeah, so this section is really specifically about um the risks and opportunities of AI and current operations. We’ve already talked about this a little bit, but one of the follow-ons to the question about AI-enabled competitors is AI-enabled customers. Have you seen any instances where your customers may have an appetite to write their own core system using AI tools or failing that, modules, right? Are they perhaps by le—do they perhaps have less appetite to buy modules from your companies because they can write some of these extensions or small things themselves? Have we seen any of that?

[02:14:11] Mark Miller: We’ve not seen any any yet. Not—we used to have uh and we still do have software that’s installed on like mainframes, AS400s, and uh we used to always have to place the code out there and some customers would have their own professional services people developing on that code uh because they thought it was cheaper than buying it from us. Uh but, you know, what happens is a few developers leave and uh they’ll uh they call you up and say, “Hey, you know, we’ve modified the uh the the software and we probably need some help making it work again.” So um I think we will see customers develop things and uh uh and they will some will have IT departments that do, and that’s that’s okay. And uh we we hope we’ll still be the the trusted partner to make sure that it works and isn’t uh, you know, messing up someone’s uh healthcare data that Santina has to worry about one of her businesses. So but we haven’t I haven’t heard anything.

[02:15:13] Will: We’ve also gotten a lot of questions, and we’ve heard a little bit from panelists um here and there, about maybe tiering the businesses, right? Thinking about strong businesses, weak businesses, especially as it perhaps as it pertains to AI. And so are these sorts of portfolio classification exercises and ratings typically at the BU level? Is it at the Constellation level, and can you talk a little bit more about specifically what you’re thinking about and looking for?

[02:15:42] Mark Miller: Well, at the Constellation—I you know, level—I’m just we’re sort of looking at, you know, how these tools are being adopted across the world. But like I honestly don’t really like to tier businesses, or, you know, some of our operating group leaders here might m do that in some ways. I think our weak businesses will continue to be weak, and we hope we get better leaders in them and make them stronger, and our strong businesses will continue to be—will continue to be strong. So I think, and that’s, you know, we measure that on a financial performance basis at the business unit level, and uh in the end, that’s what matters. It’s uh, you know, you can get good returns on the incremental capital you deploy uh on either organic growth or incremental acquisitions that you do in that particular area. So I’m not a big fan of ranking businesses from a an AI threat perspective, personally. I think the same the same applies independent of AI, whether business was strong or weak, I don’t know. But this crew might disagree with me on this, this is the wonderful thing about Constellation: we’re allowed to differ and and and do our own thing um—

[02:16:47] Bill Delaney: So at Medaxo, we have built a vulnerability assessment tool. Um not so much to tier them in terms of bad or good, but to really understand where they sit from a vulnerability perspective, and we’re using that to determine, you know, how we approach the investment going forward. So uh we have a lot of businesses that are, you know, fortresses, that doesn’t mean we won’t be spending money on AI, but it changes the way we think about uh exploiting that strength and doubling down on it, as opposed to some other businesses that may be more vulnerable, which may need a different type of investment. So we’re using it to think about—it’s not a cookie-cutter approach to how we approach the investment. So it’s a guide, uh and that’s the way we’ve adopted it.

[02:17:31] Mark Miller: It’s a qualitative assessment, right like—

[02:17:33] Bill Delaney: It’s qualitative.

[02:17:34] Mark Miller: Which, you know, I don’t love qualitative measurements. They’re like figure skating judgment, right? So, you know, I prefer organic growth and, you know, return on invested capital and numbers that you can really wrap your heads around because those you can trust, right? So your sport is swimming or track, you like to—some—

[02:17:51] Will: Yeah, i’m just—

[02:17:51] Mark Miller: Just saying, but but it, you know, I’m not saying it no anybody shouldn’t do that if they feel they do that. I think your your weak businesses will continue to be weak if they’re not led by the right leaders. So great um—

Question 6: Contract Monetization, API Access, and On-Premises Architecture

[02:18:05] Will: How are you adapting to AI agents becoming or potentially becoming part of your user base? Do you have requests, for instance, for MCP access to your software, and how do you potentially counter revenue loss from the advent of Agentic software use if it threatens sales of seats? Who wants to take that one? It’s a fun one.

[02:18:28] Barry Symons: I can start with the pricing side of things. So we’ve been talking a lot about how you deal with pricing because again, ideally you don’t want to price on seats in a lot of verticals. But you acquire a lot of businesses as we do, and you inherit the pricing models that you get and you have to be thinking through that. So we’ve been doing a lot of thinking about in this new world, if we start displacing seats using AI in our customers, then we have to have a pricing model to make sure that it’s more of an enterprise or value sold type of situation um so we’ve been spending a lot of time thinking about that and uh where it’s, you know, relevant in our businesses, implementing that. So but it’s something we’re very very cautious of because again, like you said, you could end up eating yourself from a cannibalization of seats perspective, so yeah.

[02:19:12] Mark Miller: Yeah, if your user-based pricing if—

[02:19:14] Barry Symons: Yeah, and we unfortunately—i know we don’t love it in Constellation—but we inherit it all the time.

[02:19:19] Mark Miller: Yeah, it’s hard to change those contracts.

[02:19:21] Barry Symons: Absolutely, you can’t sometimes because they’re three-year, four-year contracts and you can’t change them, right? But you got to think about it and be proactive and work on how you can change them. Damian, thoughts on that at all? Nodding your head, any thoughts or no need? They’ve got lots of questions, so—

[02:19:37] Jeff Bender: No, we uh so we just implemented a dashboard at at Harris, and that is one of the metrics that we’re trying to track. I would say when I look at it, I would say that was maybe one of the the areas that there there businesses are still struggling to to figure out. So we’ve asked them to basically let us know if other people are accessing our our system of record, how many API calls they’re making, who what organizations or or what vendors are making these calls, and whether they’re just reading or whether they’re writing back uh information. But I would say I don’t know that our businesses have always thought about it that way, and depending on what vertical you’re in, like you need to make that data available for for other vendors, like healthcare specifically, right? In, you know, a lot of cases, we have to make that data available. So I think we’re trying to understand what to do with that, so I think we’re in just in the early stages, but we’re we’re trying to to track it and to and to measure it.

[02:20:29] Will: Similar question. Other prominent software vendors have talked about monetizing API access to data or semantics housed in their systems. Is that something that you would consider?

[02:20:39] Mark Miller: We do that sometimes, some of our businesses do that, have done that for decades. But uh I I do love that idea. So where it’s possible, but it isn’t always possible. So uh I don’t know, would any comments on that? I I love the concept of it.

[02:20:53] Bill Delaney: So we work we work in complex ecosystems, right? We’ve been interacting with third-party systems for a long time, and wherever possible, we monetize it. Yeah.

[02:21:05] Will: Great, one last question uh specifically to David Nland. Are most of Lumen’s customers on-prem, and what does this mean for AI adoption rates? Good question.

[02:21:16] David Nland: Yeah, i mean, most of the customers are either on-prem or they have our own personal cloud um and a lot of our customers think or believe, and they actually are technology companies themselves, right? So um it doesn’t mean our systems are not cloud-native, so they need to be cloud-native for DevOps and for any innovation uh rhythm. Um so it’s gen—generally cloud-native architecture, but often deployed on bare metal because they own a lot of bare metal, or in their own private clouds. They’re very reluctant to do processing—core processing—in the public domain.

Now, if the World Cup’s on and we’re streaming video for Sky TV in the UK, for example, and let’s just say England make it through to the semi-finals, which they will, then we have to burst traffic above capacity, and that will be in the cloud. So so so using compute power wisely um when it makes sense is definitely what our customers do. But generally speaking, it’s it’s on-prem, i’d hold off making that order just yet, so yeah.

[02:22:24] Will: Do you think it uh affects your customers’ willingness to adopt AI?

[02:22:29] David Nland: Well, it’s just it’s going—you know, their their ability to adopt cloud-native. So se—94% of our revenue comes from tier one customers, right? So think about that profile. So um so it’s hard to get them from on-prem to cloud-native, even on-prem cloud-native or on-prem in the private cloud is that—that’s a big architecture shift, and we’re still in the, you know, 10 10-year-old journey of getting them up to cloud-native. So agentic AI native will take time, you know, in the tier ones that—that 8 6% of our customers that are not tier ones, they’ll get there much much quicker to agentic AI architecture.

But getting major platforms upgraded in tier ones to it, it’ll be a four five—they’ll want to see other people do it first. They’ll want to see stuff that’s secure, passes regulatory control, and where they understand this total cost of ownership, the tokens discussion. They want to understand total cost of ownership really well before they make any platform moves, but it will happen, but it’s going to be over the next 5 to 10 years.

[02:23:30] Mark Miller: It’s terabytes of data, right Dave? Like it’s uncomprehensible, the amount of data. There’s millions, if not billions, of real-time transactions, it’s crazy stuff. Great, thank you. That’s my section.

Question 4: Incentive Models and Reinvestment Structures

[02:23:44] Howard: All right um so I’ve got the next uh fun section of how CSI’s uh culture and operations and incentives interact with AI. Um so the first one I have is um a prominent US business person recently said on a podcast that the majority of SMBs don’t understand AI. So in in light of that, shouldn’t Constellation revise its performance incentive model to prioritize organic growth over acquired growth in order to spur, you know, subsidiaries to develop new AI offerings? And we’ve seen some examples of that of course, but I guess this questioner is asking, you know, should there be a little more priority focus on organic versus acquired growth depending on the the vertical or the—

[02:24:21] Mark Miller: This is a great question for John Beltz, uh it’s on the—you got to get him to say something.

[02:24:28] John H. Beltz: I was half asleep, it wasn’t—I heard operations.

[02:24:32] Mark Miller: Yeah, it’s true. If you don’t mind John, take that one, it’s a discuss—

[02:24:36] John H. Beltz: I’ll talk about it, i’ll talk about the uh a bit about the compensation plan um which I think everyone is familiar with. I mean, the the core of the plan, which everyone up here would be on, would be return on invested capital and then there’s a growth element to that plan. Um when you get down beneath these individuals—and they they can elaborate on this—there are various plans without con—within Constellation, and almost everyone running an actual business is only rewarded on organic growth. So if you’re running a business unit with 50 people, you have always been, and you probably always will be, uh compensated based on organic growth. It doesn’t really come into your thinking that much, the acquired growth for the most part.

Um however, that being said, there have been some changes made um over the last year to encourage more organic growth. It was implemented last year at a few of our businesses um as a trial, and then another large business group this year also put it in place, and effectively it’s a kicker for just organic growth, not only at the business unit level, but also at the portfolio leader level. Um so to answer the question: yes, there are changes being made on an experimental basis, and and people will wait and see if that has any impact.

[02:25:53] Howard: How have the experiments gone so far? Va—

[02:26:01] Mark Miller: Damian, what Damian, and then Robin. So how’s it go first, how’s it going?

[02:26:05] Damian McKay: The um organic growth, there’s a lot of opportunities and it’s really uh I think in particularly in the context of of uh you know, dynamic competition and and potentially AI, it’s how do we add more value and win, you know, being able to win more names. So getting the attention on that to be in a position, putting the incentives that will actually drive the focus on organic growth, and then having a a uh a strong focus within organic growth on new names or uh increased value, increased usage within existing customers. So fing—focusing that, I think we’ve had some some great traction. Some businesses and portfolios are just more uh a better position to grow organically. Sometimes we make investments that we know uh, you know, deliberately know that won’t grow, but we’ve got some really strong uh portfolios and businesses that are very uh tuned into organic rights.

[02:27:01] Mark Miller: Yeah, I always think organic growth, acquisitive growth, I like them both. So that that’s important, it’s always what this or that, and it depends. But but we like them both, and uh I think solid businesses also show organic growth.

[02:27:16] Robin van Poelje: Um so when we did the spinout together with Topicus, and they all historically had very strong organic growth, they had development capabilities and abilities, they developed lots of new products, new clients, so we didn’t want to—I’m not exaggerating—kill that with the typical TSS approach. So they are a standalone operating group company, they remain focused on organic growth, and we bake that into the incentive system. So we still have the CSI incentive system, but may solidly focused on organic growth. They do, by the way, acquisitions as well, but just to give you an example, and that’s where John was referring to, it’s something we implemented and we are evaluating and see what does it mean. So I think that’s great about Constellation as well, to do those experiments also in incentive structures.

[02:28:09] Howard: Great um and then maybe at the portfolio level um have uh any managers seen any, you know, be a portfolio shift under K—under KYC uh shift capital to organic incentives versus acquisitions due to a AI opportunities? So have you seen kind of slight redeployment outside of out of acquisitions and into invest reinvesting back in the business?

