All right, good afternoon. I am Sanjit Singh. I cover infrastructure software for the Morgan Stanley Software Research Team. We're super excited to talk observability again, this time with the Dynatrace management team. Welcome to have again to the conference CEO, Rick McConnell, and CFO, Jim Benson. Rick, Jim, thank you for joining us once again this year. We really appreciate it.
Thanks for having us.
Awesome.
Thank you, Sanjit.
I don't have my disclosure sign, but, for important disclosures, go to www.morganstanley.com/researchdisclosures. So with that, I finally got off my butt and initiated on Dynatrace a couple weeks ago, and one of the things that we highlighted in that report was, that if you go back to 2018, Dynatrace is among the top two share gainers in the category, and the category, frankly, has a lot of players, right? And so, Rick, maybe just to level set the conversation, at a foundational level, what's the reason why Dynatrace has been able to stand out from the pack and be the share gainer that you have been, for going on six, seven years now?
There are really three, I would say, primary differentiators for Dynatrace in the market. First of all, we have an integrated platform, and that integrated platform delivers deep contextual analytics. Those contextual analytics enable us to provide answers, not just data, not just dashboards. Those answers really make the difference of radical reduction in number of incidents and radical reduction in the amount of time it takes to fix issues. The second area is what we call Hypermodal AI. With Hypermodal AI, for well over a decade, we've had Causal AI and Predictive AI. Causal AI is dedicated to addressing root cause analysis of issues as they occur. Predictive AI is focused on anticipating issues so that you can prevent them before they occur.
To those two, we are now in the process of adding Generative AI , which then will bring in natural language interface to the platform to bring the platform to more end users. It is the combination of these three AI techniques that really becomes the secret sauce of Dynatrace. And then third and finally, is really around automation, and if you can get to the point in an observability platform where you trust the answers that the platform's giving, then you can actually automate the response and do so directly in code. So if you have the analytics, the AI oversight, providing these answers, and then trust the answers to automate response, that's really the power of Dynatrace.
Yeah, it's quite, quite, quite compelling, compelling vision. Now, a few weeks ago, you held your Perform conference, and again, just to sort of level set everyone, what were your key takeaways coming out of the Perform conference? And when you engaged with your customers, what did you learn about your customers' investment plans, for 2024?
I would say... Well, first of all, our Perform conference was record number of people, 2,500 people or so in the audience, thousands of people online. It was it was an exceptional overall engagement. Great to hear directly from customers all the time. Good to see their passion. Excellent, of course, to see their commitment to the platform. I would say the biggest theme, biggest takeaway from my standpoint, was the fact that observability has moved from optional to mandatory. In some sense, you can almost abstract Dynatrace out of the equation of that first assessment.
It is simply the fact that workloads are becoming so complex, so big, that it is impossible to manage them manually, the way that you used to, with a Network Operations Center, with an army of people staring at a sea of glass to try to look for the needle in the haystack of red, yellow, green alerts. We've instead moved to the point where end users are not gonna wait. Their experience mandates that software work perfectly, and for software to work perfectly, you can't have incidents in the first place. So your ability to reduce incidents and radically reduce the amount of time it takes to repair them when they occur is quintessential, and that's what they're after. That's why we had so many people there. That's what they're looking for the Dynatrace platform to do for them.
Yeah, that's a pretty healthy backdrop to enter calendar 2024. One of the things that's going on in observability is platform consolidation, a theme that you guys talk quite a bit about, and I wanna spend a few minutes talking about the platform consolidation opportunity. You know, the context here is that enterprises, I think I cited a report where there's 70% of enterprise businesses are using five or more monitoring tools, a lot of them using 10 or more monitoring tools. So before we get into, like, you know, what you're doing to consolidate them, how did we get into this, such a fragmented, siloed market to begin with?
How did we get there? Well, it's the evolution of the complexity of software generally, I think that got us there. It was in the olden days, we back up far enough, that it was all mainframes or simple software structures, pretty simple to control end-to-end a piece of software. If you had a bug in it, you knew precisely where the bug occurred, and so you would attack it, fast or to present. In the present day, you have microservices, you have containers, hybrid cloud environments. You have you may not have changed a piece of code yourself. You may have simply recompiled a new set of libraries that created a bug or an issue in the system that then may lead to an incident or an outage....
