Great. My name's Rudy Kessinger. I cover security infrastructure software here at Davidson. With me today, we have CEO of AvePoint, TJ.
Yeah.
Head of Investor Relations, Jamie. Thank you guys for joining us.
Thank you for having us.
Thank you.
Good. We'll jump right into it. You guys have been around a long time, actually. Since, I think you were founded in 2001.
Yeah.
I still find, even though you have several public competitors, and you're reaching some scale, that a lot of investors are still somewhat unaware of you. Just talk us through the company story, where you started and where you're at today. Let's start there.
Thank you, Rudy. We often got those commentaries like we're the best company people never heard of. We started in the Microsoft Office space, specifically Microsoft SharePoint, which is their enterprise content management platform. Initially, we start with data backup, data archiving, data migration, and then governance. When Microsoft went to cloud in early 2011, we were the first to invest aggressively there, taking our entire portfolio product from on-prem to become SaaS. Today, we're the largest Microsoft Office Cloud ecosystem player when it comes to unified SaaS data integration, data backup, data archiving, and governance player. The space that we started in is regulated industry because those are the ones, especially government, banks, care about policies of data access, data retention, record management, et cetera. That's where we started.
What we see, though, is increasingly, once we went to cloud, we were able to actually be more accessible because the maintenance cost of subscribing to a SaaS vendor is much lower. We became much more accessible to small to medium sized businesses around the world. Not only do we now sell to large enterprises, we also have nearly 50% is small and medium businesses, fully recurring. That's a really good development. Second to that is this whole emergence of GenAI. Folks are realizing that in order for AI to actually know anything intelligent about your company, it needs to grind on your proprietary data. This is the key. Also unstructured data. Most companies, their data's scattered everywhere, emails, share drives, Box, Dropbox, CRM systems. To be able to organize that in a meaningful way, to remove the out-of-date, trivial, redundant data.
To control the access also, so that your marketing team doesn't have access to your finance data, doesn't have access to your HR data. That's become something that's essentially table stakes in the age of AI. Before, AI can be secure by obscurity, you can't find it anyway. Now it's actively surfaced out by AI. This is where all of a sudden, everyone, every industry, not just regulated, they realized that the data foundation is very, very important to successful AI deployment. Today, how we actually position ourselves as part of the AI infrastructure, so you have the AI energy layer, chips layer, data layer, model layer, and application layer. We're between the data and model. Effectively, we help enterprises better curate their data estate, better manage access control, better manage the life cycle and retention of data, therefore they have better AI output.
Of course, last year's theme was people using AI. This year's theme is AI using AI, effectively agentic era. That then accelerates a lot more shadow AI risks. Of course, now also AI cost. In the last three months, people saw that as Anthropic planned to go public, the price spike is real. Of course, when Anthropic increased prices by 500%-600%, everyone else follows. What we see is that intentionality now finally settles in. Every CFO has to worry about consumption-based, token-based budgeting. That means there's a lot more intentionality towards responsible and cost-effective AI deployment. All of that works very well in our wheelhouse.
Yeah. I'll come back to AI in a bit.
Okay.
I want, actually, if you could expand on your partnership with Microsoft and how you differentiate for M365 and in the Microsoft ecosystem. I think some investors, they hear the significant-
Yeah
concentration of Microsoft. They view it as a risk. I know you guys obviously view it as a strength and-
Right
an opportunity. Expand on the partnership there and the opportunity there that still remains.
It's interesting, right? When we first went public in 2021, people viewed Microsoft as a risk. Then, of course, as Microsoft became one of the first movers, their partnership with OpenAI, at one point, everyone viewed that as very positive because they were dominating the news headline. At that time, Google was in code red. Now Google is doing super well with Gemini and NotebookLM, and Microsoft's now internal code red. Now that worry surfaced out again. We have been an ecosystem player for the last 20+ years in the Microsoft ecosystem. Now we're active in the Google ecosystem as well as Salesforce ecosystem and AWS ecosystem as well, because the world is multi-cloud. We never depended on Microsoft for business. Microsoft never resell anyone's product, us, or any other party.
