Tyler Technologies, Inc. (TYL) Earnings Call Transcript & Summary

Good morning, everyone. Welcome back to the Day 2 of the Citizens Technology Conference. Good to have you all here this morning. Really my pleasure to host Brian Miller, the CFO of Tyler Technologies here for us today. Welcome, Brian.

Thanks, Trevor. Nice to be here.

We're going to run through just kind of the basic fireside chat, and then we'll open it up to the group here for questions for Brian. So again, thanks for being here and let's take it away. Brian, I'll spare you going -- jumping right in to the AI topic because I think that's probably the knee-jerk reaction. So let's talk numbers.

I'm sure we'll get there.

Yes, we will get there.

So last year, 2025 ended non-GAAP operating margins 26%. What's your -- the target margin is 30-plus for 2030? Can you maybe just walk us through sort of how you're envisioning the pathway to get there and the different kind of aspects or avenues for that journey?

Sure. Yes. At our Investor Day in 2023, we set out those targets for '25 and 2030, with the operating margin target being 30% plus. And certainly, 30% isn't a ceiling for us, but that's sort of an interim target for us. We're well on track with that. '25, we actually exceeded where that sort of interim step we thought we would be at. We said it -- we always said it wouldn't be a straight line, so we need to average about 100 basis points a year, and we've gotten more than that in the first couple of years. We laid out a number of things. It's not one thing that drives that operating margin expansion, but a lot of it is related to our ongoing cloud transition, so underneath kind of the cloud transitions as we move. We have moved really from new business almost entirely away from licenses, but we still have a big, not quite half, but by revenue, about half of -- a little less than half of our customer base is still on-prem. So as those customers move from on-prem to the cloud over a multiyear period, that generates a significant revenue uplift and improves our profitability. Underneath that, we have moved from our own proprietary data centers to pretty fully utilizing the public cloud at AWS. We closed our last data center in the -- late in to last year as we planned. There are still some bubble costs associated with that transition that will continue through this year. So we'll get a little bit more margin uplift next year from that. There's a very active process of version consolidation going on. So not only do we have a lot of products for the public sector, but we have historically supported multiple versions of a lot of those products. So it's been very expensive from a development standpoint and a support standpoint. As we move towards being fully in the cloud and having all of our customers hosted in AWS and ultimately, everyone on one version of software, everyone upgrading at the same time, really kind of taking full advantage of the cloud, not just where they're hosted, but how we manage the software and how we manage the customers. Customers who aren't on the current version of the software need to upgrade either before or when they migrate to the cloud. So we've been encouraging that by sunsetting older versions of products and moving customers along towards that most recent version so that they're in a position to move to the cloud, but also seeing margin benefits as we accomplish that, and there's still some of that work left to be done. We've also been releasing cloud optimized versions of our products, so products that are architected to fully take advantage of the public cloud and the features in AWS and lower our hosting costs, and we continue to -- we'll continue to see progress on that. I guess lastly, there's just a lot of scale. As we move more of our customer base into AWS, the pricing is very much scale-driven than capacity driven. So the more capacity we buy, the lower the unit costs come and we continue to become more efficient. Those are really the vectors around the cloud transition and those are responsible for a lot of what will be both gross margin and operating margin expansion. We're also continuing to improve the margins around our transaction business. We've -- certainly, the margins there aren't the same as software margins, but we have the opportunity to continue to grow those margins primarily as we move away from more sort of commodity type payment processing that we, through NIC, which we acquired about 5 years ago, did a lot of. And we're really -- our model now is really payments that are embedded with a software product like a utility billing system or a traffic court system. So we're able to provide significant incremental value to customers, and we're able to get premium pricing for that. So for example, we -- our contract with the State of Texas for payment processing ended in the fourth quarter. We've known about it for a couple of years. We didn't bid aggressively on the renewal. That was a commodity type low-margin -- 8% to 10% margin business that really didn't have the same connections and opportunities to expand business that really all of our other payment relationships have. So that's one of the things that's contributing to that. But we do expect that we'll continue to see as we put more and more customers -- software customers on the payment platform that, that will drive margin expansion around our transaction business. And then lastly, there is OpEx leverage over the long term. We've been, I think, very successful at continuing to improve the efficiencies around both sales and marketing and our G&A costs. R&D, there is a bit of an expansion right now, mostly around AI, but we can still do that within the framework of the operating margin expansion. And I guess the last thing I'd talk about is in our mix. Professional services is part of our revenue mix. We typically do the services related to an implementation of one of our software systems with our own professional services staff. That is low or negative margin business typically. And we've been working really hard to reduce the level of services that it takes to implement a new system. There is some use of AI there that we expect to expand going forward around data conversions, but also just discouraging a lot of customization and the cloud being a little bit more efficient about deployment. So this year, for example, we've said pro services only grow 3% to 5%. So that change in the mix or reducing that in our mix also is positive to margins.