That world has resulted in an explosion of data and a massive increase in its complexity, and an inability to manage your software the way you used to, in the way that you used to manage those workloads. So that's sort of how we, how we did it to get to this point, is that groups, individual departments or different parts of a larger organization would make independent decisions. I need a dashboard. I need to build this internally. I need to use open source code. I'm gonna choose this vendor or that vendor, and the next thing you know is you have tool sprawl. So what's the resolution of that? Resolution is eventually it becomes the case where you are not efficient in addressing the problem, not efficient in delivering software that works right, not effective in delivering great user experience.
On top of all that, it's costing you way more than it would if you consolidated those tools to a better answer. Our view of it is quite simply that tool sprawl is created by the environment that we live in today. Tool consolidation enables two major things: number one, radically improve user experience, and number two, at lower cost.
Awesome. When I look at the players of the market, they're all sort of trying to position themselves as the consolidator versus the consolidated. And so from, you know, you guys have highlighted a number of platform consolidation wins on the last few earnings calls, or several earnings calls. Why are customers choosing to consolidate with Dynatrace? Is it the productivity enhancements? Is it frankly just about lowering licensing costs overall? What's causing that demand to funnel to Dynatrace versus some of your competitors?
Well, let's, let's go back to the answer to question number one. By the way, you're gonna ask Jim a question here soon, right? Okay.
Yeah, I'll get it.
I just wanted, I just wanted to make sure.
Yeah.
So back to... I'm just joking. The first answer or the answer to the first question was really around contextual analytics, Hypermodal AI, and automation. All of these elements are sort of the alternative approach to manual dashboards and manual resolution of issues with dashboards that show you red, yellow, green, and you're trying to architect through figuring out where a problem is. The Dynatrace Platform is more and more and more differentiable. The bigger the data flow, the more data types you incorporate, the more complexity there is there in the system. So the result of it is, in environments where larger enterprises, which are continuing to expand their workloads, the consolidating to other tools that simply offer more dashboards and more manual resolution is sort of orthogonal or opposite to what you're trying to achieve.
Mm-hmm.
which is improved customer satisfaction at lower cost.
Right.
Because you need the resolution of those issues to become more efficient, which comes from elements like AI and automation-
Mm-hmm
... Not more dashboards, more data, more insights through an evaluation of those data types. This is why we have such a strong win-loss ratio in these consolidation type plays and such a big opportunity that I like to think that we really can become the primary consolidator of choice.
Awesome. Yeah. No, it makes a lot of sense. And then maybe to toggle the conversation to Jim, but staying on the same topic of platform consolidation, which has been, you know, quite the theme in the market. On the last earnings call, you know, the theme of the earnings call was, you have multiple platform consolidation opportunities. They're larger in size, making up more of the pipeline, but uncertainty on the timing of when they close, and it caused you to lower the guidance, the ARR guidance, by 100 basis points. Jim, I was wondering if you could take us out behind the curtains on speaking to these platform consolidation opportunities you do have in the pipeline, how are customers arriving at their decisions?
What are they sort of weighing, and what's driving any sort of hesitation on these deals closing?
Yeah.
Could you actually, actually speak to what changed maybe between Q2 and versus Q3?
So that's a great question. That's certainly a question I've gotten a lot from investors over the last, you know, several weeks. That, you know, you start with, has the demand environment changed fundamentally from kind of 90 days ago? The answer is, it hasn't. The demand environment is pretty healthy for observability. Has our pipeline changed? Our pipeline has not changed. Our pipeline is still very healthy. Our pipeline is growing at a rate and pace faster than our ARR growth. So demand environment healthy, pipeline healthy. The... I would say what we've seen within the pipeline, when you double-click on the pipeline, you know, as you can imagine, 90 days ago, your, your visibility of the pipeline is really good in a 90-day window, and it's pretty good in a 180-day window.
But the deals that are kind of 180 days, the timing variability then is a little bit less certain. And so what we've seen is we've seen more and more customers. You know, we used the stat on the call that we've seen a 39% increase in million-dollar ACV pipeline from Q4 of last year. And to remind folks, Q4 of last year was a huge, huge quarter for the company. So it was a record quarter for the company. So we're seeing a 39% increase in million-dollar ACV deals. And we double-click on them, many of them, not all of them, but many of them are the one. It could be many to few. And for all the reasons that Rick outlined, that the environments that they're dealing with are very, very complex.