We are part of that trillion-dollar ecosystem where we help customers complete the last mile problem, right? Help them activate workloads, whether it's Purview, whether it's E5, whether it's Copilot. In that sense, we're actually very complementary to what Microsoft offers. Oftentimes, when Microsoft releases new offerings, we are the very small group of launch partners, because every hyperscaler, Microsoft, Google included, they rely on a robust and rich ecosystem to form their overall stickiness and competitive moat against other hyperscalers. This is where we play. We are complementary. We don't compete against Microsoft. We are helping our customers to complete that story. That's the relationship with Microsoft, right? It's actually a complementary relationship. Now customers are very much increasingly more multi-cloud. As we expand our footprint within existing accounts, we now support multi-cloud so that we have increasing strategic value to our customers.
Then just maybe one more on Microsoft before I shift gears. There's also a lot of investor questions around seat-based models.
Yes.
Specifically with M365, I think in the backup space.
Yes
potential price compression as you've seen a number of competitors add M365 as well. What are you seeing from that dynamic? I think maybe just go a step further. What percentage of seats are still not backed up by third-party vendors in M365, by third-party vendors?
Yes, that's a very good question. Microsoft Office Cloud have roughly about 500 million user seats, between 4- 500. We cover tens of millions. We're not yet in the 100 million range yet, but we're going to get there quickly. We still think there's room for growth. We know based on our cloud consumption, based on our existing contract with Microsoft, we are one of the largest. We have, for example, a much larger Azure contract than many of the other backup vendors out there. Partly because many of them are not fully SaaS. Many of them are still very much, like Veeam, for example, is not SaaS yet. Part of that is our, Microsoft is not a primary source of revenue for them, they're getting into. There's still space.
However, having said that, Microsoft recently released Backup Express, which is a snapshot technology. We also incorporate that as part of our backup stack. There are aspect of backup where you do need to snapshot the entire environment, and there are aspect of it we're doing real-time incremental backup and recovery because, you think about when you have a cyber attacker or AI damage, there's a small blast radius. It's not the entire environment. You do need more real-time recovery capabilities. It actually forms the entire stack. Microsoft told us that we actually sell more of their, this underlying snap capability than themselves.
Microsoft sellers out there are trying to sell E7 licenses. They're not specific licenses on a base level backup that's not also, doesn't do everything, right? It's just a base level coverage. From that perspective, we have some of th e fundamental core Microsoft capabilities in our product portfolio, like I mentioned. When customer have a Purview, we actually complement Purvi ew. We use Purview lately. When they don't, we have products that will classify and tag and govern data for you. Most customers in the world are mixed license types. To today, right? E5 license has been out how many years? Eight years, et cetera. The penetration rate is less than 20%. Copilot's been out, what? four years. The penetration rate now is, what, 5%?
While Microsoft, Google, Salesforce, they all want to sell the highest license type to all their customers are actually pretty sophisticated these days. They will pick and choose on the level of investment that's right for them. This is where we come in, right? Fundamentally, we exist, and we thrive because we solve two issues for our customers that the hyperscaler would never do. One, we help them maximize their existing investment across different license types and across different workloads. Hyperscaler don't do that. They want you to buy the most expensive of everything because they're becoming utility companies. Just like, if no one's in the house, the utility company rather you leave the air condition on. Second, customers are multi-cloud because they want to choose the best of hyperscalers.
You throw a stone at an enterprise customer, they most likely use email from Microsoft, but then they use probably CRM from Salesforce. They probably use Compute from AWS, and they'll most likely now start to use Gemini on top of their Microsoft infrastructure. We see that happen in the enterprise as well as public sector. Because of that, this multi-cloud flavor is something we then help customer address. We can manage and govern and secure data across multi-cloud, including even Atlassian, monday.com, and many other data sources.