[02:28:30] Damian McKay: Yes, I think I imagine we all have, and it’s something that we we want to be doing initiatives and tracking those to make sure that uh, you know, they’ve got clear business cases. But uh when you when you change the incentives and you uh give a bigger carrot for organic growth, that’s going to drive that drives the focus on that, and then we have to as leaders, and the leaders within the autonomous business units, they need to be encouraging that sort of behavior, uh and the incentives help them as well. There’s nothing like personal wealth destruction to drive focus.

[02:29:08] John H. Beltz: A member of the HR committee, that was why KYC was put in in the first place, right? It was—

[02:29:13] Mark Miller: Yeah, it was a very—it was it was an amazing tool to change behavior, right? So uh yeah, it was very helpful, yeah.

[02:29:21] Howard: Another question uh on incentives um so conversely, this shareholder is asking: could your incentives discourage long-term shifts towards AI? So for example, you know, has the board or management looked at how a bonus formula based on ROWIC plus organic growth performs under structural organic pressure? So by that they mean, you know, uh could BU, you know, just look for one-time cost takeouts from AI, you know, as in replacing employees, but, you know, trying to juice ROI, but, you know, not really grow for the long term?

[02:29:52] Mark Miller: We we will obviously be monitoring that. If we see that happening, we’ll adjust accordingly. So uh yeah, either way, if they can, you know, increase their net revenues per person, that’s always a wonderful metric uh to use, whether that’s for growing the top line or uh, you know, uh figuring smarter ways to use the team. But yeah, so I think we’ll just monitor and see if we see any uh extraordinary uh things happening, but we haven’t seen that as as of yet.

I haven’t seen—we me—we measure uh we’ve measured closely on quarters, but we don’t think on quarters. So the long-term thinking, and if you look at the people uh uh that you’ve seen on stage, people have been around the company for a long time, the people that come in the acquis—with acquisitions stay for a long time uh and and that buy-and-hold mentality sort of uh, i think, is a an advantage for us where we don’t have the short-termism, we’re going to just um, you know, strip it out and try to make a quick a quick bonus and and move on. Uh we do think long term and and we take the ownership of the businesses very uh, you know, there’s a lot of pride in the returns of the business and the long-term returns uh as well. Yeah, the average tenure of the team up here is like in the high teens, right? Uh like, you know, it’s pretty amazing group of people and they’ve been on this incentive program for a long time and have seen a lot of different things happen in our businesses. We’ve had some really tremendous successes and we’ve also had some failures, and we’re just trying to learn from those, yeah.

[02:31:20] Mike Dufton: Yeah, I I think maybe one thing on organic growth that that’s important to understand across all of our business unit leaders: they want to be growing their business. So they’re not thinking, “Oh, I just want to maximize my return.” They want to grow their business. And I would say in the last year, as we’ve really leaned much more into AI, when you’re in these events, when you’re watching the leaders of these businesses see the way that they can pivot and respond to challenges they’ve faced with their customers for a long period of time, they get very enthusiastic and excited, and they want to double down and and try and find new ways to solve those problems. So we don’t have the proof points to say that this is converted into meaningful revenue, but I would say you can see the strong desire to—”I know I have a problem I’ve been wanting to fix for a long time, i now see a path to being able to do that.” So I do think that it’s important to recognize these leaders are trying to be more meaningful, they’re trying to do what they can to solve the the business problems, and I think, yes, compensation is one element of it for sure, but but I think there’s a lot of opportunity that they’re seeing to address customer challenges in a new way and and to increase the level of satisfaction um—

[02:32:35] Howard: And uh you know, on that, you know, we’ve heard some metrics maybe thrown out—revenue per head—and are there any kind of other, you know, metrics uh regarding AI disruption that you might feel comfortable sharing? You know, churn rates, win-loss metrics, pipeline, um and are there any KPIs you’ve added or changed to the operating ratios in in light of AI?

[02:32:57] Mark Miller: I don’t think so, no. I think it’s, you know, business as usual from a metric perspective, and be paranoid. What? Go ahead—

[02:33:04] Jamal Baksh: I agree, and and just on on net revenue per head, it’s really it should be it’s looked at at the business unit level because we’re as you saw in the video, there’s a a constant uh, you know, uh um very hard to track essentially new people coming in. But when we consider that, we’re really looking at the micro level and, you know, into each business unit.

[02:33:28] Larry: So before you ask your last question, I’d like to cue the audience that after Howard’s next question, we’ll return to the microphone. So if you think about a question and go to the mics in a minute, we’ll we’ll welcome your contributions. Yeah, Howard.

[02:33:38] Howard: Thanks, Larry. So last one here um really a lot of benefits of decentralization and I subscribe to that too. But, you know, one what this shareholder asks, um at what point does a lack of a centralized AI capability become a disadvantage? For instance, negotiating enterprise-wide partnerships with hyperscalers or large language model providers so you have that scale, and have you considered a hybrid approach, you know, because of the some economies of scale?

[02:34:03] Mark Miller: Like we will will—we are like ourselves, the couple of dozen people at Constellation will uh definitely speak to them about uh about, you know, I would guess you’d say stay sort of structures that we could use that to our advantage. We don’t like to do that, but when it makes sense, we’ll do that. We’ve all, you know, so uh for sure. I mean, anyone add to that, Jamal or Bernie, or—

[02:34:28] Jamal Baksh: We take advantage of our scale. So we do have relation—I’ve been talking to Microsoft, Amazon, or AWS for years, right? And now they are—we’ve entered into contracts with them so we can get better pricing, but we do not then force our business units to utilize it, they can access it if they need to. So we we have—I think we do have the best of both worlds: you can still take advantage of our our purchasing power.

[02:34:48] Mark Miller: So one of the real advantages is we can get the really good engineers out to hang out with our business units, and that’s hard to do if you’re a business outside of Constellation, it’s a standalone, you know, business that’s—and you think of our businesses like, you know, i mean maybe that didn’t come across in the pres—presentation today, but if you looked at the, you know, i won’t quote any numbers, the average business unit size, the median business unit side, these are small businesses, right? There’s dozens of employees in those businesses, and the ability for us to get a terrific engineer from one of these hyperscalers to sit down with the head of that business, the head of development for that business—and I saw it yesterday, i was sitting at a table with a group of people who were developing uh some product that was an agent, actually. And I get to tell—what you say, what business? I thought one of the actual uh people from the hyperscale is one of our employees, and I said hi to her and she said, “Who are you?” And uh it was kind of like—it was fun to see that because that’s where I think our scale helps us the most, is—and we appreciate all of their help um and we’re happy to pick up the phone and say, “Hey, we need someone to help us with this.” Where it’s harder to do if you’re, you know, $5 million business in Cincinnati, Ohio, and you’ve got, you know, some customers who want to move. So that’s where our our scale helps us a little bit.

But you got to make those calls uh, you know, as infrequent as possible because we’re pretty—we don’t have a big head office sitting around waiting to for uh, you know, to uh yeah, we got—it’s happening at the portfolio level, right, and at the operating group level. So we’re certainly trying to take the load for our businesses and give them some help and to, you know, so they don’t all have to learn how to ride the bike by themselves. Um interestingly, we’re starting to see it in M&A conversations where targets have realized that they can’t do this by themselves and they’re looking for the help from somebody like us and join join the family and get that type of help.

Question 8: Talent Redundancies and Software Refactoring

[02:36:44] Larry: Thank you, Bill. Um let’s turn now to question number four on that side of the room from another shareholder. Hey, how’s it going, welcome.

[02:36:51] Ashwin Anamalai (Shareholder): Good morning team. My name is Ashwin Anamalai, i’m from Waterloo. Um I’m a relatively new shareholder to Constellation. I usually only stick with index funds, i’m like a huge index person, but thanks to the great discount that Constellation is offering right now folks like me um we are able to be a shareholder. So it’s not just me, I asked all my friends to buy the shares here, everyone’s here, so thank you so much. Are you our first PEMS for us? So this is my first shareholder meeting here and u well, uh Mr. Miller, please don’t do any buybacks yet, no more trying to get added to the S&P 500, just give it a few more years and then we will get there. Well, stick with us uh please, like—

So what made me change my mind is that uh in Waterloo there was a hackathon called Unhackathon that was organized for university students where they were like, “Businesses, come with your problems, use AI and solve our problems.” It’s like, this is amazing. Um so with the help of AI, with cloud code, with Codex and all these tools, they should be able to solve all these problems. So, you know what, they did manage to solve the problems, but at the end of it all, zero—zero solutions were deployed for the businesses. It’s hard to sell, and uh that when it was the light bulb moment for me and I was like, yeah, building code is one part of the solution, but having the ownership and the trust is really important. The very next day, like started buying Constellation, you’re welcome.

[02:38:20] Mark Miller: What’s the failure rate on startups, right? It’s so frustrating, it’s so like—you know, you can build something great, i’m a developer, and you could build something that—and it’s really actually very useful, but success is not judged on your ability to develop products. Success is judged on your ability to go out and convince the customers that they want to use your product. And I honestly, like over the years, I’m sure all the team up here have felt it, there’s products that we’ve built that made a ton of sense but it took so long to penetrate uh that yeah, the returns on it are—it’s it’s a difficult thing to do. So uh thank you for that, like we didn’t actually uh ask for ask you to say that, but but it’s true, and and they’re doing a hackathon. Like that’s essentially what they’re doing up near the airport now. We’ve got our own hackathon, there’s 160 people there and they were up doing pizza and Red Bull last night or what have you, and they’re trying to do that too. But they’ve got to go convince their customers that they need this, and we’ve got, as Jeff said, a great relationship, and it’s still hard, right Jeff? Like for sure, like it’s still hard. So I don’t know if we answered your question, but thank you for the discuss.

[02:39:27] Ashwin Anamalai (Shareholder): So I still did not get to my question. Yeah, what is your question? So, you know, AI has improved productivity a lot and we are seeing uh large-scale tech layoffs. Are we going to see the same in Constellation? Thank you.

[02:39:41] Mark Miller: Who wants to take that? No one’s jumping at that one.

[02:39:45] Bill Delaney: I don’t I don’t see us doing large-scale layoffs, no. We’re at the moment we’re so focused on how to leverage this capacity because what you’ve got to understand is we just have an enormous backlog of our customers wanting us to do things for them. Some of those things that, you know, just didn’t make business sense now do make business sense. So, you know, I think as it was said earlier, maybe not everybody’s going to come for the ride with us, but you know, thinking about large-scale layoffs is is not—not on certainly our mind, i never discuss it with anybody here. By the way, comment—I don’t know if he’s—

[02:40:26] Barry Symons: Yeah, I was going to just add to what Bill was saying. Uh one of the things I say a lot within the Jonas organization—it’s a phrase that my team knows very well—and I do the hypothetical: if the business gods came down from above and said, “Barry, here’s the deal for you: you will never ever ever win another customer again in Jonas’ history, but at the same time, you will never ever ever lose a customer that you currently have in the Jonas organization. Would you take that deal?” And I say 150% I would take that deal, because there’s so much more we can do for our customers, and AI just makes it so much easier. And Mark talked a lot about it today with customer intimacy. If we get this right, the utopia grid is massive, and so we don’t need—I’m don’t get—I want both, i want new name customers, too. But, you know, the opportunity is huge, so that’s how I think about it, and you think about that you need more employees, you don’t need less employees.

[02:41:18] Mark Miller: It’s such a great answer, like I tried to convince one of our business leaders of that once. Like we bought this company in Switzerland and I was like, you never know need to get a new customer because they’re—they always blew their brains out implementing large new systems. And believe me, it depends on our business, so don’t take this across constantly. Oh absolutely, but but it was like honestly, like you should just build more things for your existing customers and that would be a great business and you—your customers would actually care more about you because you’re actually listening to them and solving them, rather than running to the next, you know, the next uh you know, building to like try to pitch something to someone new. And you put all your smart people sometimes on those new things. So that’s a really great point, Barry. Very inspiring. Um let’s flip back to u question number five from the audience, welcome.

[02:42:04] Philipe (Shareholder): I’m Philipe from São Paulo, my new capital. Uh my question is regarding tech debt, the legacy of your decade-old solutions, whether that makes it uh potentially much slower to build AI relative to AI natives building from the ground up from zero. Whether you’re seeing that in some of the organizations, or maybe if AI is actually offsetting the tech debt because you’re able to modernize, you can snap stuff on it.