Great. That's comprehensive. Thank you. So it seems like 30% plus then is kind of just, as you said, a ceiling and -- or not a ceiling, it's really just kind of the start. That's great. Since you brought it up, let's talk a little bit about the transaction part of the business, 5% to 7% growth there, exiting out the Texas piece, it's back up to double digits, 10%, 12%. How durable is that double-digit growth rate? How do you see that? Again, once we sort of lap the effects of Texas kind of throughout this year, how does that look when we're sort of exiting that piece out fully?

Yes, without Texas, yes, it's kind of 10% to 12%. Again, back at our 2023 Investor Day, we said that we expected growth kind of in the 10% to 13% range CAGR over that 2023 to 2030 time period. So we're kind of right in there that -- and that's what we're very comfortable with kind of at least that low double-digit growth. We've been a bit ahead of it over the last couple of years, part of that because we've been really successful in the early days of selling payments into our local government software customers. There's still a lot of runway ahead of that. We've also benefited from some pricing increases from some of our third-party payment partners that we have revenue shares with. We think that's largely played out at this point, and that's not an area where we really dedicate new business. But -- and then I think the third thing that's contributing to -- or has contributed to a little bit higher transaction growth is around the model where we occasionally sell software, but get paid for it through transaction fees rather than SaaS fees. For example, we -- the largest that we've done there was with the California State Parks for our suite of outdoor recreation solutions. So SaaS solutions that manage all the aspects of the parks like campground reservations or tours, retail, every aspect of running the parks. But rather than California, which does have budget challenges, rather than trying to appropriate funds to pay for that SaaS fees, we're getting paid by convenience fees. So if you make a campground reservation, you might pay the state $15 and there's a $5 fee that goes to Tyler. If you tour the Hearst Castle, maybe there's a $3 fee that's tacked onto your ticket. So that way, the state doesn't have to budget funds. We get paid at least as well as we would have under a SaaS arrangement. It can be kind of seasonal and not necessarily pro rata revenue recognition, but is a very good model for us and the customer. And we've seen more instances of that. It's not the primary way we sell software, but we've done that with -- how we did in the fourth quarter with a motor vehicle registration, digital motor vehicle titling system for a state. And so there's a little bit of that in our -- that actually SaaS growth would be higher, but transaction growth is benefiting from that same impact on margins.

Great. Cool. Well, without further ado, maybe we'll slide into some AI discussion here. Actually, I wanted to start maybe AI, but also talking about flips maybe kind of together almost a little bit. So you guys -- it seems like you've had great visibility kind of quarter-over-quarter on what the flips look like. You've got kind of a good kind of sales team doing a good job of saying kind of what's coming. And so you've got that, again, you have that line of sight. And we talked about a little bit of this after your earnings call most recently. Can you tell us a little bit about when you do that flip and they're transitioning over to the cloud, is that a -- is there a room for a competitor to come in and kind of essentially since the person is already migrating -- the customer is migrating over, is there for them to just kind of come in and say, let's just see what else is out there aside from Tyler. Can we maybe talk about that dynamic, especially in this new AI world we live in, there's potential disruption. Is that a trend or a thing that happens?