This is something that the hyperscaler would never do, right? Microsoft wouldn't want to cover anything on Google and vice versa. These are two fundamental differences of objectives that ultimately we are very aligned with our own customer success and we will continue to grow and thrive in that space.
Super helpful. You guys have, as you mentioned, you've expanded to Google, Salesforce, Atlassian, other SaaS applications.
Yes.
In terms of your coverage and what-
ServiceNow as well. Yeah
what you can protect.
Yeah.
Talk to me about the level of traction you're seeing there.
Yeah.
I guess when might we hear you guys say 10%-20% of new business is coming from those products?
Yeah.
How are you enhancing the go-to-market motion from a sales enablement, from a partner enablement, from a marketing standpoint.
Yeah
Your customers know that AvePoint is not just the Microsoft-
Yeah
company now. They protect everything. Talk about all that.
Yeah. That's a great question. We have given guidance to say we will consistently grow at 26% CAGR on ARR for the next few years, get to that $1 billion ARR as fast as we can. When we do that, we expect the non-Mi crosoft revenue mix will be 30+%, right? Versus today's less than 10%. That's the guidance. In term of the product coverage, we already have that today. Everything is the business challenge, and momentum is about go-to-market. Each ecosystem, whether it's Google or AWS versus Microsoft, they're different. They're different players, they're different go-to-market players, partners, and different channel. That's what takes time, right? To kind of not only get the tech ready but also go-to-market.
We do have direct customers. The most recent earnings, we highlighted that this Middle East conflict in the early days when AWS regional data center got taken down, we see a massive spike, several hundred percent increase in orders on resiliency from our Middle East customers. The massive spike also from our European customers, and therefore non-Microsoft workl oads, right? Because we cover them. One, these type of trigger event make people realize, enterprise customer realize they cannot just rely on hyperscalers for everything. Two, they need a highly available, cloud-based, resilience partner to help them recover. Because now we also have supply chain shocks on hardware, right?
SaaS-based solution, it's much more resilient and high speed of recovery than non-SaaS-based solution, right? From a upgrade and maintenance perspective. This is where, from a direct perspective, we have that demand, but where we're going to see real momentum is this channel play, which we invest aggressively since we've gone public. By our investment in channel, we're able to lower our sales and marketing costs from 44% of revenue to now 31% of revenue. All of that saving goes straight to the profitability of the business. We're able to continue this robust double-digit growth while maintaining double-digit profitability. We're now GAAP profitable as well. We control SBC very well. That's something we feel is very unique, that also makes us very resilient going forward.
On the supply chain shocks, it is a important distinction, I think, versus your other public competitors-
Yeah
in the backup space, 80% of your revenue is SaaS. I'm curious-
Yeah
roughly, I think.
Yes.
I got that wrong.
That's right.
Higher memory prices, higher hardware prices, longer lead times for on-premise boxes, how is that impacting your business, if at all? Certainly there are some puts and takes with some of your other larger competitors.
because
in this space.
Yeah, that's a great question. Our infrastructure is all hyperscaler infrastructure, right? We have 18 instances sitting in GCP, in Azure, in AWS data centers, as well as U.S. government data centers. The nice thing about these hyperscalers, they're like very large airlines. They do their fuel hedging very well, multi-year hedging. Same thing, because hyperscalers have the capacity, they d o their hardware pricing hedging very well. When we do sign a multi $100 million contract, for example, with Microsoft, we are guaranteed pricing, right, at a better discount than most people can buy on a smaller contract. That guaranteed pricing shield us away from any these type of supply chain shocks. That works very well for us. We don't have our own infrastructure. We are a true SaaS company that's asset light, so 88% of our revenue is recurring.
Outside of that SaaS, that 8%, 7%-8%, that's term license, but that's still subscription. The only terming is term licensing is really just a revenue recognition, a distinction on accounting side. If customer uses any non-cloud assets, we have to call it some term, and this happen often with government customers.