[02:42:33] Mark Miller: Like obviously you can snap stuff onto our existing systems where it makes sense, right? So gives you an opportunity to do that uh and also, yeah, i mean, anybody want to comment on that? I I mean, I see our ability to understand exactly how the—the data is existing and snap things on top of it is advantageous, and we used to do that in the early days using Visual Basic or something, but now you can do that a little faster. So I think it’s uh it’s not something I’m super worried about, so for us. I mean, are you uh you guys worried about that or—

[02:43:07] Mike Dufton: I wouldn’t say worried about it. I think one of the interesting things is tech modernization is significantly easier using AI, right? So I think there are many people that could be looking at their existing tech stack and saying, “I can replace wherever I feel I have weaknesses around the technology choices I’ve made in the past.” I think what we’ve certainly been talking to our businesses about is not to focus just on the tech stack modernization, but uh it’s very difficult to convince a client to move to a new version of a system if the only thing you’ve done is replace the underlying technology. So it really comes down to: how are you going to respond differently to meeting the needs of the customer as a result of your tech stack modernization? So I think uh there’s no question you can enhance your technology stack faster thanks to AI, but really again, it’s going to come down to: do you understand what the customer is trying to solve, what’s the pain point they’re trying to solve, and how are you embedding that into whatever enhancements you do to the tech stack modernization? So I I think you’ll find that if we had weaknesses in our tech stacks, we can address them faster, but we we need to have a meaningful reason for the customer still to want to upgrade, which is going to come down to delivering more value.

[02:44:26] Larry: Thanks, Philip. Um we do have a bit of time for one final audience question in this segment, as long as it’s on the brief side.

[02:44:32] Shareholder: Yeah, it should be quick. Thank you for taking my question. My question is more around morale, employee morale. Given the the dynamic of the SAS apocalypse as well as like the technological shift with AI and people worried about their jobs, have you noticed a change in morale, and how do you keep morale of employees high? That someone should take—

[02:44:54] Bill Delaney: Look, what what we’ve seen in a recent major event was, you know, a recognition that people have gone through, you know, all the emotions with this, and so have we, right? Uh it’s a—it’s but, you know, what I’ve certainly picked up as the—the technology has matured so much in in in in recent times, and people can see real value and it’s starting to shake out how we can really leverage it, uh I’m seeing a shift to excitement and uh you know, the level of energy that’s come that people are coming away from from our events and taking back to their their business is literally electric. And that’s not just me saying that, that’s the—the independent feedback we’re getting from our people. So I think there’s been like a huge shift in the last 6 months or so uh as uh and and that’s—and I think that’s happening at all our levels. You know, I think some of us are are lamenting that we can’t go back in time and and uh, you know, do some things over because we would love to have the access to the tools that we’re seeing today.

[02:45:57] Damian McKay: Yeah, I think we we haven’t uh we haven’t seen leaders leave or people, you know, you morale start to impact people and concerned. And, you know, people, you know, very little sort of talk about the share price. I think most of our uh our leaders understand the power of compounding and they’re in for the—the long-term journey. They’ve got autonomy on the business that they’re working on and they can see a pathway there. So although, you know, the out external noise or changes, you know, it’s it’s uh, you know, they’re not oblivious to it, but when you’re in control of what you’re doing and you know you can compound your space and you’ve got capital to to grow and you believe in the model, uh I would say, you know, morale is is very high, and we don’t have as demonstrated by probably engagement, harder to measure, but also, you know, losing losing leaders, we’re not uh we’re not seeing that.

[02:46:48] Mark Miller: Yeah, you’ve got like, you know, dozens of customer or hundreds of customers and dozens of fellow employees. It’s it’s it’s a it’s a little bubble you’re inside one of those businesses, and we hope to allow them to continue to be inside that bubble and just learn from each other. So I think, yeah, it’s not like it’s a part of this big massive organization and, you know, it’s it’s you’re inside your business unit and you got your customers and you got your teammates, and you’re trying to figure out how to what to do for those customers. Oh, it’s a it’s a great question, and thank you um helpful answers, thank you very much.

Question 9: Rewrite Equations and Automated Code Safety

[02:47:20] Larry: Just pivot a little bit to the segment I’m going to tee up, which is um some technical matters including um coding, and we’ll start with rewrites. Um Greg touched on the topic of rewrites—AI and rewrites—in the panel this morning. We had quite a few questions around that topic, including one that came in to the chat this morning um and uh so here’s one version of it: how is AI changing your approach to rewrites? Um does the framework for thinking about the cost of rewrites um change with AI, and in in what ways?

[02:47:55] Mark Miller: Rewrites still scare me to death in kind of AI, but I don’t know about how about all of you.

[02:48:01] Bill Delaney: I think if you’re thinking about it that way, you’re sort of thinking about the world in an old way of thinking about the world. So we would rather our businesses—i think one of the panelists this morning said—step back, engage with your customer about what they’re looking for, because this technology allows us to envisage a completely different way for our customers to engage uh with our with with the with the data and and and, you know, the the services that we would traditionally provide them. So, you know, I would be putting rewrites at the bottom of the list of priorities, personally.

[02:48:33] Mark Miller: Yeah, it’s like you’d rather do more for your customer than rewrite your existing system. So as a developer, I used to love to rewrite stuff because you liked how we thought you could do it better, but honestly, you’d rather add more value to your customer than uh rewrite your code. But I think some people will—i mean, as there many rewrites going on?

[02:48:51] Jeff Bender: I wouldn’t—you know, there’s probably more than there was before. Like the rewrite’s always been an ROI issue, so the—the fact was before there was just no ROI in rewriting the vast majority of our solutions. I think the ROI equation is now different with AI, so I think yeah, we’re willing—we’re probably more willing to look at different opportunities because the return uh which I would say maybe rarely ever existed before, now in some cases does. But we still need to prove it to ourselves, right? That’s the theory. We have a small one going on and, you know, we’re watching it and, you know, it’s progressing well, but, you know, getting to 85% is is doesn’t seem to be that hard. Getting to 100% and then, you know, to getting customers to actually adopt it is is a whole different uh—

[02:49:29] Mark Miller: Yeah, it’ll be done next quarter, right? And then next quarter and then next quarter, and multiple quarters later. Thank you.

[02:49:38] Larry: Good, second question in this segment from uh the shareholders that submitted questions um refers to a recent discussion among leading programmers about a sudden shift from autocomplete to Agentic coding that um they hadn’t expected it but now think it’s confirmed, and they want to know: is that feeling common across Constellation businesses? Who wants to take that one?

[02:50:08] Mike Dufton: I could if—you know, look, we’re seeing it, we’re experimenting with it uh the jury’s still out a little bit because we’re conservative and we want to see, but we’re we’re definitely I would say uh recalibrating our expectations about how this is going to develop.

[02:50:24] Mark Miller: So, you know, my sense like generally of a Constellation—again I talk about we’re very intimate with our we’re intimate with our customers, we’re close to our customers—is there’s an opportunity to do more specific things for fewer customers than, you know, you used to build build once and saw many. More of an opportunity to build uh individual things because of it. So I hope I hope it becomes true, because having someone sitting across from the customer and sort of saying, “Hey, uh what do you need?” and uh doing it, might only be true for that one customer. We probably would have steered away from that before, but again, that is yet to be proven, it’s all all theory right now. So—

[02:51:03] Larry: And then I guess this is a inversion of that, but it’s that okay, suppose AI enables writing a lot of code much more quickly, what about the quality of that code? Is there some danger?

[02:51:19] Mark Miller: Well, it’s unpredictable with AI, like, you know, I use I use it too. Like it will suddenly do something unexpected in the middle uh you know, when it does the next write of some update you’re doing, which I’m sure you’ve seen that if you’re using it from a just from a using it for text or writing and stuff like that. So but I assume that’ll get resolved, but you don’t know, like it suddenly will do something entirely surprising. I don’t know if you you’ve seen that, but I’ve seen that doing some pipe coding and stuff like unintended consequences to that one little change that you just made, which actually fundamentally might change how you do things, but I’m sure that’s something they’ll they’ll work on.

[02:51:54] Mike Dufton: Said I think I think a lot of the conversation around the the migration of of let’s say writing code and and how it evolves, at the end of the day, there’s always a human that is integrating and making the final decision as to whether we’re going to promote that code or not promote that code. I think it’s going to change the nature of our our senior developers uh being able to oversee a lot more uh capability than they would have been able to in the past, but I don’t think it’s our view that we’re going to create systems where we just start randomly writing code and it just goes into production. So I think it will enable uh our developers to be more efficient in terms of the number of lines of code they can oversee, but there is a human at the end of the chain.

[02:52:41] Larry: Thank you. Next question that you submitted in advance runs like this: i understand that Harris recently launched a developer program to upskill software engineers with AI tools and learning. And you touched on this before, Mark, but i’d love to just hear an elaboration, if you don’t mind, even though Constellation is decentralized. In what ways, if any, is the company mandating or evaluating AI usage, product and feature development, and AI product?

[02:53:08] Mark Miller: Just looking at usage of tools, and it it’s Mike Mike’s event, but we’re also have multiple other operating groups at it. I mean, uh we’re just sort of looking at how the tools are adopted and we’ll learn from the businesses. If there’s good examples of things that happen, we’ll just learn from them and share them with everybody uh around the table here and see if they can use them. So I don’t know if that answer your question, Larry, but yeah, satisfies me.

Question 10: Pricing Structures and Capital Allocation Drift

[02:53:33] Larry: May I turn it now over to uh Will to enlarge the discussion about general operations.

[02:53:38] Will: Sure, yeah, so these are a little bit less specific to AI, they’re about operations. Um one of them I wanted to do is just a follow-up to an earlier segment that was posed in the live stream chat, which was um we talked about the breakdown of seat-based versus other types of licensing. Do do you have a general can you give us a general idea for the breakdown of pricing models in the portfolio across the company?

[02:54:03] Mark Miller: No. Jamal, that’s your—

[02:54:06] Jamal Baksh: Answer is no. I mean, we do not break it down that way um like I know there there are like we’ve just said, there are some seat-based pricing out there, there is other pricing, but I also believe that we can adapt, right? I think, was it Barry said it, that if that becomes an issue and we start providing more agentic uh employees for that and seat-base goes away, then we’ll find a way to get value for that, right? I mean, that’s that’s something we’ve always done. But no, i do we don’t track anything in that way, so I I don’t have to break down okay.

[02:54:33] Mark Miller: What it comes down to, like generally in business, you’ve got to be doing something that your customer values, you’ve got to be doing it better than your competitors um in order to ex—you know, to deserve that value, and you’ve got to do that continuously. Think about that as you go through and how you price. I think Jamal said, will it will be your based on your ability to to add value to your customers, right? And you can’t take that ever for granted, any of our business leaders take that for granted well, you know, they’re they’re not going to succeed. I hope they all uh they all remember that. So uh their customers are what uh what keeps their businesses humming.

[02:55:08] Will: Great u we’ve gotten a number of questions from shareholders who are curious about Constellation Payments as a long-running initiative internally, but there’s a sense that it’s getting more traction or becoming perhaps more mandatory in some ways. The question is: what are the incentives for Constellation subsidiaries to use Constellation Payments as opposed to other payment processors, also curious about the IRR on the investment needed to create Constellation Payments?

[02:55:34] Mark Miller: That’s your probably should take that one, maybe.

[02:55:37] Barry Symons: I think so, happy to let someone else take. Yeah yeah yeah. So um we’ve had a couple false starts with Constellation Payments, and so we realized many years ago there was a huge opportunity in a lot of the verticals within Jonas to capitalize on payments. We were always in the payments game as an ISO and reselling someone else’s stuff, and saw that there was opportunity to get greater margin if you move up the food chain, and so we did that. We partnered with uh one firm originally, and that partnership didn’t end up working out well, it wasn’t the right gateway for us. And then, unfortunately, we did a second partnership and that was working great, but then that firm was actually sold to one of the big uh payments companies, and they discontinued or end-of-life’d that product.

So don’t want to get burned three times, it’s kind of like the three little pigs, you know, eventually you build the thing out of brick, and so we got our own gateway, and uh we’ve launched that gateway, we have a number of clients on that gateway, and it’s picking up some pretty good steam right now. So we feel pretty good about where we are. As far as going across Constellation, we picked the name Constellation Payments with the dream that we’d be able to sell to all our friends up here on stage, but we realize within Constellation that sometimes doesn’t always work out that well um and so we’re really focused within the Jonas organization. So the vast majority of the Constellation payment stuff is within Jonas, but we do have a couple clients from Harris and a couple clients from Vela, and we just did a joint venture with Volaris over in the UK.

So it’s gaining some traction within the group um which is good to see, and it’s upon us to prove that this is a great solution and better than the others, versus Mark Miller dictating to the uh other people on stage to make it happen. So we feel pretty good about the trajectory we’re on um there’s still a lot to do, but uh yeah, we feel pretty good about the trajectory we’re on. We’re growing um significantly in the payment side of things in terms of organic growth, and we have been for a number of years and forecast doing that for a number more years. So uh I think that covered all the questions, but if I miss something—

[02:57:45] Will: Oh yes, you’ll have to answer yeah um historical IRR perspective.