You've got these disparate tools with disparate teams leveraging disparate tools. So when they're making a consolidation decision, going from many to few or many to one, and so there's just a lot of, in the whole go-to-market navigation with these customers, there's POCs that they're doing, and then you have to satisfy different constituencies within a customer that might be using different tool sets. And so what we have found is that it certainly does increase the variability of timing, of when you're gonna close the deal. And so, you know, we actually had some evidence, and the evidence was maybe a deal that we thought that might have had a close date of, say, January, that close date, say, moved to February. It might have been a deal that we thought was February, that moved to March.
We looked at it, and we said, "We just think it's smart, one, to be very transparent with investors about what we're seeing," because we actually think it's a net positive, Sanjit. We actually think it's good that if this emerging trend of tool and vendor consolidation, we actually think we are in a good position to benefit from that. So on the intermediate term, we said, this is actually a pretty good trend if this continues. Yes, in the near term, it does bring kinda short-term variability around when you think you can close something. So we thought it was prudent to increment that into the guide.
You know, and I—while we're not guiding for fiscal 2025, I have gotten questions from investors: "Well, if you see this continue, what do you think this might mean for fiscal 2025?" And I think what we'll have to do is we'll have to build in an incremental level of prudence, knowing that there's gonna be some uncertainty. But I actually think the general theme is positive for Dynatrace, that I think we are in a good position to benefit from that. And we'll talk about it, I'm sure, here in a bit, that some of the go-to-market changes that we're making, I think, are gonna help us better navigate those decisions with customers.
Yeah.
Hopefully that helps.
Absolutely. So on one side of the coin, you know, pricing is... timing is uncertain. We'll get to pricing later, but timing is uncertain. The other side of the coin is that these opportunities seem like what you guys were implying on the call, was these opportunities are consequential. I mean, these are real meaty opportunities, probably a nice seven-figure type deal. So if and when these things land, what do you—what—how should we think about the impact to things like ARR and net new ARR over time, as these deals start to close?
I would say that, one, I've used the phrase emerging trend because that, you know, I don't think it's appropriate to call it a kind of consistent trend because this is just there's always been a handful of tool consolidation deals in the past. The fact that we see a little bit more of them in our fourth quarter suggests that maybe they're gonna be a bit more prevalent. I do think that it will bring you know, near-term uncertainty as far as the timing of closing some of these things.
Mm-hmm.
I do think longer term, though, it. Because they are larger, and to the extent that we can benefit from closing them, I think in the longer term, they can be an accelerant, because to your point, your land sizes are greater. And just to be clear, these are not all new logos, right? So these are both existing customers and new customers that are just going through a journey of, you know, to your earlier discussion with Rick about these. How did we get here? Well, we got here because everyone used their own disparate tools. Finally, the CIO or some C-level decision maker said, "We need to make some architecture decisions," and they put someone in charge to do some of this. So I think the journey is not a journey that's gonna take three months or six months or nine months.
This is a journey that's gonna... You know, this is a multi-year journey for customers, and in some cases, customers are on the early side of it. But I do believe that with larger deal sizes should lead to an acceleration in the business. What's challenging for us to assess is the timing of that and the rate and pace of that. Because I think if you see more of these, you know, by the fact that they're a little bit longer, their sales cycles are longer, it's a little bit more difficult to judge the timing of those, and so we'll wanna build a level of variability. But again, I bring people back to-
Mm.
I actually think thematically, it's, it's a net positive.
Yeah.
The only thing that I will say is that I've gotten questions around, like, is this a change in your go-to-market that we're whale hunting? This is not a case of we've changed our model from land and then expand. The model still is very much land, and we've changed our model on the land side to make sure we're landing with the right profile of customer, because we have found that customers that land greater than $100,000 have a propensity to expand, more so than someone less than that. So it's still land and expand. So this tool consolidation that we're seeing is more customer-driven than it's been as a Dynatrace-driven, activity. So I'd say we're very positive about the intermediate term around as all these variables are a net positive for Dynatrace.
I think some of the go-to-market changes that we have a new CRO that's been in place since July. He's gonna introduce a bit more changes to get more focus on these very large... So if you think about it, we go after the Global 15,000 customer base, is what we've talked about. When you actually double-click on that, the total addressable spend is actually 70% weighted to call the top 3,000. And so one of the things that we're doing is, for our next fiscal year, we're gonna make a bit of a reallocation change around where sales is spending their time and getting more weighted to the top-of-the-pyramid enterprise accounts... that one we either haven't landed or we think the propensity to expand is greater.
