Dynatrace, Inc. (DT) Earnings Call Transcript & Summary

Welcome. Welcome to JMP's virtual 2021 tech conference. I am Erik Suppiger, JMP's cybersecurity and infrastructure analyst. And I want to thank all of you for joining us today. This is going to be our keynote with John Van Siclen, the CEO of Dynatrace. I'm going to dive into some questions in just 1 moment. But before I do, I really want to encourage all of our attendees here to please feel free to ask questions on this call. You'll find the Q&A box down at the bottom of your screen, and we will take questions in real time. So please do join us from a question perspective. So John, welcome. Thank you very much for joining us.

Thank you, Erik. Appreciate it.

So there's a lot going on here with Dynatrace, but I want to take a step back and look a little bit at your -- the background here. Can you walk us through the background leading up to Dynatrace? What brought you to Dynatrace?

Well, so I've been with the company now for 12.5 years. It was the summer of 2008. And we all remember 2008 or we'd like to forget it maybe, but it was an opportunity to either -- I had an opportunity either go in the security business or the performance management business. And I had watched several companies in 2001, 2002 when the bubble burst in Silicon Valley, that actually exploded in times of challenge. Mercury Interactive being one and Wiley being another one, early performance management companies did extremely well. And I realized that they did extremely well because when times are challenged, enterprises turn to some of their most precious assets and try to optimize them, and software is one of them. And so that sort of helped sway me to the performance side in 2008 and never looked back. Just thrilled to have been part of a nice, long and successful journey with Dynatrace.

Okay. And Dynatrace has had a long history. It actually, I think it was founded in '05 in Austria. Can you talk a little bit about the evolution of Dynatrace and how it's gotten to be where it is today?

Sure. I'd say that there's really like 3 chapters to the business. There's the venture stage, which was that early-stage from 2005 to 2011, actually. So the first real 5 years of the existence of the company, where we were establishing ourselves as an application performance monitoring company, uniquely positioned at the core of sort of the Java explosion in Java and distributed applications. That then turned into a stage where we were part of a portfolio of products at a company called Compuware. Compuware mainframe company trying to reinvent itself into a modern distributed application performance company. And I ran that organization there. And that actually gave us a very nice wide global footprint, a very mature company, wide global footprint. And then Chapter 3 has been really the reemergence of an independent Dynatrace about 6 years ago with the help of Thoma Bravo, private equity firm. And they set us up as a separate business carved out, and we've been sort of -- and we reinvented off of that moment into the company that we are today.

Okay. And when you were at Compuware, was it a fairly independent operating entity?

Yes and no. From a business standpoint, yes. But financially and operationally, it was one common platform, which certainly creates a challenge when you have a very low transaction business on the mainframe side and a high transaction, modern business on the Dynatrace side. So that carve out was really quite liberating, honestly, for being able to really run a much more entrepreneurial business, one that could -- that had the room to reinvent, not only the platform, but the entire business model from a traditional license business over to a subscription business. And the Thoma Bravo team, super supportive in that process. I will say I'm happy that we did it in private equity, not in public markets, like a number of companies are trying to do today because there are challenges. It takes a little while. But the results have been obviously quite good and really excited about the performance of the business, really in the last 3 to 4 years when we really hit the ground running on an enterprise-class new platform, subscription-based, targeting modern cloud environment, not legacy data center kinds of things and transitioning our entire customer base and company to this new model.

Talk a little bit about the contribution from Thoma Bravo. They were the majority -- they were the majority owner for -- up until the IPO. What involvement have they had from a management perspective? And how has that diminished since the time of the IPO?