[02:57:49] Barry Symons: Yeah, that’s perspective is always good. Perspective is off the charts um historical probably wouldn’t be that good, i don’t think we should be disclosing those numbers, but it’s definitely I would say below threshold, so I’ll say that on it. But uh it is the right decision, i think, about where we’re going now from an AI perspective and going back to what I talked about earlier about utopia grids and doing everything for your customer. Payments is becoming more important, as is things like hardware embedded—proprietary hardware, that type of stuff—which I know Bill knows a lot about in his business. Um and so I I see it as also a strategic wedge, attrition buster as well, and we’ve seen that over the years that clients that use both our software and our payments are much stickier than the ones that just use our software um so I’m very excited about that as well.

Question 11: Enterprise Data Sovereignty, Distressed Credit, and Terminal Value Multiples

[02:58:35] Will: But great, it’s an interesting example of sort of a cross-group functionality or a layer. Um we heard from Santina talking about data sharing inside of that healthcare group, we heard from Bill about Medaxo building an AI layer over his various business units. Really curious about the extent to which data can be shared or not shared uh across internal business units, especially customer data, and then also would love to hear other examples of things that you can now do because you have bigger business units, many more businesses at the same time, bigger verticals.

[02:59:11] Mark Miller: You mean yeah, bigger verticals, bigger business units. Sir, i love big business units, i’d much prefer to have groups of business together, yeah. Who wants to take that one?

[02:59:21] Bill Delaney: I’ll I’ll take it um look, you know, we’re uh trusted custodians of our customers’ data and we treat that very seriously. Uh but we, you know, we do have access into and can see, and with their with their permission um, you know, utilize that more broadly than we are today. Uh so it it really comes down to, you know, today where we have multiple products and even multiple Medaxo businesses servicing the one customer, you know, those systems are still talking to each other via various means. This is a much— this is a much a different level of that type of of engagement.

And then, you know, we have to get them comfortable that if we’re taking that data and gaining intelligence from it, then we, you know, we have to be doing that on an anonymized basis and we have to—it has to be a quid pro quo that they get access to the—the knowledge gained from, you know, the broader customer set that we open this up to. Our hope is that as we add more customers um, you know, we get even more intelligence and it becomes, you know, a compounder and a flywheel uh that that creates u value, even greater value for our customers, and we believe we’re uh strongly placed to to leverage that. But, you know, we definitely, you know, frankly, unlike um, you know, what’s happening in some of the frontier models uh, you know, we’re not scraping people’s data without their permission, so we will honor, you know, we’ll be very careful about treating our customers’ data with uh the respect it deserves. Any other groups have an interesting, you know, case of a of a cross-BU kind of initiative that’s—

[03:01:04] Jeff Bender: Yeah, we had some stuff in in healthcare. So obviously, it depends on um our customer contracts, but, you know, typically we we we were able to aggregate a lot of data, anonymize it, and then actually be able to monetize it by selling it to, you know, pharmaceutical organizations and research organizations that were looking to do research into, you know, cardiovascular issues or or other issues, because again, we had all of the patient data that we were allowed to get access to. So so we’ve definitely had some some successes with that, but it’s tricky, right? Because you, you know, back to Bill’s point, you like, you know, do no harm with with that data that we’re we’re entrusted with is is is the first consideration.

[03:01:41] Larry: Um before Will asks his last question this segment, I want to alert the audience that we’ll be turning to you again, and I see someone has already lined up. But Will will please pose your your final query on this segment, then we’ll turn to the audience.

[03:01:52] Will: Right, so last one on operations is on cyber security um didn’t quite get away from the AI questions. AI appears to be expanding both the scale and sophistication of software security threats. How does CSI minimize cyber risk across its business units? How do you balance the protection that comes from decentralization against the possible benefits of centralized resources such as red teams or shared vulnerability tools?

[03:02:19] Jeff Bender: I think we we try and do both, right? So I think, you know, we benefit from decentralization because a lot of our systems are uh standalone or apart, so, you know, you can’t get at all of our systems in in one fell uh in one fell swoop. You know, we do offer uh guidance, tool support, in some cases we mandate uh certain tools uh, you know, CrowdStrike across, you know, across all of our uh our businesses. And then we still also encourage other businesses to to do their own thing. So if you were to talk to Santina, you know, she has a very specific healthcare focus because healthcare uh just gets attacked on a on a just a continuous uh basis. So, you know, she would take the the corporate support and then she’s taken it down a whole another a whole another level, like she has a cyber security uh um leader within our healthcare practice, not just at the at the Harris level. So i’d say we we try and combine both to get the best we can in terms of, you know, being as you know as safe as as safe as we can, but it never seems to be enough.

[03:03:22] Mike Dufton: Yeah, and I think we’re obviously always looking at ways that we can use AI to monitor security threats across the volume of businesses we have. So it is very much a trust-but-verify mindset as it comes to things like security. We want each individual business unit to own the importance of maintaining the security of their environment and their data, but we do have the benefit of having the knowledge across a broad range of individuals and some very talented security experts that can go in and can work to to validate that the businesses are doing what they need to be doing. It’s probably a good example actually where there’s CSI involvement, operating group involvement, group in portfolio involvement, and business involvement. I would say there’s—that’s not often the case, but in this case actually, because it’s so important, you you actually see that going through the whole the whole stack.

[03:04:13] Larry: Okay, excellent, thank you. Uh so let’s turn to—I guess it’s question seven from the audience. We’ll start on that side over there, hi.

Question 12: The Chapters Group model

[03:04:24] Shareholder (Vela Software Employee): Uh thank you so much. I just want to say it’s uh uh wonderful to be uh here in my first in-person shareholder uh meeting. I’m an employee at uh Vela Software, and uh I also own uh shares in in Lumen. So my my question’s kind of geared toward uh David Nland um so uh with regards to the WideOrbit uh acquisition, my understanding is that at the time um, you know, Lumen was valued around like 18 times uh EBITDA, and then WideOrbit was at a lower multiple of EBITDA, like 13 times um so when they did the transaction, they essentially you were using a higher multiple public platform to acquire um, you know, a lower multiple asset.

Um I guess my my question to to you and the other spinout uh companies is: there are companies like Chapters Group where they raise capital at the TopCo level—and that can be debt or equity—and essentially that’s at a cost below the acquisition yield, and then they funnel that to HoldCos that can buy assets at lower multiples. Um I guess my my question is: why hasn’t Lumen and the other spinouts leaned further into that model? Do you see that evolving that way, and and just how you think about um, you know, creating those value, meeting those hurdle rates through multiple uh arbitrage and those types of structures, like similar to Chapters Group? Um i hope that’s—

[03:06:03] David Nland: I feel I need GPT to help me answer that question. Need a phone—yeah, I know. So, obviously, we were valued at the time as a compounding acquirer, taking a long-term view of what that value creation would be relative to a standalone asset and how you would value a standalone asset, and it was a premium asset so we paid a premium price for it. But a significant piece was the rolling shareholder investment in Lumen, so it’s a very unique deal. Um would we do unique deals like that again? I think we possibly would if we found something extremely interesting that created compounding value for our shareholders. Um you know, at $54 a share, we probably should have done it um it wasn’t the right time for us to do that. And I think we, you know, but at some point in the future there might be, if we trade above intrinsic value, I think there might be an opportunity to use script in deal. But that’s not currently our investment thesis; we’re we’re compounding with cash and very traditional approach. But something surprising, interesting, and large happens, yeah, we would definitely consider it again in the future.

[03:07:14] Larry: Thank you, David. Um question—

Question 13: Changes in M&A Approach

[03:07:21] Hernando Silver (Rosemar Capital): Hernand Silver with Rosemar Capital. Um you guys recently changed something in your M&A approach, which is you included a section on AI and risks and potential benefits. And I’m sort of going to turn this question around to the operating group heads and say: if you had to re-underwrite sort of all of your BUs, what percent would you say the underwriting would change today with LLMs versus say four years ago?

[03:07:47] Mark Miller: i’d say wicked—I wouldn’t uh yeah. I mean like Bernie, it would be a good question for you, we were talking about the last—

[03:07:59] Hernand Silver (Rosemar Capital): I think he addressed the operating group manager intentionally.

[03:08:04] Mark Miller: Yeah, I tried to ask the question for you because I think it’s if you’re—if you bought all the companies.

[03:08:08] Jeff Bender: Mark, I’m happy to start. You want to start? Yeah, because this was done—i’m I’m laughing because I’m looking at the individual who did it, i won’t disclose his name, but he did it within his own portfolio, and he said he he developed a test on AI threats and opportunities on how he’s going to approach it for for new acquisitions. Then he went back and he said, “Okay, if I would have applied this lens to my own portfolio um what are they, you know, where would they fall into?” And he obviously he had a select few companies where he said he wouldn’t have invested in under the new environment. And when I dug in a few layers, it was really a case of: these were already poor companies, what Mark alluded to earlier. They had high attrition, they weren’t great businesses, they didn’t have good moats, at the end of the day they they’ve turned out to be bad investments—not bad, i mean they still have good returns, just not great returns. And so I—it was hard to—it was hard to figure out what was really AI um that was impacting those businesses, or were they really just poor businesses that we didn’t fix and improve the moats. So it’s a—it’s a difficult question to ask, and I think everyone up here will tell you that there’s probably a few like horizontal solutions that they wish they wouldn’t had bought, but in reality, those weren’t great companies anyways, um and I think that’s kind of the the discussion that you end up going down.

[03:09:27] Bernie Anzarouth: I think poor businesses remain poor businesses, and you’d like to sometimes have another shot at whether you bought them or not, right? So sure, any kind of point solution that has high attrition, AI or no AI, is going to have issues. So uh looking back, I don’t know, the 30 years where we’ve acquired these businesses—some of these businesses um sure, maybe we shouldn’t have, or maybe there there could have been ways to to fix them. Maybe now with AI tools, we can protect them better um all sorts of possibilities. But it’s it’s certainly those um troubled businesses that aren’t great um we still get decent returns because it’s all a matter of of of the pricing and how much we can fix these businesses. But I think with AI tools now, we could probably do something better for those businesses, they’re not um they’re not gone uh but then again, we can always try to go back and revisit: should we have done that? Happy to say that it’s probably a very small minority of the businesses across the board. Excellent, thank you. Uh anyone else?

Question 14: Executive Remuneration and Governance Alignment

[03:10:31] Larry: Thank you for that. Thank you, Dan um question eight over on the right.

[03:10:35] Danny Poland (Shareholder): My name is Danny Poland, i’m a shareholder from Pittsburgh. I believe in the circular it said that Mark Miller uh elected to uh forego his salary and bonus uh this year um so that seems like a great deal for the shareholders, thanks for that.

[03:10:53] Mark Miller: Yes, I just I just carried on what Mark Leonard did, you know, and I uh I really care a lot about this company, and I took this job to help this company and the shareholders in it. So uh it was—

[03:11:05] John H. Beltz: A tough negotiation. Mark wanted zero comp, and we wanted to pay him something, and yeah—

[03:11:12] Mark Miller: So it’s really just uh for how, you know, I really just want to try to help this company uh improve and continue to increase its intrinsic value over time.

[03:11:20] Danny Poland (Shareholder): Sounds like yeah, my question was just what the motivation was so—

[03:11:21] Mark Miller: Yeah, it’s really just in the best interest of the company, and I hope uh I hope uh it can help in some way. Thank you so much for that question, and switching over actually—thank you, thank you Mark. Thank you Mark.

[03:11:35] Larry: And it’s actually question 10 here in the middle, and I think it’ll be the last one before come back to the panel.

[03:11:40] Shareholder: I had a question for Mark Miller. Um how different Constellation 2.0 under Mark Miller would be as compared to Constellation 1.0 under Mark Leonard? And similar question on the same line: how do you think about retention of the people, keyman risk? So none of your experienced guys go out there and create a copycat.

[03:12:03] Mark Miller: You know, it’s I think it’s it’s really continue what we’ve done for the last three decades, and uh I uh I don’t think there’s any fundamental changes to what we’re doing. I think we’re trying to learn some new skills, for example uh which maybe were some old skills, but we’re refining them, as for example with PEMS. And uh I think uh uh you know, we’re we’ve got uh we’ve got to learn and uh continue to grow that part of our our uh if you want to call it muscle, because we’ve got to um we’ve got a lot of capital to deploy, and I hope uh we can continue to compound uh for our shareholders uh at uh you know, and uh we’ll have to be become better at some of those things than we were before, I think. And uh I’m very uh looking forward to that, so.

And I think AI is just a nice—I think I said that earlier—it’s just uh for me it’s helpful because it, you know, it shakes up some of our businesses and gets them thinking about customer-driven mentality, and we just got to make sure that we get a good return on that uh investment in in what we’re doing right now. So I don’t think much has much has really changed, and you can see the team up here is uh expanded and uh you know, with a couple of dozen people at head office, there’s uh really we depend on each of our business unit leaders and all the operating group leaders to continue to uh the ones who generate all the cash, and um we just are fortunate enough to be able to help uh deploy it um so honestly, I I don’t really think things are going to change much and uh focus on developing people.

