Dynatrace, Inc. ($DT)

Welcome, everybody, and thank you for joining us. I'm [ Marty Vouin ]. This is [ Jane Patel ]. We are the JPMorgan equity research team covering enterprise software. And it's a great pleasure to be here with Rick McConnell, CEO; Jim Benson, CFO; and Dan Zugelder, CRO of Dynatrace. Rick, Jim Dan, really appreciate you guys coming back to TMC.

We'll start off on a good note. You guys finished fiscal 2026 surpassing $2 billion in ARR, fourth consecutive quarter of 16% ARR growth and a fiscal '27 guide that bakes in net new ARR acceleration. So it's been a banner year for you guys. But then just to start off, why don't you guys give a quick introduction and just let us know what Dynatrace is solving for the world's enterprises today and how that's evolving in the AI world?

Sure. DynaTrace is in the $75 billion plus observability space. Observability is targeted at enabling enterprises to run their software better. So think about us as helping to prevent incidents to remediate incidents and to optimize your environment. This is what observability software generally does. We often think of Dynatrace's vision in delivering the world software that works perfectly. So that would be our objective. And obviously, as we evolve so rapidly into an agentic world, this brings a huge array of new opportunities to the observability space. Because agents enable us to actually take action and to execute those actions in a more autonomous way. So you can imagine systems that are self-healing effectively to enable software to always work better than it's worked before. So I'm sure we'll talk more about this, but it is a very, very exciting space that we believe is a is a tailwind or receives a tailwind from all that is happening in AI based on the expansion of workloads to drive towards autonomous operations and in general, the necessity of delivering software that works perfectly.

That's great. And I think the self-healing part is a good way to describe it. I'm sure we'll touch on it, but the more that AI accelerated CoGen, I think it's made that observability even more important and critical and something that you need a third party can do it excellently to do it. But let's just start with the fiscal 2027 guide. Jim, if we think about the nuances from the earnings call last week, the guide implies net new ARR growth of 16% to 23%, accelerating from 12% in FY '26, reported ARR of 15.5% to 16.5%. That range from 16% to 23% of the ARR, it's wide. Can you give us a sense of what drives the high end versus the low end of that and what you can control in that equation?

Yes. I'll start with fiscal '26 was a year first. You mentioned in your opening remarks that we had consistent ARR growth of 16%. So we grew net new ARR at 12% in fiscal '26. We haven't done that in 3 years. So one of the things I talked to investors about is the phases that we've been going through. Fiscal '25 was a fixed year where we did some things on the go-to-market side. Fiscal '26 was stabilized and fiscal '27 was accelerated. So I've talked about that over time. We mentioned logs. We -- Dan will probably talk about it in a bit around platform consolidation, end-to-end observability is certainly something that more and more enterprises are looking to do. So that is certainly a tailwind. So relative to the range, I think what you see is that at the low end of the range, it still implies an acceleration of net new ARR growth of 16% and you mentioned 23% on the high end. Some people have asked me, they said, geez, Q4 landed a little bit lighter than maybe people would have thought. And I would say it did land a little bit lighter than our internal expectations. And I'd say largely that is we had softness in our EMEA market, notably in the Middle East. The difference between growing double digits and growing 9% is a few million dollars. So when I look to fiscal '26, there's a lot of momentum building in the business. And we expect it to continue. All the things that I just mentioned will continue. Relative to a range, I mean, our range is usually 1 point, plus or minus. So I think what you get at the high end of the range is a continuation of what we've been doing and building. Even at the lower end of the range, you're getting that. I think you know that historically, I've used the word prudent. that was called out on the earnings call. This guide is no different than any other guide I provided every year. So there is an internal plan and then there is a haircut for what we do that we're going to tell the investor community. And it's no different than what it always is. So we certainly have line of sight to doing better than the high end of the guide. Obviously, we got to see how the year progresses. I mentioned in the call that we would be a little bit more half 1 weighted. And just to clarify that, our historical linearity is like 44% of our net new ARR in the first half, 56% in the second half. That doesn't mean we're going to do more net new ARR in the first half. It just means that we'll do a couple of points more, maybe 46%, 54%. And it's because we have strong forecasted pipeline entering the year. And our pipeline visibility is always tighter in a 3- to 6-month window than it is 12-month window. So we actually just believe there's just a lot of tailwinds. And as the year progresses, we'll update our outlook accordingly.