Okay.
Yeah. From that perspective, we feel that we have advantage where we're true asset light. We have this visibility into our cost management. Of course, increasingly when we also now use tokens, that's something that we are actively managing when it comes to token optimization.
Yeah. Got it. Okay. I want to shift gears to your Control Suite or your-
Yeah
governance security solutions, because this is an area where I feel like even sometimes myself, I get a little confused, right?
Right.
Just help us understand, like where exactly you play there. Are you a data governance-
Yeah
player? Are you a data security player? Are you both?
Yeah.
Who do you compete with? What's your competitive advantage? Why do you win there?
We have come from a enterprise content management space. In the old days, you think about OpenText, Documentum, right? Of course, Microsoft view themself as that as well. That's our core DNA. What that mean is we have to care about the entire life cycle of unstructured data, whether it's your emails, your chats, your contracts, your internet portals, your project, et cetera, a ll the workloads in your office. That means going from data ingestion, right, data creation or data migration, classification, tagging, workloads, workflows, to then resiliency, backup recovery, ransomware detection and recovery, and then to then governance. That's life cycle management. How long does this data asset live for? How long does this chat channel live for? How long does this data room live for? We even have a data room product in that regard.
How long does this project live for internally, externally? When does access begin, and when does access shut off?
Yeah.
When this person move from department to department, who do then we recertify this asset, right? All of these things that we do pre-GenAI days now matter even more in GenAI days. Because what AI does is actually exemplify, accelerate, right, all these issues at a much faster pace because AI is always running. This is why what we do today matters to our customers, because we do the end-to-end. We don't just do governance. We don't just monitor access control and do remediation. We also recover for you if you actually, AI have damage or external bad actor have damage, or internal bad actors leak data. That's why we say that from data ingestion, data integration, data migration is basically another word for data movement. Data will never stop moving from platform to platform, from tenant to tenant. Think about acquisitions, divestitures.
Think about moving from Google to Microsoft, from Box to Microsoft to Google, from Salesforce to other CRM systems. We think all of this holistically as one problem set. Thanks to AI, all of these areas are converging. We use d to have very different players in different space, competing competitors. In migration, people who only do migration. In backup, who only do backup. In governance, who only do governance. Increasingly, that's converging, and we see that, and we are the first mover in that space. It's good to see that industry recognize this different aspect of the same problem set. Right? You see that some of the backup players are now going towards, last few years, security. They're buying security companies to get into the Data Security Posture Management that Gartner talk about, which we're in as well.
Now this year, thanks to the AI costs and AI agentic era of shadow AI, now all these guys all start talking about governance. What do they mean by governance? They don't really have a robust end-to-end, highly scalable to hundreds of thousands employee governance framework that we have always been running for 10 plus years. Right? For example, when Deloitte deploy Copilot for Teams, they first deploy our governance solution for Teams first, so that the Copilot in Teams doesn't recommend things to you that you're not supposed to have access to. Regular industry care about this. Banks care about this. Government care about this. Then also when some other very large organization, they want to discover the tens of thousands of agents running in their environment, want to bring them under control, they come to us. Right?
We operate at that level and at scale. Interestingly, we also take that capabilities and sell that capability to small businesses via managed service providers. We're able to both attack the large enterprise through enterprise-grade software, but also offer them to small businesses. This is where we see we have advantage, a segmentation advantage, a global presence advantage, and also this end-to-end thing that we truly do what we say and already have all the aspects ready for this agentic era.
Yeah. You do have every company that you compete with coming out with this AI discovery-
That's right.
governance.
Everyone's now coming out with the same language.
Right. your guys' AvePoint AgentPulse capability that GA'd several months ago, tell us about that, why it's differentiated?
Yeah
Why you have a right to win there?
So-
We got a second part on that.
Perfect.
Go ahead.
AvePoint AgentPulse only GA'd Q1 this year.
Right? Not seven months ago. It's like three months ago.