Question 15: Underwriting Frameworks for Non-VMS and Style Drift Assets

[03:13:36] Larry: Thank you, that’s actually a nice segue. Uh we’ve completed the first um round and a half. Um we now like to move to M&A and uh I’ll turn it to Howard for the first segment that section.

[03:13:50] Howard: So yeah, so the other uh other key branch of Constellation of course, and uh you know, it’s not mey’s getting ready now um and of course uh the first uh pertinent section of course relates to AI and M&A. Sure um and uh so the first question here um is about terminal value. So, you know, there’s uh you know, lots of AI innovations and so there’s a worry maybe, you know, as you evaluate software businesses, maybe you should be concerned about terminal value. So how has your thinking on that evolved, and how have you kind of integrated that into your process of uh of evaluating companies?

[03:14:20] Mark Miller: Yeah, there’s been a fundamental shift in terms of how the uh investing public views publicly traded companies, and and everyone has subsequently taken a haircut, including um us. Um I I don’t I don’t know, i I I see the businesses within Constellation and the folks at the AI accelerator that we saw out uh near the airport yesterday, and I see a whole bunch of people really pumped about what they can produce for their customers. And if you go back to that earlier question, you know, “Are we going to be laying off people?” i see the tremendous amount of capacity that we have going forward available for our guys to develop more and more product for our customers. So they’re going out, and the—the general managers of the BUs are are going out to speak to their customers and trying to figure out what their pain points are.

And all of that backlog of stuff that they’ve always wanted to do, and but never had the chance to deliver because it was just so tough to do, they’re jumping in right now uh both feet, getting everybody involved in trying to figure out what to do for uh for their customers. So I don’t see that terminal value diminishing the way that people have seen that in the public markets, and I think there really is a disconnect. Now, maybe there are some businesses that are on the fringe that maybe it could be copied really quickly, very easily replaceable, i just don’t see that. And so to me fundamentally, I believe in the businesses that we’re running uh here, and uh just the shift in in this um this tooling that we have in our shed that we could just take advantage of and deliver everything that we need for our customers.

[03:15:59] Howard: Is it fair to say that then for the kind of broad M&A process across different operating groups, you haven’t really put in any new, you know, kicker for terminal value assessment or anything like that?

[03:16:12] Mark Miller: I haven’t, have you?

[03:16:13] Bernie Anzarouth: No.

[03:16:14] Mark Miller: i’d love to no no no none of that, no so yeah.

[03:16:20] Howard: Um and then on assessing, you know, AI businesses uh you know, this shareholder is kind of asking, you know, is have you built any framework for assessing acquisitions of AI-first businesses, and how is that maybe similar or different to previous iterations of technology—SAS, mobile, you know, all of that?

[03:16:42] Bernie Anzarouth: Yeah um so AI-first businesses. So um right now what we’re looking at in terms of acquisitions, when whenever uh a prospect comes into our sights and we’re we’re going through the motions, part of our diligence now is assessing their vulnerability to AI and um it it varies from one operating group to the to the next to see how um they um they assess it. But that definitely goes very much into our thinking to figure out whether or not there is a vulnerability.

But not only that, um we also look at a lot of these businesses that are really in their infancy stage in terms of using the AI tools, and because we’re gaining all of that experience of using this stuff internally within our businesses, we’re also looking at how we can apply this stuff to the new businesses that we’re acquiring to see if there’s any upside. So we’re taking both of those into account, yeah.

[03:17:35] Howard: The due diligence is giving you internal lessons as well, right?

[03:17:38] Bernie Anzarouth: Absolutely, and and what we love to see—now we have, you know, 1,500 plus business units, we love to see the results internally so we can use those lessons and apply them to uh future acquisitions, which is how we’ve done everything from the start. We always learn, and some of these businesses come in with their own um lessons that we can take from them as well, and fabulous people of course.

[03:17:59] Mark Miller: Yeah, we learned everything from uh the businesses we’ve looked at, Bernie, since ’95, right? Right, so AI is just another aspect—ect to that. And um, you know, kind of a offshoot of that is uh if any of these kind of AI businesses you’re evaluating, they’re, you know, using one particular frontier model, you know, how do you kind of underwrite is that a risk, is that, you know, opportunity, something to keep in mind of?

[03:18:24] Bernie Anzarouth: The code is transferable usually between the models, to be clear. Multiple platforms, um it’s just a matter of trying—code is code is code, exactly. So yeah, don’t know if anyone else has anything to add.

[03:18:37] Jeff Bender: I think, Howard, we’re not really looking at that many AI-first companies, let’s just be clear. Like, so I think we’re looking at a lot of companies that are saying they’re doing things with AI, to Bernie’s point, and we get in and we try and understand what they’re doing, comparing it to what we’re doing and making our assessments. But like I can’t remember the last time, you know, one of the ones we looked at was what what we would call an AI-first company, you remember? You have to remember who we are and and and how we value and and what we pay. So I think that’s that’s that’s not where we are, not yet.

[03:19:04] Bernie Anzarouth: Yeah, absolutely. I mean, just just underline that point, the bulk of the businesses that we see are not AI-first businesses. These are small businesses, they’re doing the right thing within their customer bases, and they’re they’re um doing the meat and potato stuff, uh but generally they’re not um on the AI bandwagon yet, or very few are, and the ones that are are just using the basics. So we’re not looking for AI-first businesses per se.

[03:19:29] Mark Miller: We’ve se—we’ve seen a few um but they’re usually they don’t have customers and they’re sort of out of money and valuations, so I wouldn’t you know, I don’t know. I don’t think there’s a huge we’re not seeing huge competitors in our segments where this AI-first company’s killing it. We are seeing people who’ve put some money in something, try to build something, and can’t get the distribution right, yeah.

[03:19:52] Jamal Baksh: You could probably point them to the SPAC market instead of—um they could change their name and add AI to it, easy to do, that’s easy.

[03:20:01] Howard: Uh so um maybe uh switching to hardware. So you mentioned last year that you studied Motorola Solutions and Hexagon, and, you know, do do the developments in AI make you more interested in hardware? And I know Mark is a big fan of hardware.

[03:20:14] Mark Miller: We’ve always—i mean like the Harris through Trapeze had always had a lot of opportunity to deal with hardware because we dealt with a lot of uh of um buses and trains and there’s a lot of in-vehicle devices, and uh I’ve always liked hardware um I always think it’s uh I think it’s it’s it’s, you know, you want to do as much as you can inside that vertical. You obviously want proprietary hardware, hardware that uh is special and unique to that niche. So I really I really like hardware a lot. Um does AI help us with it? You know, i mean, you know, when you’re doing hardware, it’s a harder business, you got to worry most a lot more about working capital, and uh generally the hardware has embedded software on it. So it might help you write some of the embedded software outside inside the hardware, but uh yeah, I’ve never had a problem with it um maybe we’ll do more of it, maybe we won’t. It’ll depend on our uh what becomes available to acquire. So uh but it’s uh perfectly fine for us to do it, so it’s more of opportunities.

[03:21:09] Howard: Yeah, i think so. It’s AI doesn’t help or doesn’t change that game, yeah. Uh and then maybe just the last one on the section for me is on horizontal businesses. So we’ve kind of talked about a bit about that um and, you know, there’s a belief that obviously VMS is more insulated than horizontal businesses, and does that, you know, change your framework for acquiring horizontal businesses given AI, or, you know, has it always been the same?

[03:21:33] Mark Miller: It depends on the situation. I mean, if you’re horizontal in a specific geography like, you know, that’s one thing. You know, so it depends on the on the situation and the price and uh what we really think the long-term value of that business will be. So I don’t think it’s affecting us at all in our decision-making. We’re going to look at that business uh on an individual basis and think about what the next decade or of that business might look like under multiple multiple scenarios, right?

[03:22:03] Bernie Anzarouth: Yeah, i I I can add that many of the horizontal businesses that we do buy have their own moats despite them being horizontal, and that’s what we do look for—something that’s defensible. If not, we’ll model it up accordingly, exactly.

[03:22:16] Mark Miller: Which you mean affects what you can pay for that business, obviously, if uh the moat is not comfortable or deep. Makes sense.

Question 14: Permanent Engaged Minority Shareholder Strategy

[03:22:23] Howard: All right, thanks guys, i’ll turn it to Larry.

[03:22:31] Larry: Okay, thanks. Um my segment is on PEMS. We heard a presentation about the topic earlier, which provided I think extremely valuable information. Nevertheless, we got a very large number of questions about this topic, and so we’d like to pose a few of them. Um we’d like to front that by emphasizing that we are very grateful for Mark Leonard’s continued involvement in this endeavor as an adviser, very um we’re also happy with your keen interest in this topic um but we alo—also want to stress the need to be particularly careful in this area around um proprietary information related with this uh strategy, as as well as the fact that um it can involve two public companies and concern about revealing non-public information. So with that front, u a few questions if I may. Mark, here’s a simple one, at least to to read. Uh ready: what edge does CSI bring to the PEMS strategy?

[03:23:30] Mark Miller: Well, i mean i mean we’re clearly—we have capital, a lot of people have capital, but we really understand vertical markets, vertical market software businesses, um I think reasonably well. We’ve had a fair amount of experience over the last three decades understanding vertical market software companies, so I think those two combined are are are are useful. And uh we also know some sometimes a little bit about uh the verticals that these companies are in, which I think can help. And uh um we’re also comfortable buying uh or being in investing in businesses that require some some help. Bernie, i mean, maybe you could elaborate on that.

[03:24:21] Bernie Anzarouth: Yeah, absolutely. I mean, our our deep expertise in software for—we’ve been at it for over 30 years, i think that uh that helps us a lot. I think we have a lot to um to share uh in terms of uh how we do things, and and so our expertise in that, plus with serial acquirers. Uh you don’t see very many serial acquirers in our M&A conferences. In the past, we’ve we’ve uh invited um heads of serial acquirers that are non-software as well, and you’ve been to a couple of those and uh there are immense um tremendous similarities that I think our experience uh could could help, and uh I think we have a lot to offer. We thought a lot about compensation, about, you know, uh investing uh incremental capital at uh intelligently and measuring those things, and we think those are useful for people who care about these businesses over the long run.

[03:25:14] Larry: Thank you, so I think it’s a um is that—yeah, that’s a good second question. In a PEMS structure, how does the underwriting process discount or account for the friction of having to persuade a legacy board or management team to adopt Constellation’s capital allocation discipline? Sure, want me to take that one? Yeah, you take that one.

[03:25:38] Jamal Baksh: So if you think of the way that we’ve done acquisitions um to date, uh we buy businesses lock, stock, and barrel, and it uh it takes uh one, two, sometimes three years to get them up to speed with the best practices that um uh that we use uh to run our businesses. And so if you think of a minority shareholder and how much influence you can have on a business that you do not own um 100%, it’ll obviously take a little while longer uh in order to influence management to do the right thing. And so what we believe is that we have to get them at a lower price than the businesses that we are actually acquiring at 100%. And so you will see, if we measure ourselves at the same IRR uh that we measure 100% ownership, uh we would have to get these businesses at a lower price.

[03:26:24] Mark Miller: Yeah, to be patient, patient costs time is not your friend. Okay uh—

[03:26:30] Larry: Many—this third question of this segment: many conglomerates with meaningful public equity holdings trade at a persistent discount to net asset value. As Constellation builds out its PEMS strategy, are you concerned the market may apply a similar discount to these holdings over time?

[03:27:01] Mark Miller: Yeah, I don’t think so. I mean, uh you could see uh funds that are out there that are um closed funds that invest in individual publicly traded businesses, and you can go invest in them. But yes, the closed funds themselves have a discount to that. I think that over time um if you look at uh how our performance will um um just per how we’ll perform over time, um I think uh I don’t think there would be a discount. And if there is, so be it um it’ll vary over time, just the same way our intrinsic value is one thing and the price is a different thing, and there’s nothing that we could do to predict that price and to change it. And I think uh conglomerate—it’s still vertical market software what we’re doing, so it’s not some totally different business.

[03:27:54] Larry: Um what what if the market fully reflects the value of the company and you no longer see the investment as attractive? Um would you depart or protect your reputation as a permanent owner? I mean, part of the PEMS strategy is permanent, I think um so, and being a permanent owner, you’ve emphasized it several times this morning: buy and hold forever. Um how would you think about that permanent—

[03:28:24] Mark Miller: Yeah yeah, i mean, selling the businesses um if the price uh runs away with us, that that just doesn’t make sense tax-wise. It just would be tax-inefficient, and I think what we would do is um have or or just see like-minded investors uh pile into the business that we’re in, and uh just keep holding on to it for the long haul. We’re not in the business of, you know, selling high-priced shares to people that are unaware of what’s going on within the business; we’re we’re literally in there permanently. Yeah, with better things to do with our time, yeah.