Yes. And just to be 100% clear, your guidance philosophy essentially is unchanged from what you guys have done historically?

That's right.

Yes. Yes. If we just double-click on that 9% net new ARR growth in Q4 and then going forward, I know you mentioned there was some softness in the Middle East. I assume the Iran conflict doesn't help that situation. But if you can just talk about okay, if you have those slight headwinds, out of those tailwinds that you're seeing, which ones do you think are going to be able to drive a lot of that? What should give investors confidence in what we're seeing through the rest of the year?

I think the biggest one is I think we will continue to do what we did with logs. Our expectation is that logs will continue to double. So we ended the year well over $100 million in our -- we have a lot of confidence that we will continue that trajectory that we have the right product with the right pricing and packaging at the right time. And I think we've shown it in and ability to go into our installed base and sell it and actually even introduce new logos. So I think logs will be a big source of it. I do think we'll continue to see workloads grow with end-to-end observability. More existing customers and new customers looking to consolidate. One, they can save them money. You consolidate multiple vendors and you could save money just through software costs. Beyond that, we allow their environment to run more efficiently because once you're on Dynatrace, you can -- you have a platform where you have deterministic answers to figure out issues as opposed to people chasing alerts and dashboards. And so it really is just a -- this is not like new plays. It really is and it's a continuation. Dan can talk about the go-to-market model is largely unchanged. We're going to go down with our strategic accounts. We've had Global 500. We'll probably go down another 100, 150. We'll have 4 to 5 reps in that part of the pyramid. So you'll have fewer accounts per rep. We'll get better penetration. The penetration that we've had in those accounts since Dan made those changes, the fastest-growing segment that we've had in the company. And it's about running that play again. And we think that, that will continue to be another source of productivity improvement.

Yes, and having the continuation of the existing tailwinds feels a little bit more comfortable than anything new having to kind of layer in.

One other thing that I would mention is that you know that we've been at the Dynatrace platform subscription for a while. And so we now have over 75% of our ARR on that contracting vehicle, a little over 60% of our customers. Fiscal '27 is the first year where you're actually going to go through three annual reset cycles. We haven't had that in the past. And the first year of EPS was fiscal '24. Most of our DPS customers have 3-year cohort classes. They are coming up for their actual renewal in fiscal '27. So there's a significant percentage of our installed base with DPS that is going to go through either a renewal or an annual reset. It's a huge opportunity for an expansion for us for all the reasons we just said. Consumption is growing at a very rapid clip. It's been growing north of 20% for well over 4 quarters. And if we can continue with that. Again, it will be a source of expansion opportunity.

Yes. Yes. It sounds like a lot of upward pressures in the right direction. Touching on some of the other things you were talking about, Dan, I think this is your wheelhouse. Dynatrace, you guys started this go-to-market transformation a couple of years ago, shifting reps to focus on 4 or 5 accounts in the Fortune 500, building the partner motion, comp plan changes. And as Jim said, you just finished kind of the stabilization phase. So when you're looking at that pipeline coverage entering FY '27 is healthy enough to guide net new ARR modestly to accelerate. One of the themes in our partner work is that the pipeline quality matters more than the quantity. I assume you guys kind of have the same perspective. Can you talk about what you're seeing in terms of that quality and how you measure it? Is it the stage progression, deal velocity, close rates? What is giving you confidence in that pipeline?

Yes. I think it probably starts with the methodology and the inspection, the quality of in, what goes in, the approach we take with that. But then the analytics play a role as well. The motion for Dynatrace, if you go back a few years ago, was a very different selling motion regarding pipeline. It was a very mid-level enterprise engagement, typically very feature-based. And over the last 3 years, it's transformed into a much more enterprise value selling motion. So we have a different approach, a pretty dramatic different approach. And that's matured at this point. So we understand what goes into our pipeline. We understand the risk of what matures, what doesn't, at what pace. So I think it's just the maturity of the motion that gives us the confidence in the pipeline.