Several, I think it's several.
Several.
Sorry.
Yeah, yeah. Several month ago.
Yeah, yeah.
We did do preview since end of last year. It's a very robust set of cloud capability that allow customers to discover whatever that's running in their environment, whether it's cloud, Gemini, or Copilot agents, and then bring them under control. What do we mean by that? By basically controlling what kind of access these agents will have, what system they're allowed to t alk to, what kind of data they have access rights to, an d also at the same time, alert customers on their consumption, their token consumption, and shut them down if they need to. The difference of our governance framework that's highly scalable to hundreds of thousands of employees is that we actually do this concept of delegated administration. We allow business users to add context to the data and to the agents that they have. Because IT doesn't know what business users use things for.
IT always afraid to shut things down. Right? IT always become the last layer of defense when it comes to costs, when it comes to some of the misclassification. Our governance SaaS framework has been running for 10+ years for some of the biggest organization and for governments of the world. That's our differentiation, and that's our robustness. Of course, no w that applies to the agentic era in spades. Right? Agents, I was actually just sitting down with the CISO of Temasek two weeks ago in Singapore. She said this is priority of priorities, right? The hair on fire problem for all large enterprise CISOs and even board level conversation is one of AI risk, and the second is AI cost management. Because shadow AI is happening everywhere. Everyone's spinning off their own little agents.
Even people using little hard drives to spin off open cloud, right? Because now we got into the device management, we can then go to device level to uncover and discover if they're running agents locally, right? That's also important to monitor and manage. This is why we feel we have the right to claim that AI trust layer in the AI infrastructure story, because we have truly all the capabilities. There's stil l some gaps I want to fill, especially when it comes to multi-cloud. We think that we actually, obviously everyone's moving, accelerating at the speed of AI. Two-thirds of our employee population are dev. We're moving very quickly. You will see a lot more product releases coming out of AvePoint in the next few month and quarters to really lean into that whole AI trust layer story.
Mm-hmm. The comment you guys made at AgentPulse, I think on this most recent earnings call, was that 50% of the Control Suite pipeline, Control Suite I believe is about 40% of your total pipeline.
Yeah.
50% of that pipeline is now for bundles which include the AvePoint AgentPulse capability.
Yes.
I believe AvePoint AgentPulse is still not something you're directly monetizing. How do you envision it impacting numbers?
We do monetize it in the sense that in order to get AgentPulse, we have the good, better and best bundle. You have to buy the best bundle, right? That then gravitates, allow us to elevate the deal sizes for Control. When we talk about 40% of the pipeline being Control, that's already a market step up because you look at since last year
Right
our blended revenue contribution from Control is 26%, right? Resiliency is close to 60%. What we see is the trajectory to have far more Control mix, in our customer sets, our partner sets, and within that there's a very high inc lusion of AgentPulse, which elevates the deal sizes for Control. The demand for AgentPulse is quite high. We're actually already looking at, I can say this because my CFO started to say this, to specifically skew that
on its own and sell that on its own.
Yeah. Got it. Okay. Your guys' guidance this year calls for organic ARR growth FX adjusted to accelerate from 23% Q1 to 26%-
at year-end.
Yep.
What's informing that? What's giving you confidence in it? I think some investors look at that, it's a steep ramp-
Yep
in net new ARR growth, especially in the second half.
Yep.
Talk about what drives the achievement of that.
First of all, we completed 13 quarters of outperformance. That means we say we're going to do something, and we did it, and a little better, right? Then 12 consecutive quarter of double-digit new ARR add. That should give people confidence that if we say we're going to do something this year, we'll do it. Now, why is there seemingly to say that high level of confidence for second half accele ration? It's one, our business is seasonal, because we have a large enterprise component, we have a large government component. Q2 is bigger than Q1. Second half is bigger than first half, and Q3 is our public sector year-end, U.S. public sector year-end, and Q4 is most of our large commercial enterprise year-end. Last year our public sector is pretty flat, and we actually even call it out, right?