Yes, not really. That almost never happens. Generally, the decision to flip, and flip is our term for migrating from on-prem to the cloud. That is -- it's changing where the software is hosted and ultimately changes the way we manage the software in terms of how we deliver releases and upgrades and how we manage the customer, all designed to make it a better client experience as well as a more efficient experience for Tyler. But from the user's perspective, there's not really a change. You don't have to retrain people or convert data. So the decision to flip to the cloud is really kind of stands on its own. I just can't think of any situation that I know where the customer has taken that as an opportunity to think about changing. What it has provided for us is an opportunity to sell more stuff to them because they may have a core system from Tyler that they're moving to the cloud and there are ancillary products either that they don't have today or that they have from other vendors and maybe those vendors don't have a cloud solution. So we've been more intentional about trying to take advantage of those or look for those upsell opportunities. So for example, they have our court case management system, which the hub of the courts, but they might have someone else's jail system or someone else's jury system. So this is an opportunity. Maybe they're not looking to replace those right away, but this is an opportunity to say, okay, you're taking this system to the cloud, obviously, you don't want to leave -- you probably have some long-term strategy about moving those others, but this is an opportunity to move those with Tyler to one integrated system. Our biggest flip last quarter was Los Angeles County, which is our largest permitting and licensing customers, so the system used to manage all the licensing in the county like building permits and business licenses that sort of stuff. They flipped to the cloud, went from paying us, I think about $950,000 a year in maintenance to a little over $2 million a year in SaaS fees for the same product, but they also added a product we have that -- a SaaS product that we acquired a few years ago for fire prevention, fire inspections, and that added about another $750,000 of ARR. And they signed up to do payments with us. So we'll be processing all the payments for all the licensing and permitting in LA County. So a very significant -- ultimately, when it's all kind of at full run rate, a pretty significant uplift in revenues. And we think that it's much more of an opportunity to expand our relationships with customers as they move to the cloud and bring more of their products into Tyler. As far as AI, just in general, in sales processes, we were not really seeing it be a factor in new sales processes or really in flips. Public sector, not really surprisingly, is slow to move, slow to adopt new technology, risk averse, like things that they already see working somewhere else. And it seems to be the same with AI. There's certainly a lot of curiosity. We talked about a lot last year at our user conference. We have talked to our clients about, with each major product, with our road map for integrating AI into those products are. But -- and customers are curious, they want to know more, but they're not really looking for AI in a big way right now. They want very practical kinds of applications that solve real-world problems, that have -- provide value, have a return and they're still learning a lot about that. We believe that Tyler will be the company to provide that around our products. We're not seeing in new sales -- people pause a new sales process because they want to get AI into the ERP system or the public safety system. We're not seeing people going out and hiring, either doing it themselves. That's not something you're going to see, especially at the local government level, or bringing in third parties or start-ups to kind of drop AI on top of an existing Tyler product. When we talk about why we think Tyler shouldn't have the right to win the opportunity to bring AI to our clients, not necessarily every last AI application, but certainly around Tyler's mission-critical applications, there's kind of 3 things. One is the domain expertise that we understand. These are typically very complex workflows that we have a deep understanding of. And so being able to integrate AI into those, they trust us to do that because we understand that. The second is the relationships and trust. As I said, they really -- relationships, references are really important. Trust is important. So these things have to work really be right, and they have to trust who they're getting it from. And so doing -- expanding the relationships with Tyler and getting AI around our products from us is something they're comfortable with. And we have decades-long relationships with many of these customers. So we're not a startup. They've known us for a long time. And the third thing is really around data that we have data not from one customer, but from hundreds or thousands of customers running these systems. So we really believe that that's -- those are the keys to the opportunity for Tyler. And we have some products that are currently fairly widely in use. Some of them came from acquisitions in the last 2 or 3 years that are kind of use AI at their core. And then we're doing development around adding AI sort of features or add-ons to a lot of existing products. So they're typically pretty -- at the early days. It's very kind of boring, very kind of basic things that we're using AI to do. So going back to the licensing and permitting system, a big pain point for local governments is around application reviews, especially around building permits. So in a lot of places, there's a many month wait to get a building permit because there aren't enough clerks to review those. It's a manual process, and they have a lot of backlogs. That impacts development, that impacts property taxes and citizen satisfaction. And so application review is something that is one of our early things that are in pilot right now and expect to be more widely available this year to use AI to automate that process and solve a problem, clear value. And it's not necessarily that they'll go hire a lot of clerks, it's that they don't have enough to start with. Staffing shortages are a big problem for governments, especially at the state and local level. Report writing in our public safety system, our records management system. The average police officer, I'm told, spends 2 to 3 hours a day writing reports. So integrating AI into that system to assist them with writing those reports, creating a lot more time to be more productive and on the street. Data entry, using AI to automate data entry in the court system. We have a product there called document automation that uses AI to enter data in the system and solve the problem of not enough clerks in the courts and backlogs of cases waiting to be entered.

Great. It sounds like -- and I'll open up for questions in the last few minutes here. It sounds like the use cases you went through are very important, but almost niche to the point where it wouldn't necessarily be a viable business model for one startup just to do that particular application review, right? But you having the knowledge, you're already there, like it's most like kind of a walk -- crawl, walk, run type of approach, which is almost a moat in and of itself compared to others coming in, right?