That's great. I want to turn it over to [ Jayden ] here. But Rick, complementing on that enterprise motion on the shift to the left, developer side, you got DevCycle, MCP server, cloud code, cursor. Is this an evolution of the enterprise motion or a different velocity engine kind of targeted AI natives and the AI native buyers?

Great question. Definitely the power of the and. I learned this from my earlier career in Cisco, and we were always talking about the power of the and. This is definitely it. We don't foresee a shift from the enterprise to AI natives, for example. We believe that it is the enterprise plus AI native. We have been working very diligently over the last 18 months to build developer capabilities into the platform. And you've seen those over the last couple of quarters through a myriad of announcements related to integrations, for example, into AWS, Bedrock Agent Core into GitHub Copilot into Cloud Code into a myriad of different other elements. All of these enable our platform to be more developer ready. And we didn't grow up on the developer side like some others in our industry. We grew up in the enterprise side, focused on IT ops selling to the CXO. And this is why out of our customer base, we have an average selling price of now about $500,000. It's because they are large end-to-end enterprise deployments. And they -- these customers use us to make sure that their software always works. And they want to deliver end-to-end observability at multiple levels. completely integrated data lake house, integrated fully across logs, applications, infrastructure, real user monitoring, et cetera. and then integrated for all personas. And this is what we do better than anybody. This has done a traces superpower. We want to take that superpower that we've delivered to the enterprise and bring that to the developer community, bring that to AI natives. And that really is the next step in our evolution as an organization.

Great. Dan, I think this one is for you. Let's talk a little bit about the disconnect between what we hear from investors and what customers or what you're seeing with customers? One consistent theme we hear from investors that customers are more reluctant to spend because of all the uncertainties due to AI. A lot of this discussion centers on the application side, but I think it's worth touching on here. Dan, I assume you're the one across the table from CIOs and CROs every week. What do customer conversations actually sound like right now versus what we're hearing from the investor community?

Well, I think there has been -- I don't know if it's new, there is cost concerns. People are looking to -- and that's why consolidation has been such a play is that -- no question it gets people's attention when you're saying I'm taking two or three tools and making them into one, and there's going to be a 10% to 15%, 20% cost savings. So there's an attractiveness of that conversation. When you can couple that with saying I'm going to give you a better outcome, that's usually the combination that people are looking for. So we're not looking just to do things to cut costs, but if we can deliver a better observability outcome. So I think as we enter AI, and we just came out of our sales kickoff, I reiterated that our motion of tool consolidation cost out and logs is a great example of that, where we sometimes take 40%, 50% cost out because of combining logs with metrics and traces, you just need less of them in a more efficient way. So that -- it does resonate with our customers. There's no question. It has been, and I think it continues to be -- get people's attention. What you do in this world that's been created with AI is a fantastic tailwind because you're creating then budgeted new projects their AI projects that need observability. So you continue to run your entire play around tool consolidation, the enterprise end-to-end observability. And now we have a motion that we've added that is more about saying, going after these specific AI projects that need observability. So you have a little bit of a bespoke process in go-to-market, and then you have our very mature enterprise market. So that's created additional opportunity for us. And then Rick talked about the AI natives to have their own specific market as well. So AI has created additional markets and for observability for sure.