So Thoma Bravo never really tried to manage the company. It's one of their unique attributes, unlike some private equity companies who try to come in and grab hold and drive the car for you, they step back and just ask your questions and show you others who have done something similar successfully so that you can make your own decisions. They bet on the management team to know the market better, the product better, the -- what they need to operate better, but they help with the guideposts. And they've been phenomenal at helping us build a much more efficient software business than I ever thought I would run. It is truly more efficient. There's some very simple things that we did that maybe give the investors some ideas here. So when we first carved out, more than half of our R&D was done in North America. It's cost a lot of money, lots of turnover and hard to get the very best engineers if you're in Boston and Detroit. And we have now moved more than 95% of R&D to the Central Time zone in Europe. Austria, Poland and Barcelona, Spain. House nearly 8 plus labs now over in Europe. Phenomenal talent, great innovation and a cost base that we just couldn't dream of here. I mean, if we were trying to do the R&D footprint we have in Europe in North America, it'd be at 30% of our revenue. Today, it's only 15%. So very efficient that way, and I could give you more examples of how we intelligently rebuilt a very efficient, reimagine a very efficient model, which gives us this great balance between growth and profitability. That's unique, frankly, and within our peer set.

Okay. Very good. So I want to move on. I think it was around 2014 or '15 that Dynatrace went through a pretty major product transition. Basically, you reengineered the entire product. Discuss the transition there, discuss the -- what was going on in the market that really prompted you to do that.

It's a great question. I know it's a little bit of throwback for some of you. But back in 2013, '14, everybody could spell cloud, but had no idea what it was. And we weren't really that different at the moment, and we were obviously all reading about it. And after discussing it several times, we decided, look, we need to go do a study and really understand, is this going to impact our business and in which ways. And a dozen of my top folks, including the CTO and founder, who's still with me, Bernd Greifeneder, went off and 3 months later came back, and I remember the meeting actually. It was a dinner in Linz, Austria, which is where our major labs are located. And the team said, "We need to reinvent everything. We are not going to evolve here. We need to start from ground 0 new code from scratch because the cloud is going to disrupt everything in a massive way. And we can either be disrupted or be the disruptor. Which do you want to be?" So after 2 days of meetings of exactly how we would do it and what the key components were, we decided we'll be the disruptor and I think we've proven that out over time. Some of the things that we saw, first of all, that the cloud is all software. And so what used to be a nice defined layering of applications and infrastructure and network, we're going to all collapse into a software virtualized environment, and it's going to be super complex to distinguish what's touching and disrupting what in that cloud. So we were going to have to think about an entire full stack view, network infrastructure and apps, all at once, not just applications. Second one was, it's all going to be dynamic. Things are going to be moving. There's going to be dynamic orchestration. And so therefore, instrumentation has to be automatic and continuous. There's no manual opportunity to place things anywhere. And the third big one was the scale. Instead of these nice sort of concise stacks of apps on infrastructure and then you add another one and another one, it was going to be the data center all in one massive web scale cloud. So we're going to have to increase scale by about 100x. So all these things sort of swirled into new code. And it took us some time to mature it. And the rest, but it was about 4 years ago, we went enterprise. 2016, I guess, almost 5 years ago now. And it really has transformed the company. And as I said, it wasn't just the platform. It was the couple of other dynamics, one was, instead of being in the data center side in application only, we went to the modern cloud side in all of our customers, which is where the money is being spent. The second thing is we went to a subscription model and away from traditional licensing model. And that, of course, gives us accelerated growth and much more predictability with the subscription model, both required to be a public company.

Talk a little bit about the old product. How is that performing? And then how did the transition? It was really a multiyear transition. The product development was a while and then also the transition in the sales organization. Can you walk us through that?

Yes. So the -- let me walk through the transitions a little bit here. So R&D was the first to transition, but we still needed a foot in existing products because we were still growing. If you look at the Gartner Magic Quadrant of application performance monitoring, you'll see Dynatrace in the upper right, all the way through 2014, '15, '16, '17. So we're still a leader in that segment. But the reinvention allowed us to be a leader in a much broader modern cloud segment. That's now being -- we used to call it full stack -- a full stack view beyond APM, we didn't call it observability at the time because there was no such term, but now that's become the term that we were already in and defining for our customer base. So R&D was the first one. Sales then needed to transition, did a very nice job transitioning both new customers and some of the early transitions of existing customers. But over the last, starting in, what was it, 2019, we really started converting the base in a big way. So today, there's virtually no classic product left in the base. It is all Dynatrace, the new platform, all modern cloud. And again, just a massive transition from an old sort of multiproduct group of things that inherited through those Compuware years into a truly platform driven Dynatrace business today.