So we're introducing, I think, some go-to-market adjustments and go-to-market focus that we will be in a better position on these tool consolidation opportunities as they continue to grow, and I think we'll be in a better position to navigate through them.
Is there anything on the go-to-market side that's coming online outside of the, you know, the focus more on that sort of top 3,000?
Well, I'll start, and then maybe Rick can offer a few other. So one is what I just said, that, so we're going from a model of at the top of the pyramid, we may have had, say, eight accounts per rep. The reality is that the rep was really only focusing on, say, four of those customers. So they had eight accounts, but they were really focusing on four. Well, if these are really big customers, and you're only focusing on four, not eight, you probably should be redistributing some of the four that they're not working on to other reps. So we're taking some reps that maybe were focused below that tier-
Mm-hmm.
And we're gonna get more focused. So we'll reduce the accounts per rep. Less accounts per rep means better focus with those. And then we're gonna go after the longer tail, call it to the Global 15,000 with a more of an inside sales motion. We have an inside sales motion today, but we'll just beef it up. That's one component. The other component is we've talked about channels and partners being an improved part of the model. And the good news is we continue to do more of our business through partners. What we need to do more of, though, is we need to get more deal origination from partners. So, in many cases today, a partner may be involved in the deal, but really the pipeline was generated by us.
We got to get to a model where we get better reach from the partners. So we hired a new channel leader about three months ago, and his focus, Sanjay, is to address this, which is getting better reach, and in particular, different aspects of the partner model, 'cause all- not all partners are created equal. You got GSIs, you got hyperscalers, and then you call it a regional channel partner. So there's a different model for each one. But the whole premise is to get the partner model to give you more reach, so you're getting more pipeline from the partners. The good news is today, they're involved in a lot of deals, and they may be helping us accelerate a deal, close a deal faster, maybe make the deal a little bit bigger.
But we want to get to a model where they're not just doing that, but they're also bringing more deals to the table with us, so we get, you know, can get bigger, bigger, bigger pipeline. Bigger pipeline will yield kind of an acceleration in the growth rates that we've seen. And, Rick, I don't know if you'd offer anything else.
I mean, I would, I would simply offer we're focused on scale and acceleration. And, and that means you go after the money where the money is, which tends to be in major accounts, enterprises, global accounts. And, and then the question is: How do you get there? And through, through channels such as global system integrators, you can move the deal for observability dramatically forward relative to closing it after the fact. So those are some of the themes we're driving.
Yeah, it makes total sense. I do want to spend some time talking about all the various product capabilities of the platform.
Mm-hmm.
-that kind of maybe sets you, sets the company up well in this age of AI. But sort of last question on this topic, just around, you know, maximizing the opportunity. I think the investor base, you know, appreciates Dynatrace for the profitability and the competition of growth.
Yep.
20% is probably a threshold of growth that people, you know, want to see out of Dynatrace. So in the context of like net new ARR being down for the last six quarters, you're guiding it to be down next quarter, and hey, now, these quarters, we could probably see a pretty nice inflection. But to get to that sustained sort of 20%, above 20% ARR, what needs to happen? As sort of a question on, like, how is the model set up? Do you need to put out a string of quarters to sustainably be above that 20% threshold?
Yeah, it's a great question, and I think just to, you know, to back up to your... kind of reiterate your point, that from an investor lens, the beauty of the Dynatrace model is you get both growth and profitability. Because we and oh, by the way, profitability and significant free cash flow generation, so much so that we, the government actually collects a lot of that in the way of cash taxes. So, but that's a net positive, that we have it's a very healthy model of both balanced growth and profitability. We are very mindful, Sanjit, that growth has moderated. Now, obviously, we've been dealing with a macro environment that certainly has not been an accelerant to growth in any case.
However, I think we're in this period of what got you to $1 billion, which is effectively where we were, is not how you get to $3 billion, $4 billion, $5 billion . So some of the changes that we're making on the go-to-market side, that I mentioned-
Mm-hmm.
Those are the, call it, the changes that we need to make to drive more consistency, which is what you're getting at. Driving more consistency in the model, building more pipeline, being able to get partners to get involved and help bring business to you and helping you close business in a more accelerated manner, that the product capabilities of the company are vast. What we're trying to tune is we're trying to tune the go-to-market model in a way that Rick uses the term flywheel with the GSIs. And whether it's with the GSIs or whether it's through a reweighting of focus from our reps to higher propensity to spend accounts, that there's a bunch of things that we're doing that we believe will lead to 20%+ growth.