[03:29:04] Larry: Um I’ve got one or two more questions on PEMS, and then we’ll return to the audience. So if you want to think about uh questions and and lining up, now is a good time. So either the ultimate or penultimate question in this one is: how can you generate cash from public investments from PEMS? Are these—are there strategies with special dividends? Are you as you technically do not generate free cash flow from the public investments, which won’t help you grow free cash flow and reinvest the proceeds, how do you how do you think about that aspect of it? Sure.

[03:29:37] Bernie Anzarouth: What the way that we run our businesses is exactly that: we look at areas to invest in. So if it’s R&D, if it’s particular initiatives that we’re doing within our businesses, so that’s where uh some capital goes. Then there’s capital deployment in terms of acquisition, and so we’ll take a look at that and see uh what’s available for acquisitions, and so we would apply those lessons to PEMS as well. And when there’s excess capital within the operating groups, what do they do? They send it up to headquarters. And I think what we’ll try to do is influence um businesses that uh have excess capital to return it to shareholders if they can’t find anything else to uh that that gets a good return on invested capital the way that we measure it. Um they should be measuring it themselves as well, so any excess we would hope to uh to get it to be returned, but we would want businesses to continue to invest in what they’re doing as long as the returns are appropriate.

[03:30:38] Larry: That’s excellent, thanks. I I think I will ask one more if if you don’t mind, because it’s it I think it’s got a subtle learning in it. The questioner is is looking for clarity, I guess, in her understanding of the blueprint of the PEMS strategy. Here’s how they describe it, then they want to know if her assumption—if if their assumption is right: so the plan is to buy shares, see if a a company is okay with being taken over, but if not, try to steer decisions in a certain way and reap the benefits of the improved operations. Is is that a fair description, or are they missing something there?

[03:31:15] Bernie Anzarouth: If you think of the process of the way um companies are for sale, it it doesn’t work that way. You don’t knock on the door for in a very large business multi—we’re talking about multi-billion dollar businesses—you don’t knock on the door and say, “Are you for sale?” and and, you know, “Yes we are, okay here’s the check.” Um nearly always, and we’ve had the opportunity to to uh bid on companies, publicly traded companies that are for sale, they are hand-in-hand with typically an investment banker that takes them out on a roadshow um with uh a document and an investment—a confidential investment memo that that talks about their business, and they do an appropriate job of scouring the world for appropriate buyers. Once that happens, you have to pay a premium over market u to get that business.

We’re not in the business of pay paying up for these businesses in an auction, it just doesn’t make sense for us. We want to go to the places where they’re undervalued. So it’s not—that it’s really not so we’re looking for undervalued businesses that are out there, and would like to get in and try to influence management um and what we’re looking for is that the appropriate managers are in place, the incentives are in in in line with shareholder expectations, and that their capital allocation is sound. It’s it’s very simple really, want to align those businesses with shareholder um requirements.

Question 15: Public Target Arbitrage and Organic Versus Acquisitive Reinvestment

[03:32:40] Larry: Thank you for that clarity, that clarification. So let’s turn to the audience again. I think we’re up to question 10, this one will come from the middle aisle.

[03:32:51] Andrew RosenBlum (Bonsai Partners): Thank you. Uh Andrew RosenBlum from Bonsai Partners. My question is around the compounding engine that we have. I think it’s one of the most important things that we do, we understand where capital comes from and how to redeploy it. PEMS is different because it doesn’t return the capital back to us, and it doesn’t give us the capability to take it from potentially cash-flowing but low uh incremental ROIC opportunities in that business to put it somewhere else. Um so my question to you is: do you have to, because the compounding engine sticks in the business for PEMS that we are buying in the public market, is the bar just intrinsically much much higher that we we have to be completely confident that that business itself will be a compounding engine like we are? And therefore, we need to hold it to a different standard of business, not just a VMS at a cheap price.

[03:33:41] Mark Miller: I think Bernie, you sort of answered that, right?

[03:33:47] Bernie Anzarouth: Yeah, that’s that’s a that’s that’s a very good question uh so you have to remember how we got here, okay? So um well we had a slide up there earlier that said we had uh how much—$3 billion of dry powder? Yeah um we’re we’re trying to invest that capital, and it’s it’s a very tough slog. It’s it’s tough despite the number of um VMS businesses that are out there um it’s tough to use to c to redeploy all of our capital, and that’s why we came up with the idea of PEMS. So the idea of PEMS is to use some of the capital that we have to find undervalued opportunities out there. Once we find those undervalued opportunities, the idea is well actually trying to find those those opportunities is to figure out um the businesses that have the highest probability of being receptive to our influence. So that goes hand-in-hand with what we’re looking for uh folks that have been shareholders of these businesses are probably looking for some kind of catalyst for these businesses to to improve themselves. And like I said earlier, there are a number of ways that these businesses can invest their capital: and so it’s it’s R&D, it’s various initiatives within the businesses whether it’s geographic, product etc.; there’s buying back shares, there’s dividends, all sorts of different avenues. So we believe that over time, using a number of different uh ways to improve businesses um that we will reap the benefits of the improvements of those businesses, i hope that helps.

[03:35:16] Larry: Thank you, Andrew, thank you very much. Great uh question is it 13, lucky 13 for you?

[03:35:25] Joseph Shapnik (Rainwater Equity): Joseph Shapnik from Rainwater Equity. Question on the same topic: uh you may generate as much or even more than 20 billion in free cash flow over the next five years or so. Yes uh that’s a really really tall task. I wonder, because it seems as though uh buybacks are probably unlikely given that you haven’t done one at this level, dividends have culturally not been very high on the list, so just wonder outside of PEMS, do you have any other ideas or evolutions that you might introduce over the next five years to help with capital allocation?

[03:36:01] Mark Miller: Sure. Mark, do you want me to take that? Yeah, you can, okay um—

[03:36:07] Bernie Anzarouth: So uh what uh so PEMS was is one of them, obviously. Uh if you dialed back the clock, I don’t know, a couple of quarters, uh Mark Leonard also mentioned style drift. Uh I think we’re looking at other areas as well um um services, tech-enabled services, and other areas as well, so it’s really in its infancy stage and so we’re experimenting um with different places, areas where we’ve had a little bit of experience in um through our software—software uh investment, and so we are looking at other avenues. It’s not just going to be software, of course, the bulk of it is going to be software because that’s what we’re really good at and we’ve studied that over the last 30 years. So you can—you’re going to continue to see a lot of that going forward, but there uh there’s a possibility of of new stuff that’s coming up.

[03:36:53] Joseph Shapnik (Rainwater Equity): One follow-up on that, you can follow up, go ahead, thank you. Uh follow-up on that: do you think you’ll see more PEMS investments over the next three years or more style drift investments?

[03:37:05] Mark Miller: Positive answer: i cannot predict that. There is no—we were talking about it last night at dinner, actually, and uh you know, you can’t really predict that, right? So it’s just nice to have different tools in the shed that you can use based on the current situation, right? So yeah. Now, PEMS is still an experiment, was just starting out, and just like um a couple of other areas we’re poking at um still very early. Excellent, thank you.

Question 16: Horizontal Integration Challenges and Autonomous Execution Checklists

[03:37:31] Larry: Joseph, question 14 in the middle.

[03:37:34] Daniel Lee (WCM Investment Management): Hey, how’s it going, hi. Uh Daniel Lee from WCM Investment Management. Um thanks for hosting the meeting, good to be back since 2019 um thank you. Kind of a follow-up to the prior question on—I think I might be wrong—but I noticed there’s a business unit that’s focused on industrials um and so it’s kind of to that style drift point. I know there’s a lot of experiments happening, but could you help us get a sense of how you think about approaching kind of new industries that are outside the direct kind of vertical market software realm? Um do you approach it first with the same hurdles, like the IRR hurdles first, or is it more of a business model thing first? And how do you track those experiments to see whether you should kind of dial it in more or kind of pull away?

[03:38:18] Mark Miller: Well, every investment we track and have tracked forever, or so uh and I think in our decentralized structure, there’s some experimentation just that naturally happens and ideas that come out. And we uh, you know, depending on the operating group leader and who who, you know, which portfolio, and we sort of um will allow those experiments to happen, and then we’ll measure how that’s uh how that’s working out, and then consider whether that’s maybe that’s something we could do more of. But it kind of, I would say, it happens organically uh wouldn’t you say?

[03:38:52] Bernie Anzarouth: Yeah, and we’ll measure them the same way, same IRRs, everything’s the same, right? Um it just so happens that VMS is a really good industry to be in um uh so we we’ll just have to measure them accordingly with the same um level um as our own VMS acquisitions, for sure.

[03:39:10] Larry: Thank you, Dan, thank you very much. And we we’ll actually come back, so we’re going to pivot back to the panel now um and we’ll actually have some more questions about investments outside of VMS in in a moment. But first, we’ll turn the table to um Will with some questions about general M&A activity.

[03:39:28] Will: Right, so um VMS is a great industry, there does seem to be some sort of gradation inside of VMS. The question asks: the phrase mission-critical is used to describe just about every software business. Um when you are evaluating businesses to buy, how do you assess the criticality of the software to the end customer, and can you illustrate with some examples of businesses where you’ve walked away because the software didn’t meet the mission-critical criteria?

[03:39:52] Mark Miller: I can hand it over to the other guys, does anyone else want to take that? It’s a great question. Oh yeah, Barry.

[03:40:03] Barry Symons: Yeah yeah, happy um yeah so we always think about if the system goes down, does the customer go out of business, can they not run their business, right? And so if you think of it that way, that’s the ultimate mission-criticality, and then you sort of take that part of the continuum and you go to the other side of the continuum, which is if the software stops working, does the business just pull out a piece of paper and hand write things down and and keep going? And so that’s how we think about it, and then we have that judgment that goes into it, and most of what we look at is on this end of the continuum, which is the business really stops working and you’ve got to be able to fix it and all that kind of stuff.

Um but you know, on this end of the spectrum, we do have some businesses—and, you know, we talked about marketing this morning um that’s one of the businesses that probably pivots closer to this side of the spectrum than this side of the spectrum, and we have some of those in Jonas as well that are closer to this side. And when we go I guess closer to this side of the spectrum where it’s less mission-critical, we often do it in a vertical we’re already in, so it’s like an add-on product on top of a mission-critical system. Um so that’s how we think about it and uh you know, if we’re going to this end of the spectrum, we hopefully price appropriately, and we think about that in the price we’re paying and what we’re willing to buy the business for. And if we get to this side of the spectrum, then we know it’s probably inherently good business, and, you know, one of the key indicators is obviously attrition rates, right? So when you’re on this end of the spectrum you tend to have very low attrition rates, when you’re on this end of the spectrum tend to have higher much higher attrition rates. And um yeah, I met—i was talking at dinner last night to one of the Harris folks, and they have a couple businesses with 0% attrition, i’ve never seen that. So I’m just jealous of Jeff over there, i’d like to know what that is.

[03:41:48] Mike Dufton: Yeah, so so the middle of the spectrum, just to add to that, the middle of the spectrum is a say a departmental solution where we have the—the main uh system of record, but there’s a little department solution that could be mission-critical to a handful of people within that organization. So the business doesn’t die if the—if you pull the plug on that add-on uh but that’s mission-critical to that group um so then that’s kind of somewhere in the middle, I suppose.

[03:42:13] Will: Unasked but implied in this question is, you know, to what extent are you looking for mission-criticality uh specifically as a qualitative measurement, as as a qualitative criterion versus as a, you know, it happens to be mission-critical because you’ve looked at the attrition rates and those tend to be low, do you know what I mean?

[03:42:33] Barry Symons: Yeah, we we use a business quality checklist that we’ve used for I don’t know how long, Bernie, but and it’s in there. But there’s other aspects to it and so it’s not the only thing we look at, but it’s one of the key things we look at is that mission-criticality and the business quality checklist. And obviously, the higher you score on the business quality checklist, the higher we believe it’s a great business and we might want to stretch to make sure we acquire that business. And if it scores lower, then we hopefully price accordingly, but it’s not the only criteria on the checklist, but it’s definitely on the checklist and a big one. Great.

Question 17: Post-Acquisition Integration Timelines and Operational Spillover

[03:43:03] Will: Uh Robin, a couple questions for you. Slightly over a year into investing in a SEO, what lessons have you learned about minority investing that might inform your approach next time around? Can you elaborate on how a SEO has changed for the better following the advice of Topicus executives that have joined their supervisory board?