The one thing I would add, just to append to Dan's remarks is, we absolutely believe that observability is a beneficiary of AI evolution of AI-first organizations. It is critical in our view that in an AI-first world, you need more observability, not less. We do not see observability as a market and infrastructure environment, which is going to get disintermediated by AI. Quite the contrary, we see that in our existing customer base, maybe 30% or so of workloads are actually observed in a sophisticated way like using Dynatrace in a probabilistic world where you're using lots of LOMs, agents to build software, we believe that you actually need more observability to oversee those workloads than not, number one. Number two, we bring incredible contacts, billions of interconnected data points in real time, very specific to a particular organizations environment. And number three, we bring domain expertise of understanding that environment specifically and how that environment is operating. So observability is going to benefit from a tailwind in AI, not just in deployment of AI natives, but also in the enterprise itself, where obviously, enterprises are deploying more and more AI workloads that are inclusive of AI, largely built by AI, coded by AI and over time, will be operated by AI in a very agentic world headed to an autonomous environment. All of these factors are driving an environment for observability that has become more and more mission-critical to organizations day by day. And of course, it's our commitment to do everything we can to be a beneficiary of this trend for observability overall at Dynatrace.

Yes. It seems like there's a lot of AI tailwinds that are benefiting the business on top of the execution that you guys are providing. Jim, on that point, when do you expect these AI tailwinds to meaningfully contribute to accelerated growth?

Well, I think we're already starting to see it. But you got to remember that the way our model works is you're signing up customers on a contract with DPS and then they have 3, 1-year commitments and then they push consumption. So the underlying component of DPS is consumption. It's not a seat-based model, it's a consumption model. Now we recognize revenue ratably but we consume based on what customers are -- what workloads customers are adding and the growth in those workloads. So we're already seeing some of the benefits that -- again, I mentioned that it's been over 4 quarters that consumption is growing north of 20%. And the way to think about that is that -- and I think this is where people sometimes struggle is why is ARR not growing 20%. And there's a lag. There is a lag with just the nature of the way the DPS contracting model works. But if you can continue to grow consumption north of 20%, you will see a convergence of ARR and consumption over time.

And if I our selling motion is just as much about bringing workloads onto the platform as it is expansion of ARR because it's it's the preemptive piece of expansion on ARR. So our salespeople are out there just trying to get workloads on to Dynatrace, we they're AI workloads, whether they're logs, whatever expanding to more applications in general. Their motion is like, I know if I get more consumption -- when that DPS comes for renewal, that will give the fuel to that fire to be able to expand more. So you can understand how we are very consumption-based even though we're an or an ACV and ARR as far as the way we measure the business, the focus is now -- is all on consumption.

The other thing that I would say is that we have integrated account teams that Dan's account reps are the quarterback for us. So we have customer success teams, they are measured on consumption. They are compensated on consumption. Dan has bespoke strike teams or bespoke product areas. They are measured on consumption. They are compensated on consumption. So this consumption mindset is -- exists throughout the company. And I'd say we continue to run plays now. There's going to be plays on consumption. So again, at its core, get them on the DPS platform, have teams of people work with them to get more value to Dan's point, reps ensuring they find new workloads. We've got customer success teams that are helping ensure that they're getting value out of the different product areas. And then specifically on the strike teams, they're both -- they're on the front end with sales on deals, and they're also on the consumption side around driving more consumption. So it's -- I'd say we have a lot of the ingredients in place. They've been in place now for a year. I think it's led to what has been stabilization, and we're quite confident that it will now lead to acceleration.

Great. Let's tackle sort of the bear case on observability head on. Rick, this question is for you.

I must give the bear case to Jim. I'm just kidding.

The question that comes up most often on our end is this outstanding idea that as the cost of code goes to 0, customers can use open telemetry plus Vibe coding to roll their own observability layer and an LLM can reason over that telemetry. We've written tens of times that we think this thesis is off base, but better to hear from you than anyone else. What is your thoughts on that? And why is that true or not true?