Now, what I haven't said is that that's gonna happen in fiscal 2025-
Mm-hmm
... because I haven't guided yet. But I also believe these changes are gonna, they're gonna have to mature into the model before you see it. So I think we're in a great position from a product perspective. I think we're in a great position from a leadership perspective. We have new leaders that have seen growth at scale. They're trying to introduce changes for the company that,
... build muscle that probably we haven't had in the past. And so I'm very optimistic that we're in a good position to be able to capitalize what I, you know, on what I think is a really large growing market opportunity and get the company back to 20%+ ARR growth.
Yeah, that's, that's fantastic. So let's talk about the breadth of capabilities of the Dynatrace platform. Again, doing my work on the initiation, where I came away really impressed by Davis AI, the OneAgent architecture, Smartscape, the Grail data lakehouse, and the Automation Engine. And so with all of these capabilities already in place, and then we're working on layering Generative AI into the platform. Given this foundation, do you feel like you have a leg up on the competition as we head into this age of AI? Or does a new compute cycle force you to re-architect in some way, just like the movement towards microservices and Kubernetes, you know, force the company to rethink the architecture of the platform?
No, absolutely, no re-architecture of the platform needed. In fact, we just completed a re-architecture of the platform on Grail to deliver these sorts of capabilities, and it is the integration of each of these core technologies, Sanjit, as you defined it, that really established the foundation for Dynatrace.
Mm-hmm.
So, we're very happy with the state of the platform. We can evolve it in application security, evolve it in log management, evolve it in other areas too, such as Generative AI with Davis Copilot, to make the platform even more accessible. But each of these elements provide an opportunity for expansion of the existing architecture in an extraordinarily compelling way.
Mm-hmm.
So we're excited about where we are. In fact, I gave you the multiple differentiators at the beginning, but each of those really are represented by a completely unified platform. We have been very disciplined through the course of time, even post-acquisitions, to maintain the efficacy of that platform. Because it's really been critical to our customers to have a single underlying data store, where each of the modules access that data store through the same set of core technologies, because that's where the platform can really automate. If you have a series of underlying data stores-
Mm-hmm
... that are only aggregated at the user interface level, then you have a lot of manual tagging that has to do with data types. You're never gonna get automation out of that, and that's where customers wanna be.
Mm-hmm. You know, you mentioned that, like, in one of your answers at the top of the conversation, the concept of Hypermodal AI, which kind of sort of combines Predictive AI, Causal AI, and Generative AI in a unified platform. You know, with the hypermodal approach, how is Dynatrace's approach to AI different than your competitors, and how will this ultimately benefit customers and users?
The easy answer to the question is, nobody has causal and predictive AI. So the foundation for Hypermodal AI that we've had for over a decade is our ability to use AI intelligently against an integrated dataset to deliver causal AI for root cause analysis, and predictive AI for anticipating issues to reduce incidents or eliminate incidents before they occur. This is completely unique to the Dynatrace platform. Now, many others in the observability market have since said, "Well, we're adding Generative A I." But Generative AI against a dataset that is not empirical, that is not deterministic, is just accessing an uncertain data type through correlation, not causation. Not trying to get technical-
Mm-hmm
... but think about it this way: you could use ChatGPT or some gen AI engine, in concert with an LLM, against a data type that's wrong, and get a wonderfully worded, fabulously viewed incorrect answer. And the same thing is true with observability. If you apply a natural language interface to an erroneous dataset, you're gonna get the wrong answer, and you can't automate. Customers will not allow you to automate against answers that aren't certain. In our case, through causal and predictive AI, we know deterministically the answer to the question, what went wrong? Was it a router in this location? Was it storage? Was it this line of code? Wherever it might be, based on those AI analytics. And so the result of that is that you can trust the answer.
When you then apply Generative AI to it, you're applying it to a deterministic dataset.
Mm-hmm.
That's the difference.
Yeah. Let's do one more question on AI, and then we'll go to the audience and see if they have any questions for the management team. You know, Davis Copilot and additional monitoring capabilities for LLMs and the Generative AI stack, when should we expect a general availability of these capabilities? That's sort of, you know, the first part of the question. And how are you thinking about these as growth revenue opportunities in the context of, frankly, multiple opportunities that the platform provides from a revenue perspective?