[03:43:23] Robin van Poelje: So it’s it’s less than a year, because I believe we’re a little bit over half a year in it, so that’s one um and uh it’s a great business uh it’s uh I mean, the founder, Adam Goro, built this business, it’s a kind of uh Topicus, and he built it out with his team, so that’s it’s a great business, we like it. And what we try to do now is uh as as Bernie and Mark Miller referred to, is to have discussions with the management team. So that’s the part engaged, and, you know, we show them our tools and the way we think and all that things, and it goes step-by-step. And u you know, and they’re trying to get their head around things we do and things they do, and you want to continue them doing what they do very well, and hopefully we can add some things to it that one plus one is more than two. I think that’s very important in this partnership, and, you know, there are also partnerships when it’s less than two, but you try to do that together with the company and and the management, and we’re uh yeah, we’re on that path. And we have three persons on the board and, you know, we have good dialogues and I think uh uh it’s slightly different than like like said before, buying a whole company. Uh but like Bernie said, you know, is the feeling that people are open to what we do and how we think, and I think uh we have great discussions there with management. But like I said, it’s half a year that we’re in it, so it’s a bit early to tell where we’re exactly standing.

[03:45:03] Will: The second question on this is um if, to the extent that they do M&A at ACO, how do you address potential overlap and targets? Do you have some sort of deal registration database the same way that, you know, the whole company sort of registers its deals and which ones are carved out to the different groups?

[03:45:20] Robin van Poelje: Yes, so they’re not part of uh the Constellation system, so they have their own uh they, you know, they did M&A and they can continue doing M&A. And then we have resolutions about if there are potential conflicts of interest, we have to solve that uh but they do what they do, so they’re not part of the Constellation let’s say uh database or whatever you want to call it uh they’re not part of that.

[03:45:41] Will: Okay um this one is a question about private credit uh particularly around software. So to what extent has the dislocation in private credit, particularly around software assets and portfolios, created new potential investment opportunities and ways to deploy capital? I suppose the questioner is asking if Constellation is open to investing in credit as opposed to equity.

[03:46:05] Bernie Anzarouth: Historical well, we know that the opportunity is out there. We know that there’s some distressed debt out there with respect to software investments um but it’s a completely different world from equity investment, and so right now we’re looking at things but we’re not diving in. We know that it’s a scary world um just because of the way that businesses restructure debt um sometimes it’s just difficult to uh hold on to a piece of paper and make sure that that piece of paper still exists at the end of the restructuring, uh or whether it’s worth that paper at the end of the restructuring. The other aspect of it is it’s a it’s a very short tenure uh tenure type of investment. So debt is typically what—five years out or three years out, and if it’s distressed um there’s going to re be a restructuring that that happens in short order. So we’re in the business of long-term investing, so permanent investing. So it’s a very difficult to shift ourselves from going to from from long-term investment into debt that has maybe two three years um and then you get, you know, all of your money back, more of your money, more money than you expect, some of it back anyway. It it would go back to, you know, we did 25 years ago of these these short-term equity investments, and I don’t think that really fits with our model, but never say never. Great.

Question 18: Corporate Venture Capital and Long-Term Capital Sufficiency

[03:47:40] Howard: We had a live question come in: “What is your view on the market’s consolidation rate and hurdle rate? In other words, will you run out of small, profitable companies to buy soon?” So, a question about your runway. You guys want to take this or—take it.

[03:48:01] Bernie Anzarouth: i I could, yeah. So our runway is still huge. I mean, going back to profitable, we don’t—I think we’ve said this various times—we don’t look at profitability, we look at the a business mission-criticality, attrition rates, u loyal customer base, employee base um all sorts of aspects of these businesses um and our database of these software companies is still quite large and it’s still building. Uh and if you think of what the software world is all about, especially now with AI tools that are out there that you can get more software businesses that are created on a regular basis, the number of acquisition targets that are here today um will be multiplied several fold over the next 10 years or so. So I think that runway is still exists. Don’t know how software or what software will look like 10 years from now, it could be a completely different story, but I’m uh I firmly believe that those prospects are still there.

[03:48:54] Mark Miller: Yeah, there’s new startups all the time, some work.

[03:48:58] Larry: Great, thanks Howard. Can I just u the—the panelists had a bet about the likely population of the online group, and um Howard won the bet. There are 1,150 people online at this point, so congratulations Howard. What what was the deal like, can’t—can’t disclose yeah. Once it matches, once in attendance matches price of Constellation’s one share, slice of pizza, pizza. Uh so um so—

[03:49:38] Howard: This section uh the last one for the M&A is more about larger V M&A and also outside VMS. Uh so on the recent earnings call, you asked about, you know, valuations and kind of said you hadn’t seen that much movement, and if this kind of persists um do you see uh valuations start to decline more for smaller acquisitions or the larger acquisitions in in private markets? And, you know, does that mean, you know, you should focus your efforts towards one or the other?

[03:50:06] Bernie Anzarouth: Yeah, so going back to the debt discussion that we just had, the leverage discussion. A large of the large acquisitions are levered, right? Uh they’re private equity uh acquisitions, and some of those um uh some of that leverage is up for refinancing probably within the next 5 years or so. So uh it all depends on what those debt markets look like in uh terms of influencing whether these businesses are put up for sale um otherwise they will just continue to refinance if they’re able to do so.

But with the SAS SAS apocalypse uh that’s going on, who knows uh how that’s going to work out. I mean, there’s billions of dollars out there in leverage uh for some of these software businesses, so um time will tell um i think we might see something in the next 5 years, or or it will prove out in the next 5 years whether that’s up or down, impossible to tell. But we stick to our, you know, bread and butter—those the smaller businesses—that’s where we get most of our acquisitions from, and I think those um will be coming up for sale on a regular basis, and and we hope to be um front of the line um ready to acquire more of these businesses um—

[03:51:11] Howard: Maybe an update on VMS Ventures. So the shareholder’s wondering, you know, how many investments have they made, you know, have they scaled successfully, and any kind of spillover benefits? And also, you know, some—are there any kind of AI-focused ones, and I know there’s been—

[03:51:30] Mark Miller: a I’ll let Don answer that question. Do we have Don’s in the crowd here? Microphone, mic. I’d love—you flew all the way over here. So Don is uh—I get to participate in VMS Venture calls us e once a week or once every couple weeks, and I just love uh love uh Don, and I appreciate him hopping on a plane and flying over here. So okay, thank you um yeah we see—so introduce yourself first, Don, just—

[03:51:53] Don: Well, I’m Don, nice to meet you all, yeah.

[03:51:55] Mark Miller: What you do though, yeah.

[03:51:55] Don: i do something with topic topic run it, yeah. So, and a couple of years we started the venture filmed with Carl Shabas is also here and Mark Leonard, mhm. And uh what we see, we started very slow, looking for the right angle to uh to make this corporate venture uh running. And in the past year, we see additional traction in AI-specific uh companies. And as you might know, it’s an internal uh corporate venture uh investment uh committee that we that we run, and in the past weeks months we close around two uh of AI-like startups. So um one of those was already mentioned by ClickDimensions uh with the agents, it’s a product uh of uh VMS Ventures that they actually use, it’s called Rhea uh and it’s a low-hanging fruit way of um yeah, deploying AI agents, something like that.

[03:53:02] Howard: Excellent, thanks. That was good on the VMS Ventures yeah, anything else Don?

[03:53:05] Mark Miller: Thank you so much, it’s great your son came over with you as well, really appreciate having you guys here. u And then just uh the next one to uh to David um so this shareholder is asking, you know, in many carveout situations, the seller’s objective is to make sure there’s a smooth transition with the buyer um that has a credible operating model and execution, which which you’ve shown um are there carveout situations, you know, where Lumen already has the support of the seller, but closing is deferred mainly because Lumen doesn’t have the organizational capacity yet or timing alignment to absorb the asset?

[03:53:42] David Nland: Yeah, no, definitely not. So we, as I said earlier, we’re always one year in advance on building placement infrastructure so where to put companies and and obviously M&A capacity to deal with opportunities. So yeah, we got lots of places to put companies and we’ve got capital and the will and the credibility, and and the bar—barriers to entry are very high on carveouts because they’re extremely complex, especially the larger ones with uh dysfunctional sellers um so it’s a yeah, it’s a big focus area for us, and we’ve got—i wouldn’t say unlimited capacity to deal with it, but we’re far away from being organizationally constrained.

[03:54:22] Howard: You’re scaling up the team thinking way, yeah. Um and um maybe a really broad one to uh kind of round this out: you know, if uh Constellation’s objective is trying to deploy 100% of free cash flow over the next decade without materially increasing dividends, what do you think the company needs to do differently from the last decade to make that possible? Is that, you know, uh acquisition criteria, which we’ve touched on a little bit, lowering hurdle rates, organizational changes?

[03:54:54] Mark Miller: It’s really developing people, like we got to continue to develop people like we have, because you know, we we deploy all the capital because we have a deep bench of people throughout this organization. And if we don’t continue to develop that, if the leaders of our leaders of our leaders aren’t developing their people, that’s going to be our our challenge. It’s—and it it sounds like an apple pie motherhood thing, but it’s it’s very true, especially if you’re decentralized capital deployment, right? So you need to have people out there that uh you measure and you trust, and you develop as as they uh as they learn the ropes of capital deployment like all of us did that are up here.

Question 19: Leadership Identification and Executive Succession Blockers

[03:55:31] Larry: Want to cue the audience chair, so um Howard will ask one more question in this round and then we’ll turn to another segment of audience questions. So if you think of one and join the mic, you have the last question.

[03:55:42] Howard: Yeah, and uh the kind of flip side of that is uh if you cannot deploy all your capital um what will you do? And, you know, if uh it’s below your magic buyback or above your bionic buyback number, you know, what else will you do with that capital?

[03:55:55] Mark Miller: Answer we had for PEMS, right? So yeah, i mean, we we’ve done special dividends in the past, so there are many ways to allocate capital, and and obviously we wouldn’t endeavor on a buyback if it didn’t make sense for our shareholders with our high rates of returns. But we have used special dividends in the—we’re not going to buy the jets though, so you might have to return them, right? Yeah, all right.

[03:56:15] Howard: That’s that’s it for me, so thank you so much Howard.

[03:56:23] Larry: So um it looks like there are no shareholders at the mics um but one is one or two are traveling, so we welcome uh so it’ll be question 15 from the audience uh next. Let’s in the middle aisle again from Wim investment management.

[03:56:41] Shareholder (WCM Investment Management): I guess a quick one on just PEMS. Um given how it’s kind of about uh deploying Constellation’s own operating best practices into these companies, is it fair for us to kind of look at it like let’s say five six years out if we look at these companies then um given the kind of time delay it’ll take for these best practices to get absorbed, that’s fair. Would that be a good indication of how some business units actually are transformed inside Constellation? Like, in a way, it’s for us we can monitor it in a public sphere how these operating practices actually get digested and how the actual financials start evolving, or would it be a little too different?

[03:57:24] Mark Miller: Personally, I hope that the performance of the business will show up in the financials. I don’t know about the share price, but the performance is in the financials and that’s what we really care about, and and that’s what shareholders really care about, right? Proving the intrinsic value. But I think the difference though is that when you—we have control, I’m not sure we would be five years patient. Yes, it will take—in PEMS, I I think we we would be, but I I don’t think internally we would typically be that patient. Thank you.

[03:57:53] Larry: Thanks, Dan. Uh question 16 to the right side.

[03:57:59] Shareholder: I just have a question on uh talent development. You mentioned about leadership development is extremely critical, yes um but everybody has their DNA: some people are born to be better operators, some people are born to be capital allocators um just wondering how much patience do you gave to younger leaders in terms of how do you measure them, and what happened after a while that uh it just they’re not successful, and then how do you—

[03:58:20] Mark Miller: Retivis a couple years ago, and we did an event called Quadrants in London, and and uh I decided I’d invite like uh four or five people on stage and just interview them about their careers. I think three of the five started as interns and were running portfolios. So, you know, we have a list of who you are, we have a list as to what you’ve done, what returns you’re getting, what uh performance you’re getting from an organic growth or perspective, and we just want to develop those people. And some are—you’re right—some are better operators and they’re better at making businesses run better, and some are better at allocating capital, and we’ll just we’ll just try to filter that out throughout the organization. I think all the people up here do that well. So so that answered your question, but uh yeah, thanks, we really want to do that. It’s very important part of our culture is—is the question that you do you think you have that people have to go into M&A to be successful or—

[03:59:20] Shareholder: No, I’m there are people interested in going M&A, right? DNA, everybody’s interested until they get there.