Yes. I mean I sort of gave the precursor to this answer already. So I can -- I think I can do it briefly. But -- the starting point is I sort of oversimplify to bifurcate the world into application seat-based software with standardized workflows and highly dynamic software -- infrastructure software that needs to take real-time data into account. These are very, very different models. And in the former case, it is much easier to [ VIB ] code a standardized workflow than a highly dynamic workflow based on context. And that is sort of the underlying thesis and our view is the combination of significant domain expertise, highly specialized for organizations like JPMorgan, for example, or the largest organizations on the planet to be able to integrate and observe their environments is absolutely nation critical to those organizations and are you really going to rely on a vibe-coded standardized piece of software. It doesn't have the same degree of domain expertise. Doesn't have the same degree of real-time context based on billions of interconnected data points. Doesn't have the ability to oversee or observe probabilistic models and by the way, there's an even added element here. For the history of Dynatrace, we have been focused on a particular question or answering a particular question for our customers. And that question is, is it working? If you think about what does observability do, is it working or maybe is it working well? Is it optimized? That's what we've been focused on in an LLM world of models that are providing input to allow trustworthy actions to take place through agents, you have to answer a separate question. And that is, is it correct -- in other words, is the information coming from models actually accurate so that it can be trusted and implemented. We have customers that are using us in payment terminals that have to work every time. They need agents to be able to take immediate action to provide corrective action to make them work. We have banks who are using them in mobile using Dynatrace as part of mobile apps to assist customers to transfer money from one account to another. That's got to work every single time, whether it's health care, manufacturing, their use cases in each of these environments that are critical to be driving trust worthy outcomes based on AI and LLM inputs. And you simply can't rely upon an LLM to provide that degree of context, that degree of domain expertise to support an environment that then was created with a standard LiveCode. It just isn't going to work. And so -- and if there's one answer to all that, I would say, context is it.

Great. On that point that these tools need to just work, the place where that makes the most sense is the enterprise, right? So Dan, this one is for you. The most striking thing in your recent quarters is the consistency of 7-figure deals, $200,000-plus new logo lands and as you said, roughly $500,000 ARR per customer, which Jim has framed as having a path to $1 million plus over the long term. A partner told us that consolidation is the only conversation that matters right now. CIOs are tired of paying for 5 tools that overlap. Is that sort of the right read on buyer mindset right now?

Consolidation is definitely on their top of mind. But I mentioned this before, it is obviously cost. It obviously has to do with if you are a C-suite, up to the CEO of organizations, and there is a major outage. CEO will be -- the first question is, what's the problem? Obviously, that I/O, if they cannot answer that question very quickly, their job is on the line. So that is -- when you have a fragmented observability stack, it means that, that answer becomes that much more difficult. because everybody is looking at their individual tools and saying, I'm good. And they're asking the question, they're going around the room and people are answering, I'm good, I'm good. And somebody have stopped saying, but we have a major outage going on right now. So we deliver that so you have visibility across your entire application and infrastructure stack. So the consolidation plays on a cost basis. There's no question it, but it plays on delivering a better outcome. I think that trend is not going away. It's a bit -- some of the C-suite I speak to all the time is that they're just trying to continue to pull things together, pull optimize costs but deliver a better outcome. That is not -- that's been there. I think it's catching when. I think there's a lot more. Logs have played a big role in that, that if you look at it, we were 18 months ago, we really weren't playing in the logs business. So that's accelerated this view of a better outcome. Most people are using another -- some of the traditional log providers, and that was redundant to observability. So that was actually -- it's fairly low-hanging fruit for them.

I would just add that -- it seems to me that the holy rail that Dan and I hear from CIOs, CTOs, CXOs basically every day as they want to get to autonomous operations. They want to get to autonomous operations because the cost savings are extraordinary. And by the way, they're having more difficulty finding head count to actually manage environments rather than less. So they need to be able to do more on an automated basis. By the way, to the extent that, that can be predictive, it's even better. so that you eliminate issues before they even occur. In order to get to autonomous operations, you have to have to Dan's point, end-to-end observability because it is the confluence of elements that enables you to have confidence and trust in the outcome that enables an agent to take action. And if you don't have that degree of confidence in trust, you cannot rely on the agent to take the correct action. So the mechanism to get to autonomous operations from accounts is to begin with end-to-end observability. And this is why, last quarter, for example, 22 deals of greater than $1 million, which is pretty substantial. For run rate, all driven by really this end-to-end observability motion, all setting up not as the end but as the mechanism to drive toward autonomous operations as we look at.