So the Davis Copilot is gonna be in tech preview by the end of this month, and we expect it to be GA in our first fiscal quarter, which is second calendar quarter. So it is imminent in availability. So that's that's the first piece. In terms of the opportunity, I think there are several opportunities. As there are more Generative AI workloads, one of the things that I didn't answer on your Perform questions earlier, for our Perform conference, was that we announced AI Observability. AI Observability is an observability solution built into the Dynatrace platform to observe Generative AI workloads. So as more and more companies develop their own Generative AI workloads, we'll enable the observation of those. Why is that important?
It's very consistent with the story that I started at, at the beginning, which is explosion of data, increase in its complexity, inability to manage manually. If you think about Generative AI overall, what does it do? Radically increases productivity, which drives an ability to get more work done faster, which results in more data, more complexity, all of these elements that make manual oversight more difficult. So you're gonna need to be observing these workloads in the same way that you observe other workloads, but it may actually be creating or they may actually be creating data even faster than the workloads are being created today.
Excellent. Let's go out to see if there's any questions from the audience for the management team. Come up front and just wait for the microphone, please.
I'm gonna be a little specific. How do you see the competitive environment? Essentially, there's Datadog, New Relic, and now AppD, Splunk, Cisco, and in particular around the consolidation part of it?
Sure. I'll take that one. So, question is, what is the competitive environment, what I think about it? There are two main thoughts that go through my head when I hear this. The first one is, wow, observability is becoming more and more commonplace, more and more standard, and the market is noticing. You have private equity companies, you've got Cisco consolidating. There's a lot more activity. For those of you who flew in through, through SFO, you see Cisco banners talking about observability in the United Terminal. It's becoming more prevalent that observability matters. And this was back to my very first answer to your question of why does all this matter? It's because software working perfectly is critical, and that's what we're trying to deliver.
So it's at the overall first answer to the question is just that observability is becoming more and more common to think about. At the individual competitive level, I would say that all of this consolidation and activity creates an opportunity for Dynatrace. So that's my second thought, because many of these organizations are going through a bit of a chaotic transformation of their businesses that I believe represents an opportunity for Dynatrace to take advantage.
We'll go back to the audience for some questions, but I wanna spend a minute or two on the security opportunity, and I'll, it's a two-part question. I think at a base level, what gives a leading observability player like Dynatrace the license to-
Mm-hmm.
- provide security for the enterprise? And the second question is, you put out a $100 million milestone by the end of fiscal year 2025. Are we on track to hit that $100 million run rate?
So on the, on the first question of do we have license to go into application security, I would say absolutely, but with limited scope. We have been very disciplined in application security to addressing those spaces of security that are directly benefiting from observability data, like, for example, Vulnerability Analytics , like, for example, Runtime Application Protection . These are areas that without the underlying observability data, you can't be effective in delivering security solutions. Quick example, in the Log4j days of late 2021, early 2022, we offered Vulnerability Analytics . We had many of our customers come to us and say, "Oh, my God, I've got this vulnerability in Log4j libraries. I can't tell where I'm using it." We sit with OneAgent at runtime.
We know exactly where you're making Log4j library calls, so we could take thousands of calls that you could search for in text in your code and tell you the only place in the last 24 hours you actually called Log4j library was in these five places. So you could then directly focus on patching the library and the code in those five areas. Incredibly powerful. We had... technology company comes to mind, many others, that had said, "We solved in 15 minutes what would have taken us days and days and days." That's simply an example, but it is a powerful one of where observability data adds enormous value to certain application security aspects or areas of deployment. We're happy with where we are.
To the second part of the question, in terms of security and its evolution and where we still believe we're on track five quarters out, I guess, by the end of fiscal 2025 for us to deliver against the $100 million goal.
Excellent. Let's go back to the audience and see if there's any questions. If you raise your hand, we'll get the microphone to you. All right, does it look like we have any? Oh, just one here in the middle.
Yeah, maybe you could just speak a little bit about DPS and to what extent that makes you excited to cross-sell the new products that you have. And also, I'm sure a lot of us are wondering, too, if you see it in general being an expansionary event for the customer, or if it's something where they're committing at roughly the same level.
Good question.
Oh, you want me to take that?
Go for it.