[03:59:27] Mark Miller: Yeah, i mean, they get there, but some of time they run into difficulties of not being able to find the right totally opportunities so—

[03:59:34] Shareholder: Well, some people want to go places they they probably shouldn’t have went, and that’s—

[03:59:41] Mark Miller: That’s just how it is inside of a massive ecosystem of Constellation, certainly. We’re trying to not create like—we want if if people are really great operators, you know, we love them to continue to be really great operators, and M&A has a slightly different skill set. Some people are, you know, adapt to that and some people don’t. What I don’t want to do is lose a great operator by forcing them to adapt to be an M&A if that’s not in their DNA. But a great operator also has very good judgment, you know um so they’re somehow helpful at M&A as well because the judgment on whether to make an investment, right, is is it requires, you know, some some thought and judgment despite having all of our metrics and experience. And, you know, you’ve really uh it’s a really difficult problem when you know, whether you want to like promote a great operator into helping you allocate capital is one of our one of our challenges. Thank you very much for the great question.

Question 20: Options Dilution and Intangible Valuation Metrics

[04:00:40] Larry: We’re gonna getting close to closing time, 1 1 PM. Got about a quarter of an hour, and we’d like to pivot to the final segment, the final topic: governance, you know, sort of defined broadly. Um a couple—I think this is a relatively quick question and it’s probably for Jamal, but what is the total percentage ownership of Constellation stock by employees, and in 2026 how much stock will employees rep uh purchase as bonus reinvestment?

[04:01:13] Jamal Baksh: Yeah, so employee ownership today is around 6%, and that excludes Mark and Mark Leonard in his family office, which was another 7%. Um in terms of the bonus, like again, there’s a lot of the leaders that are on the CSI bonus plan, but the number of people I actually bought bonuses for this year is only like 4,100, right? Out of 69,000, so you got to remember it’s still only like 6% of the total, and that represented for Constellation employees about 80 million bucks Canadian, and for Topicus employees about 15 million, so a decent number, but yeah. But it’s not pervasive across all of the 69,000.

[04:01:47] Larry: Thank you very much. Second question in this segment is asking for an update on your scheme to allow employees to invest into the underlying businesses directly. I think a Damian Jeff question.

[04:02:03] Damian McKay: Yeah, so do you guys both want to do? And and Jeff, so yeah, it’s uh it’s going well. It’s it’s it’s for employees in in uh in Canada and and also the US, so it’s geographic—geographically specific. Uh we’ve got a group of uh engaged employees who are who are um in this experiment, and uh you know, early days, i think it’s it is again it’s that long-term compounding outcomes. Jeff, you want to—

[04:02:32] Jeff Bender: Yeah, I would say um so year two, year one was uh Canadians only, year two is Canadians and US-based employees. The size of the pool quadrupled, I think for us in terms of of dollars available. But the first pool was uh was post-tax, the second pool is pre-tax, just the way the the rules work out. So I think it’s it’s going well. The people who are in the pool seem to to enjoy it, you know, we’re I think, you know, a lot more conversations about investing and being an investor, and a lot more engagement around the uh around the investment, so I think, you know, we’re we’re happy.

[04:03:09] Larry: Thank you. I’m going to ask two questions to John and then invite an audience question, and then ask one final fireworks question. Big question, fireworks. The questions um for John um thank God there there are two, and and I’ll uh the first one concerns um CEO succession planning. We we’ve just been through one, they’re already planning for the next one. I’m going to let it pass, this one’s going very well so far um but this is how it’s written: “In light of um Constellation’s increased scale, complexity, and decentralization, um could you comment on the approach to presidential succession planning, in in particular, what leadership attributes do you believe will matter most for the next president, and what processes are in place to ensure that qualified internal successors are being developed over time?” A lot of questions in there, Larry. Um Larry’s actually on the HR committee, i’m not, so I feel like I can’t answer my own question, seems to be—

[04:04:13] John H. Beltz: Happy to answer it, and and I just wanted to echo Mark Miller’s earlier comments that um all the employees and shareholders couldn’t be more grateful to Mark Leonard for his 30 years of leadership, and uh we wish he was here today, but he couldn’t make it. And if you have his email, by—I’m sure he he’d love to hear from you.

And the process, obviously, in the last CEO transition was was, from our perspective, pretty pretty clear, pretty quick. We had a process in place, it obviously happened under circumstances no one wished for, but it was pretty smooth from our perspective, and obviously we’re very grateful for Mark for stepping in um moving forward. I think one of the parts of the questions were the attributes, and this is—this is something I think is pretty critical the way we think about it um and there’s—there’s a few of them that are absolutely critical. One of them, which which Mark has alluded to a number of times, is the important of importance of autonomy in the way we operate in the decentralized model, and um it’s always been the case where um it’s very very difficult, I think, for a normal CEO to walk into Constellation. That would—you would never see that because the first thing they would like to do would be to centralize a whole bunch of things. It may not mean building up centralized departments, but it’s likely around decision-making.

And I think that’s the hardest thing for a CEO in our environment to deal with—not only Mark Miller, but but everyone else here um is you have to enable your people beneath you to make decisions, uh coach them, and that’s the only way we’re going to continue to grow, and everyone up here knows that. The benefit of that is you’re developing many tiers of very good decision-makers that we can see their track record over time, and um so I feel very comfortable that there’s a large bench of people that we can put on the 5-year track, the 10-year track, the 20-year track, and that’s largely attributable to the type of organization that we’re in um and that we’re developing those leadership skills through a career, and we’re able to track that that progress over time um so that that is a key part.

And the and then then the fundamental decision that a CEO makes, and that everyone everyone up here makes, is around capital allocation. I think that’s the number one job of a CEO of a company like CSI. That obviously is around investing in acquisitions, but also uh if you’re running a business unit, it’s investing in organic growth initiatives um and then it’s also at the top level it’s thinking about share buybacks, dividends, it’s really looking for the largest returns for our shareholders for every incremental dollar invested. And I think that’s another another key attribute that that we look forward in as we’re thinking about successors for all the CEOs up here.

And then it’s the intangible stuff, you know. We’re all—we’re all cheap, none of us fly private jets um we’re all shareholders, we’re all big big shareholders because of that, we we act in the best interest of our shareholders um and it’s it’s it’s all the intangibles, it’s low ego—the things you, you know, you don’t find the big American CEO with uh making the big salary, negotiating his his option package every year. That’s the number one thing they’re they’re thinking about um they’re using share buybacks as a way to avoid dilution on their option plans. There’s a whole bunch of things that I think all the shareholders who have been in the stock for a long time know and understand, but I I think are reiterating when as a board we’re thinking about succession candidates and what we’re looking for.

[04:07:56] Larry: That’s a long answer, Larry. There were many questions and you gave many answers, so thank you very much for that. This—the next question for you, John, concerns buybacks, a word that’s been uttered many times, and this one gets a little analytical um well there there many came in, and so I’ll actually pose two of them: one’s a little broader, one’s a little more analytical. The the broader one is: at what price—I don’t know if you can say this or set of condition, they they know you’re not going to answer that—or set of conditions would the board consider? Would you consider temporarily buying back shares while simultaneously continuing your deployment efforts, after all you have a lot of balance sheet flexibility here? Here’s the more precise one, this came in this morning on the chat, so I’ll read it slowly: “When you compare the after-tax compounded return on a $100 million VMS acquisition closed at today’s prevailing multiples versus repurchasing $100 million of Constellation shares at today’s price, which two or three inputs do you weigh most heavily, and how do you think about the certain a certainty differential between the two cash flow streams?” Okay.

[04:09:13] John H. Beltz: I’ll give my answer then maybe if anyone else wants to weigh. I mean, the way we think about it is we’re using the same hurdle rates for making both decisions. We’re buying businesses um at an at a value that is far beneath what we’re trading at and what other companies are trading at, even after we fix them. So we still believe we can generate much higher returns by buying companies than we can by buying our own shares at these prices around your discuss around the question around uh confidence in cash flows. We have high confidence in both. When you’re talking about VMS, we we obviously have a slightly higher confidence in our own cash flows, but when we’re looking at the types of companies, which are 95% of our investments, are industries we know, we know how to fix these companies, we’re pretty confident in those cash flows, and our track record would suggest that we’re reasonably good estimators of that. Jamal, do you want to add to that at all, or—

Jamal Baksh: I think John covered it well, yeah, I don’t think I can add.

John H. Beltz: As well, anyone else? All Charlie Munger, nothing to add. Sarcastic.

Larry: Is there an audience question, or shall we go to the final final question? There is an audience question, so it’s question 17, lucky lucky 17 going third time. Hey, ju—just second second, this is Alex, yeah.

Question 21: Options Dilution and Intangible Valuation Metrics

Alex Captain (First Shareholder): Thank you. Oh um this this is not really governance, but it’s related to the capital allocation. Um you you made some interesting comments on the most recent call about having built your capacity to do larger M&A relative to uh the past um and obviously public market valuations have come down, and those are many of the larger companies. What do you think is the outlook over the next few years for you to do more public-to-private larger M&A?

Mark Miller: i’d love to be able to predict that, you know. It’s, you know, we’re we’re just keep plugging away. The good news is we have—we just have, i think, a the best team we’ve ever had on it, so Alex. But I I can’t forecast the future. If Bernie could, I’d love him to, but he can’t either.

Bernie Anzarouth: i can definitely predict the future, no sorry about that.

Mark Miller: But we can’t really say we’re gonna we’re just—you just, you know, one of the things you hate in business is when you’re not on the field uh for the game, and at least want to be involved, and that’s one of the things that we’re going to work hard on here. And if there’s available uh, you know, possible uh transactions to do, we just want to be playing the game. Thank you, and I think we’re probably better—

John H. Beltz: i might want to add around um it’s one of the reasons that Constellation has kept a pristine balance sheet, and this goes back to Larry’s question about: why not borrow to buy back shares? I mean, we do believe that we want to have the opportunity to take advantage of which could be a dislocation in the market for a period of time. So keep selling vertical market software businesses, we want the valuations to press as long as possible, yeah.

Mark Miller: i don’t want to buy back shares, definitely don’t. Thank you, Alex. We’ve got time for at least one more, maybe two if they’re short, okay.

Gabriel Boni (IP Capital Partners): Hi, Gabriel Boni from IP Capital Partners from Brazil. My question is about AI opportunities for the whole M&A function. How how Constellation can enhance the job of the M&A and the business development guys over time and accelerate the process?

Mark Miller: Bernie’s on that, yeah.

Bernie Anzarouth: So we definitely have tools that are made available to our M&A um associates um and so uh it’s a matter of using those tools, and there’s all sorts of parts in the workflow uh from finding the right leads all the way down to to doing the diligence and and closing the deal. And so we’ve um built internal tools and we’ve used tools from the outside to um make that workflow a lot faster, a lot smoother uh whether or not it proves out in in um increasing the volume um that uh that we actually get to close is up for debate, we don’t know yet. But those tools are being made available, and we’re we’re uh rolling it out to all of our M&A people.

Question 22: Competitive Landscape in M&A

Larry: Thank you. Question 19, probably the last audience question.

Francisco (Rothschild): Thank Francisco from Rothschild. Just given everything that’s going on, do you expect any changes when it comes to competitive landscape in in M&A changes? Are there going to be more copycats, you mean? Well, even all the money that was flowing in, is there do you expect less money to flow into this space?

Bernie Anzarouth: Uh yeah, hard to tell. I mean, right now we’ve in the last decade or so, there have been a lot of copycats that have been popping up. Uh I’ve seen some exits of those copycats um I think XO is one of them, actually, uh sorry um yeah, XO that we bought through Harris. It was a rollup of software businesses in Quebec, so I’m personally hoping to see more of those copycats maybe throwing in this to howl, that would be nice. I don’t think that’s going to happen right away u but over time, I think some of these uh rollups might uh be looking for an exit, and uh if um if that is the way it’s going to go, hopefully we’re first in line for the first call that uh that goes out. But um I think there will be some more copycats that will pop up, uh maybe they’ll do a different version of what we’re doing uh but it’s uh we we’ll be out there looking out for them.

Larry: Thank you very much, it’s exactly 1:00 PM. And so i’d like to turn the meeting back over to Mark Miller.

Mark Miller: Yeah, i just um let’s just want to thank the whole team here, all of our shareholders. Mark Leonard obviously, our panel, which I think let’s give our panel a hand of—that’s hands, that’s like hard to do, you guys awesome job. And uh yeah, and I hope um I hope you found this useful, we look forward to your feedback. And uh obviously, I have to thank, you know, the the tens of thousands of Constellation employees across the world that are, you know, really make this all possible for us all. So uh have a great trip back home wherever you’re going, and uh thank you very much for all your time today. And there’s a lunch outside, i should say, there is some lunch outside, and uh yes, it won’t be roast beef on a platter with uh but it’ll be it’ll be something to snack on. Thank you, hey guys, thank you so much.

Good Investing’s disclaimer

Please be always aware that this content is no advice and no recommendation and make sure to read our disclaimer:

Good Investing disclaimer