What were the biggest challenges that you faced with that transition?

The biggest challenge is we're not really internal. The sort of -- the biggest challenge was really with the customer set because we had to move from an existing Dynatrace champion or sort of the classic product champion to the cloud teams. And that was probably the biggest thing. But the whole company was lined up to do it. Everybody was certainly excited about it. It was clearly the right thing to do, even when we're doing this in 2016, '17, '18 and then the big conversion years in '19 and '20. And so it's never an internal challenge, is it the right thing to do. And sometimes, that can be the biggest impediment sometimes of change, but it's really getting our customers to think about us differently. And we accomplished it. We did it really well. And that set us up for where we are today. I mean, if you really look at the spend and what's happened in the pandemic, it's the digital transformations that have accelerated, but they're underpinned by modern Cloud. If we still had a big footprint in a classic product space, there's no way we would be enjoying the growth that we are today.

How did you attract the engineering talent? This was well before cloud was very popular. Where did you get the engineers?

Well, as I said, we were in the process of also transitioning our engineering from North America over to Europe. And the -- all of our labs sort of sit in technical -- near technical universities and technical centers. And part of it is culture and part of it is working on sort of new modern technologies. And part of it, frankly, for the European technical teams is having a platform where they can impact the global software footprint, a lot of them get sort of caught in only being able to impact the European footprint. But Dynatrace being a global footprint was like, "Oh, my software can go globally to the world's largest companies as I build it. I like that. I'm going to join Dynatrace," and we've had a very high filter for top engineers. I say all the time, we couldn't have gotten the Google engineers in the U.S., but we can get them in Europe. And it's -- and they're super loyal. So our turnover is very low single digits. And so that's a big advantage over other parts of the world, as you know, from a consistency of R&D. And these kind of platforms that we're talking about, these take architecting and thinking about them for 3 to 5 years out. They're not things that, "Hey, I'm just building the latest features here." It's thought about over a long horizon and consistency of team matters.

So let me move on to some of the market dynamics. First off, penetration, how do you think of penetration of observability in the broader market? And then how do you think of your penetration within your installed base?

No, a great question. It's hard to call. I mean, it's early days for observability. It's early days. But everybody has some alternative. They're all scrambling to try to get some window into how these clouds are working. For a while, people thought they were just going to take their old tooling and pull it over, that, that would work. And the reason was because there was a lift in shift concept. "Oh, yes, just take that big monolithic apple, just lift it over, and later, we'll figure out how to carve it up into a cloud-native architecture." Well, those -- that thesis is gone. Everyone knows now that in order to take advantage of the modern cloud, you're going to have to refactor your app and rebuild it natively in a cloud environment. And so that's changed the dynamics of being able to take old tooling and just move it across, people are hitting the wall. And as they hit the wall, they need something. And so most of them are -- if they don't know who Dynatrace is, they're trying to cobble something together. And usually, we compete with what we call a do-it-yourself approach. I'm using this kind of open source that -- I have an open source database. I'm trying to do this kind of correlation engine. I'm throwing a bunch of people at it. I'm not really focused on my innovation as a bank or an insurance company or a logistics company or whatever, I'm actually getting into the monitoring business. I don't really want to be there, but there's no good alternative, and then we show up, and it's like, "Oh, gosh, I had no idea, somebody already had a platform that could lead me ahead and keep me up-to-date all the time. So I just have to do some tuning around the edges, and this thing can work for me." And so that's why it's sort of this early days of observability, but it's a great opportunity, certainly, as everybody sort of hits this complexity scale wall in the cloud for us. And it's been driving our new logo growth. As you can see, it continues to drive our net expansion rates to be world-class. So those are some of the dynamics. Underneath is that it is a disruptive moment right now as these clouds scale. And the challenges of the complexity start to overwhelm the IT teams and the digital transformation teams. And it's just -- it's a great time for Dynatrace to enter more accounts. And so I'm sure we'll talk a little bit about some of our strategies there.

Well, what about in your installed base? How much -- how broadly is Dynatrace used relative to the opportunity?