Exactly. I mean, from a technical standpoint, someone could come in and drop it on, theoretically they could do it themselves. But then they got to take care of it. And we've got -- again, we have thousands, hundreds of thousands, maybe millions of building permits that are going through our system. So we get the data to train that. We understand the workflows. As those processes change over time, laws change, processes change, we can manage that for them. So there's just not a reason for them to go somewhere else and to create more complication and uncertainty. And these are things they need to get right. I mean, for that function, permits need to be right. So -- or police report writing needs to be right. So yes, I mean, we believe that's just -- we liken it a bit to the transition to the cloud. When the earlier days of the cloud, when we were mostly on-prem or all on-prem, there was a narrative that cloud-native competitors, there'd be lots of start-ups that would emerge and develop software really quickly and create lots of new competition that would take market share. And some of them did emerge, but generally, they didn't have sales channel. They didn't have the relationships or references. So it's a long, slow go. Even though they might develop software quicker, they didn't have the deep domain expertise so the software wasn't as deep in functionality. Some of those little narrow point solutions became really good acquisitions for us that we could then grow within Tyler. And some of those just never really -- a couple in the public safety space that kind of made a splash for a little bit, but then really stalled out. Meanwhile, Tyler was able to move to the cloud and move our products to the cloud and bring our customers to the cloud. And so we sort of see a similar situation here as opposed to being real deep threat.

Great. All right. I'll open it up to the group. Sir.

[indiscernible]

Yes, it's primarily cost reduction. So they're looking for something that has real value. So I'll give you an example with the document automation solution I talked about. So we're the biggest provider of court case management systems in the country. Our systems are entirely digital. So there's no paper in the courts anymore. That's one of the big advantages to the new -- to replacing that technology. But a clerk still -- when a case comes in, the lawyer files a lawsuit, a clerk still has to set up a case in the system. And so it's a manual process. So document automation uses AI to identify all those data elements in the document and populate the case file. So they can have a really clear understanding of how many fewer clerks they need. And again, usually, those clerks aren't there already. There's just not enough of them. So they may have a budget for that, the labor budget, but trouble finding the people. So we saw that last year, one of our deals was Hillsborough County, Florida, which is Tampa, so one of our good-sized court customers. So they're paying us -- they pay us, I think, about $750,000 a year for maintenance for their court system. They're paying us about $950,000 a year for the document automation AI add-on, but they are comfortable that they're going to save well over $2 million a year in labor cost. So it solves the problem that's a real problem. They can really understand the value. There's a relationship between what they're paying and the value that they're getting. I think over time, some things will become more embedded in solutions and not necessarily separately priced SKUs that will increase the value of a solution. But we have another product called priority-based budgeting, which uses AI to help a government, once they identify sort of their priorities in their budgeting, helps them identify places in the budget where they can reallocate funds. And so high value, solves problems, practical problems they have pretty quickly implemented. So we've seen really good uptake with that. But usually, it's around a very practical problem that's recurring pain point for them. To solve problem and AI is a tool that does that. They're not saying, "I just -- I keep hearing about this and so I want a bunch of AI here." They don't have AI strategies typically now, maybe some bigger places, state government. The other thing we're doing is a lot of resident engagement portals. So to your question about better service to citizens, so we now have 6 states and then just last quarter, signed our first County, Fairfax County, Virginia, where we're providing resident engagement portals, which use AI agents to answer citizens' questions. So rather than having to call around and find the right office, call the tax office and get a question answered or search the web, it uses AI agents to answer questions and working really successfully. That's -- again, that's something that is not a unique concept to us, but we have this deep understanding of how state governments work and how county governments work and how all the -- where all those answers need to come from, how they need to be right. So that's something someone else could theoretically provide, but we think we have an advantage there.

Yes. Great. Maybe one quick one.

Do you have any strategies internally since you have all the [indiscernible] knowledge to actually automate some of your customer support? [indiscernible] down there or...

Yes, good question about internal use of AI. So 3 primary areas, development where we're already using it, and we will continue to expand the use of that, but we're getting efficiencies around developers. Still a little hard to measure exactly. And we don't have fewer developers so we're getting more work from the same developers, and we have to decide whether we want to cut the cost by that amount or do that much more work. It's probably a blend somewhere in there. Support is certainly a big area. We're using AI agents internally right now for our customer support reps to answer -- to find the answers to questions and to work with each other. We have not widely deployed AI agents for direct customer interaction, but I believe that is coming at some point. And the third area is around professional services, which is a lower, no-margin business for us. And using AI to automate processes like data conversions, which is a manual labor-intensive thing that we have to do in every implementation. And so we're expecting to broaden the use of AI there.

Great. Thanks, Brian.

You bet.

Appreciate it. Thanks a lot.