Beginning of last year, we say, "Hey, we see the noise in the space. We anticipate a 2% ARR hit," and we were correct, right? We give the guidance, and we also hit and beat the guidance. This year, just that reverting back itself will give you that accelerator. Just revert back to mean, right? Things going back to normal. Public sector will grow again, not like last year. That itself goes, and then also obviously we have the natural seasonality across other sides of business. This is why second half will be bigger than first half.
Has always been.
Yeah. Just remind us, I guess, what percent of ARR is public, excuse me, U.S. Fed, and then are you expecting U.S. Fed growth to outpace total ARR growth this year? Like, what's the gro wth expectation in federal this year?
I think U.S. Fed is mid to high single digits of the total ARR.
Yeah.
In terms of specific growth, it probably will outpace almost just because the comp is, I think, benefiting us obviously, because there really wasn't that much growth last year. Historically, I think we've seen, in Q3, it grow in line with the North America ARR growth. I think, this year, just given the dynamics that we're talking about, that we would expect to potentially outpace the total.
Got it. Gross and net retention rates, both have been improving the last several years. Still though, gross retention, 89%, FX adjusted, 91% if you adjust for that headroom from the migrations business-
Yep
which of course is not permanent recurring business.
Right.
Still lower than a lot of SaaS companies or at least what people would consider top-tier SaaS companies.
I know that the long-term target is 90% + in gross margins.
Yep.
When do you think you can achieve that, and what's more of the realistic ceiling if we're thinking about that 2029 target model? Can it get to 91%, 92% before adjusting for that two-point headroom from the migrations?
That's definitely our goal. That's our stated goal for our medium to long-term trajectory. We do have a fair amount of SMB business, right? That's contributed to the churn. In our large enterprise segment, we already exceeded those goals. O f the now 91% GR outsid e of the migration, of that 9% churn, if you will, half of it is actually down sell. That's effectively either seat reduction or product red uction. Half of it, 4.5%, is actually logo churn, and we see that mostly in the mid segment and the SMB segment. The way we remediate that is in the SMB segment, increasingly, we're giving more and more business to the MSPs, managed service providers. They become our end customers, and they also become SMB aggregators. This way we don't touch any SMB customers directly. They're managed by these managed service providers.
They're basically outsource IT. They have managed services, and they love us because for every dollar they spend on our platform, they gain $5 managed services revenue. That also alleviates this churn, right? They're addin g seat cou nts because they're adding co mpanies, and there's also MSP roll-up in the PE world that's happening. That's growing nicely. That's SMB, true SMB, less than 500 employees is already just about 20% are recurring. We think that has the opportunity to get to 30%, if not more. Because for Microsoft, that's actually 40% of their revenue.
For mid-size companies, we still use channel to sell to them, and the benefit of channel is that we can scale quickly, but the downside of channel is that you don't have too much of a direct relationship, you could have risk of churn. There we're a ctually looking at different AI-enabled, because they're still not that large, we cannot afford to have humans to cover them. We're l ooking at different AI-en abled early warning systems to be able to enable our partners, to give them, saying, "Hey, this customer, they have gone quiet," or, "Their activity has become lower. You should take a look at it," or, "We should go take a look at it." We're introducing more smarts there to make sure that we reduce that churn there. Lastly, is continue fast-paced development and roll-out of new capabilities.
That ultimately is what's going to improve as well because we need to matter, right? We need to be relevant to the customers, to the partners. You will see a lot more capabilities coming from us, in that regard, to be a even more critical player in their whole AI journey.
Right. You brought up seats, so I do want to ask on that because I haven't yet.
Yep.
There's a lot of investor fears around seat -based models, not only on seat reductions within the tech vertical, but also potential pricing pressures on a per-seat basis.
Yeah
You guys having to evolve your pricing to a value-based or outcome-based pricing model.