Today, the way -- the markets thought about sort of penetration of our type of observability, sort of application deep version, full stack version, penetrated in the 5% to 10% range. Our customers are probably more in the 20%, 25% range, but they all want to get to 60% plus. Even our largest customers with the widest footprints, none of them are saturated. So I talk about this application space as really an evergreen kind of space. I mean, I think it was IDC did a study 3 years -- or a couple of years ago, they came out with it last year and said that based on their study, they believe that there would be as many or more applications built and deployed in the next 3 years as over the prior 30. So -- and it isn't just more apps, they all scale wider, and being in a consumption-based model, both vectors work for us. And that's just the application side. Let alone some of the other platform modules we've been expanding into, like infrastructure, like digital experience and now application security.

Okay. All right. So let's talk a little bit about the competitive dynamics. There's a lot of competitors. It's a bit of a crowded market. How do you think that just the competitive dynamics will evolve over the course of the next several years? We've seen a lot of convergence of products, where APM converging with infrastructure and network. We've seen outside players that are expanding into this market. What do you think are going to be some of the dynamics as we look ahead?

Well, I don't have a perfect crystal ball, but I definitely have some opinions. So first of all, I think to understand this market space, there's a couple of key things. Most people just think on the technical vector, like what your core strengths coming in, ours would be application, some others might be log, some others might be infrastructure metrics, that kind of thing. The other one though is market attack. We're an enterprise based company with a direct sales team, more like a Splunk or a ServiceNow. And others are Freemium, come in through sort of developers and up, sort of grassroots and up. That works great for long tail, broad market, global, 100,000, 200,000. But the enterprise level is different. Because they don't look for the individual productivity tools of developers, they're looking at how can I automate things, how can I drive consistency, how kind of consolidate disparate items, how do I move faster. Their concepts are a little bit different at the enterprise and the scale and complexity is greater. So we've stayed focused on the Global 15,000 enterprise direct. I believe Direct is a perfect fit for our product, which is very different. It's not just observability. It's actually automation and analytics, which is where the real value and uniqueness comes from. And so we need to be present to position it properly and make sure customers understand the unique value or the value advantage of Dynatrace. And so those are some of the dynamics. So think market and think technical vector in. Once you do that, you realize $30 billion or $50 billion market, nobody's even $1 billion. So there's just a lot of running room before there's collision. And when there is, then the question is, okay, what's going to be the key differentiating factors. And if somebody is just doing data collection and putting it on a dashboard, that's going to be very commodity. We've always thought of ourselves as an automation and analytic company that happens to solve the observability, challenges as well or better than anyone. And we believe that the automation and analytics is really the long-term value opportunity that will continue to set Dynatrace apart from the observability only players.

Would that suggest that other data analytics companies, like a Splunk, would be in a strong position to enter the market?

Depends on how automatic, depends on how real-time, depends on how scalable across all the different observability data sets. Can't just be logs. Okay? That kind of thing. They're a great company, great engineering and don't underestimate sort of their capabilities. But they also come in off of a enterprise analytics platform, and we come in off of a cloud digital transformation platform. A little bit different problem set, a little bit different design center and a little bit different way the customers think and guide us because we both get guided by our enterprise customer base to solve what really are 2 adjacent problem sets.

Okay. So how difficult -- as I mentioned, we're seeing convergence between infrastructure and the log and the APM and the network. How difficult is it to move into an adjacent market? And I'm thinking of infrastructure for you. You introduced the infrastructure maybe a year ago. And I think it took a little bit to get it going. It sounds like it's going well now. But how do you think of your ability to move into adjacencies? And how difficult is it for somebody else who's moving into APM?