... So just to make sure we're level set for everyone that might not know what DPS is. DPS is a Dynatrace Platform Subscription . It's basically a contracting vehicle. So we've historically sold SKUs. So you want to buy our products, you buy a SKU. Whether you want to buy application, performance monitoring, infrastructure monitoring, logs, et cetera, it used to be SKU-based. So one of the complaints the customers had was, "We love your products, we don't like your contract processing, because it's, it's difficult. Everything needs to be a recontracting effort." The Dynatrace Platform Subscription model was intended to address that, that you basically, you commit to a dollar amount, over a term, whether it be a year or, say, three years. You know, for a fine item, you commit to more, you get a better unit price.
You commit to less, you get higher unit prices, and you get access to the full platform. And it's a very similar model from an accounting perspective as our SKU-based model, which is you sign up for a one-year contract or a three-year contract, it's still annual billing in advance. It's ratable revenue recognition, so free cash flow is all the same, revenue recognition's all the same. The benefit is it gives the customer full access to the platform without having to buy, you know, à la carte, this, this, and this with a recontracting process. The whole premise of it was to satisfy a customer pain point. We're early in the journey. It's generally available was our first quarter of this year, so in April.
And so we have about 10% of our customers that are now on, now on this vehicle. The whole premise of it is that we believe the thesis was that you'd get a higher propensity to expand, because they would consume faster, because it's an easier-to-consume model. And we're tracking the cohorts, like the Q1 cohort customers, Q2 cohort customers. We haven't kind of anniversaried yet kind of a year. But the good news is, right now, we're tracking that those customers that are on this vehicle are actually consuming at a greater rate and pace than their cohorts that are not on a DPS contract. So it's early days, still only 10% of our customer base, but it's behaving the way we had expected.
And we're optimistic, and it's certainly gonna be a vehicle that we're gonna introduce to customers, won new logos and existing customers as they expand. I don't know whether you'd offer anything else.
I think that's good.
Maybe to wrap up on the coverage, my first real CFO question, right? To wrap up. The trajectory of operating margins and free cash flow margins against what's, you know, an exciting opportunity in observability. You talked about maybe some of the go-to-market changes, and maybe that requires some investments. We're on the cusp of a new compute cycle with AI. So in the context of all of that, how should investors be thinking about operating margins and free cash flow margin? Here, given that you've given us two... You're on track to give us 200 basis points of margin expansion in fiscal year 2024. So if you can contextualize that for us.
Yeah. The way to generally think about it is the way we think about it internally is we do look at Rule of 50 within the company internally. We look at Rule of 50 being defined as kind of ARR growth plus pre-tax free cash flow margins, because we can't control taxes. And we're mindful of both. And, you know, what you've seen in the model over the last few years is this, it's pretty healthy margins. But, you know, our pre-tax free cash flow margins on a trailing twelve-month basis is 30%. So very healthy, with 20+% ARR growth. Now, obviously, our guide was 18%-19% for the full year for ARR growth. But we, to us, it's all about making sure you're looking at both, right?
And we guided to 27%-27.25% operating margins, and we've had some investors ask: "Well, if I do the math on that, it means your Q4 margins based on what you've guided, your Q4 margins might be 24%. Should I believe that that means you're gonna go into fiscal 2025 at maybe a 24% exit rate?" And the answer is no. You know, the way we look at it internally, and the way you should think about it, is the full year operating margins of, call it, 27%-27.25%, is call it the baseline that we're gonna operate against. You know, then now you get into a discussion around, well, how much accretion do I want to drive? Well, now there's a balance now between investing to drive more growth.
And for us, the last couple of years, we've been investing. We've been investing much more in R&D, frankly, than we have in sales and marketing. And we have because with in a macro environment that's been uncertain, it hasn't been a good use of spending to put more money into go-to-market in a period of uncertainty. I think what you should expect going forward is that as we see the demand environment kind of improve, we will make more investments on the go-to-market side. That doesn't mean we'll grow sales and marketing greater than revenue growth, but it means we're gonna weigh more investments there. So we are striving to get to the point where we can stabilize the ARR growth.
Mm-hmm.
and then get a kind of re-acceleration. We believe it's the right model for the company. I would agree with you that this market can and should allow us to be a 20%+ grower. I think we have great set of capabilities to be able to do that, and I think the benefit you get with the Dynatrace model is you get really healthy margins and an ability to drive 20%+ growth over the intermediate term.
Well, let's leave it there. Thank you so much, Rick and Jim, for joining us on the conference and shedding some light on the Dynatrace story.
Thank you. Thank you, all. Appreciate you for coming.