Yeah.
on both of those, I guess, what percent of your seats come from the tech vertical today, and how are seat counts in general trending at renewal?
Yep.
second piece is just on the pricing point, how should we think about your pricing model going forward? Will it evolve from seat-based to something different?
in general, we don't see major seat reductions across our customers. Tech is about 8%, less than 8% of our recurring, and only we see big tech shedding seats, but that's more of an artifact they overhired in the last few years. I think last year, the hype was overdone, AI's going to take over people's jobs and seat reduction, and then all software companies are going to be gone. I think everyone now realize that is o verdone, especially now we don't have enough token s, we don't have enough chips, we don't have enough everything. turn out that AI could be, a very good AI could be more expensive than person. Right? That actually focused companies to really look at responsible, cost-effective AI deployment. We can no longer do token maxing anymore, right?
Last year, all that experimentation is underwritten by the frontier model companies. This year, they have to actually make money. That's not happening anymore. We don't see seat count reduction as a major trend. Having said that, we follow the hyperscalers who are market makers. On the productivity side, like email, it's seat-based for every employee that has email. On the compute side, we call IaaS, PaaS, Infrastructure as a Service, Platform as a Service, that's consumption-based. Think Azure, think AWS, think GCP. Compute is think of Office Cloud, think of Sorry, productivity is think of Office Cloud, think of Salesforce, think of Google Workspace. Those are seat-based. now increasingly with agents run on compute cloud, and agents more consumption-based.
Yeah.
We already have a mixed license type. Lastly, you talk about outcome as a service. We do have a 12% revenue mix that's services. Previously, we say, "Hey, that's going to go down to under 10%." Increasingly, it's no surprise you see that Anthropic and OpenAI all started service businesses, right? Funny enough, you need humans to land AI. You need humans to be on-site, like the FDE model that Palantir does, to understand customer needs, to understand customer-specific envir onment, uniqueness, and complexity, and then stit ch it together and continue to refine AI to work better for that customer. That is something that we have in, especially in Japan, in Singapore, we have a very robust services business that does $30 million projects even to generate new IP.
Now that capability, we can bring it to North America and EMEA as well, and set for our biggest customers to act as that lander for AI governance and AI trust. We call that outcome as a service type of capabilities. That also allows us to stay ahead of the game, follow where the puck is going, continue to innovate. We also are very unique as a product company that actually has a serious world-class services organization that will continue to deliver with our largest customers.
Got it. Last question, growth profitability balance from here. The last three years, you guys have sustained, very much to your credit, mid-20s ARR growth while taking EBIT margins, I believe, from -1% in 2022 to 19% last year.
How should we think about that going forward? You're obviously taking more of a breather on EBIT margin expansion this year.
Yeah.
If you see opportunities to sustain or accelerate growth beyond what you're guiding to this year, should we expect more of a muted pace in EBIT margins, or are you fully committed to that 2029 target of 25%-30% EBIT?
Yeah, that target is something we're obviously still very much committed to. This year, as you point out, the guidance is expected to be flat for non-GAAP operating margins, so effectively about 19% versus last year, as we invest more in the business, as we've talked about. T he one thing I think that we are trying to emphasize more is also the fact that we are profitable on a GAAP basis, which we think mak es us certainly unique, especially for a company of our size. We manage our stock-based compensation pretty effectively, we think, which is now less than 10% of revenue. There's implied expansion of the GAAP margins this year. Then, to get to the 25%-30% longer term that we're talking about in 2029, I think, yeah, we would expect to see a resumption of the expansion taking place next year.
It doesn't need to. We have a very good sort of profitability infrastructure in place already where we can realize a lot of leverage from the sales and marketing line and G&A to get us up to the remaining 750 basis points, call it.
Yeah.
That will, I think, start to pick up again next year and resume over the next three years.
Got it. Okay. Well, thank you guys for the time. We'll go ahead and wrap it up there.
Thank you. Thanks, Rudy. Great hearing you.
Thank you.