You mentioned logs earlier and a bit on this last question as well. Logs has definitely been one of the biggest surprises, positive surprises in the Dynatrace story over the last 12 months with over $100 million of annualized consumption, which growing north of 100% year-over-year. Rick, you were out there quoted in the ether around the $250 million number, which we interpreted as aspirational. The underlying question here is, let's say we pass that $250 million, where can this business grow? What is the overall market here that you can drive this to be?

I'll start and look to Dan and Jim to comment further. But I mean, the logs business alone is a multibillion-dollar market. I mean you look at Splunk, you look at others in this market, and it's already many, many billions of dollars of log observability -- and we believe that -- we absolutely believe that our business can continue to grow toward that $1 billion mark and over the course of time, even beyond that. And the reason is because we believe that we have a solution that is better than a stand-alone log solution. Stand-alone log solution is focused only on logs. What we can do, which is consistent with what we've been discussing here before, are two things related to logs. Number one, by integrating, as Dan was talking about, logs with traces, metrics, really user data, behavioral analytics, et cetera, you get a better outcome. you can see the entire environment by seeing the entire environment inclusive of logs, you deliver a better outcome. Those answers that come from that better outcome lead to the autonomous operations of the environment that I'm describing. The second thing is you actually -- because you have traces, metrics, logs, real user data, it actually is more beneficial to have multiple different data types than just more logs. So you can actually accumulate fewer logs, but when combined with these other data types, you actually end up in an environment where you get a much richer environment to be able to provide analytics that allow for things like auto prevention, auto remediation, auto optimization. And so better outcomes, lower price point and you have lots of our existing customers that are using a log vendor for logs us for observability, and they're increasingly moving that log workload to us in order to get the benefit of end-to-end observability that we've been talking about for all the reasons we've been discussing.

Great. I have one last question, and I can hand it off to [ Ari ]. Jim, this one is for you. We would be remiss not to ask about this. The Starboard team published a fairly detailed letter on April 28, and both of you and Starboard have described the engagement as constructive. Rather than dwell on the letter itself, help the audience understand how you're thinking about the value creation framework for Dynatrace over the next several years.

Yes. Well, I mean, we think of value creation beyond all investors. So that's kind of how we're viewing Dynatrace that we want to provide shareholder value for all. And I think at the end of the day, when you think about what we're trying to do, everything we tried to build and outline was an acceleration in the growth of the business. This business can and should be north of what it has been delivering. And so for us, there's been a big focus, again, going back to my fixed stabilized accelerate, putting things in place to go on the offensive to go after the opportunity. I think Dan talked about some of -- the model we had before was good for when the company got to $1 billion and we needed a different model to be able to scale. And so when I think about acceleration, I think it's in our sights. And you've seen that for sure, at the high end of our guide, we expect that, that will happen this year. On the margin front, we are a rare company that operates at 29% or 29%, almost 30% operating margins. We've driven 400 basis points of leverage in the model. In each year, we'll tweak it a little bit because sometimes you'll make investments in 1 year, you'll get a return the following year. So driving efficiency and driving leverage is always part of the story that Dynatrace is a balanced growth and profitability story and it's a sequencing of when you would do that. So I think we've demonstrated that we can do that, and we will do that. I mentioned in the call that we're driving 150 basis points of operating expense leverage in fiscal '27. Offsetting what is the near-term gross margin headwind. So again, margin and leverage very important. And then we believe the shares are undervalued. We have significantly bought back stock we increased the pace of our buyback 40% from Q3 to Q4. You know that we doubled the authorization to $1 billion in February. And so we believe that at these prices, it's very attractive, and we will continue to put capital to use in the form of buyback. We have -- we generate a lot of cash flow. We have a fair amount of cash on the balance sheet. So accelerate growth continue to focus on efficiency and then put your buyback to use at very attractive values. And I think that's what we've been doing.

Yes. Well, Rick, Jim, Dan, it's been fantastic to have you guys here and sharing your insights with us. A lot of stuff going right in your business. We're really looking forward to where that goes. So thank you.

Thank you.