Well, so we've always -- let me start with saying that we've always thought that the application layer is the high ground. And it's the high ground because it's where business meets the infrastructure. I mean the CEO doesn't ask, how is the Kubernetes environment running today. They ask how about this application, that application. So it really is a super important area, and it's always been sort of rocket science to be able to do very precise lightweight instrumentation, scale it, don't impact the application load significantly and do it at a very high fidelity rate, right? So that's been a characteristic of Dynatrace in that environment. And I believe that -- and you can look at the -- some of the industry analyst reports about who are players in the full scope of application performance, and you're going to find a very short list all the time. On the infrastructure side, it's different, it's a wide coverage range. It's a different problem set. It's not quite the technical rocket science. It's more about just making sure you have a very broad coverage set. So we found it to be pretty straightforward to go do. We just had to think a little bit differently. We were trying to take all of the infrastructure information and append it to the application set. And we realize, no, no, no, you have to have all of that for the personas that care about the infrastructure who don't care about the applications and that generalization of the platform has been what we've been going through over the last couple of years. We always had the data, we just didn't have it generalized in a way for those personas who don't -- really don't care about the apps. Other teams put the apps on top. They're providing cloud platform for their company, if that makes sense. And so I think that we're doing well right now. We have about a 40% of our customers use the infrastructure extensions for our platform. Still early days as far as monetizing it. I think we're only in the $0.10 to $0.20 on a dollar that we can go get there as we expand out but that's great opportunity for us. It's the fastest-growing module from a percent of ARR growth, obviously, coming from a little bit smaller than some of the other ones. But it's a great opportunity for us. We're doing well, and that generalization I was just talking about is starting to play out for us.

What about other areas, network or log? Are those similar to infrastructure?

We put all that in infrastructure. So that's a really good question. So that's a good clarification. The network monitorings in there, the logs and log analytics are in there. One of the things that is different about Dynatrace and our other competitors is they want to charge for every little different type of data. And that works great in the mid-market, but enterprises don't like that. That's like death by 1,000 cuts. They want it packaged up for these use cases, and I want to know predictably what it's going to cost me for my various years of my contract, and they don't want surprises. And so that's very enterprise thinking, which is our customer base. So we've always packaged by use case, not by a little slice of data type. And so that's, I think, a really important consideration for us. It also means that just because cloud adoption is going this pace, maybe you're going to get some of that. Ours is a more gradual curve underneath in order to support the predictability and transparency, pricing packaging requirements of our customer base.

Okay. Great. You touched on this a little bit earlier, but your focus is the -- I think you talked about the Global 15,000. This is the high-end enterprise. Have you looked at expanding down market? Is that something that you would do sometime down the road? Or why are you so focused on that segment?

Yes. It's another good question. So there's nothing that precludes us from going down market. Product would be the same, everything would be the same. We have free trial, all those kinds of elements. But we do the enterprise extremely well. The enterprise is where the money is. It's also where the complexity and scale problems are greater. We happen to have a fantastic platform for handling the complexity with the automation, analytics, built-in and the scale. We're unique at sort of the high end, ability to continue to scale for larger and larger clouds. So it works well for us. The other thing that I've learned over my career is that just because you go after something bigger doesn't mean you go faster. It's like a river. I try to explain it to people. It's like a river, the wider the river, the slower the water goes. You start pushing it through the gorge and it accelerates. And I think that we have so much more room in the Global 15,000. I mean, we're just approaching 3,000 customers. We have lots of room to go. And our ARR per customer is $250,000 average as opposed to what we believe could be over $1 million ARR per customer. We got a lot of running room to build a big company, just staying focused. And then when we need to expand, we'll go to Global 20,000, Global 25 I don't think you'll ever see us just go -- let's just go everywhere. That wouldn't -- it probably wouldn't be to our advantage nor would it be to our economic sort of shareholder value.

So do you see some of the players that are more in the mid-market, do you see them moving up? Do they start to bump into you? Or how do you think of the prospects for other vendors that are targeting that segment?

Well, the mid-market will always want to move up because it's where the money is. So there's a lot of -- "Well, look, I have this many hundred thousand dollar customers and this many million dollar customers" and that kind of thing. So there's definitely a -- I get more value if I can climb that ladder. But it also has a different set of expectations to it and different sort of approach to it. And there'll be some inroads, certainly. But again, being enterprise-focused and just stay on that track will allow us to get sort of that, we believe, the dominant footprint, which always is sort of an accelerant to additional business as well as command to premium in the market. So I like the strategy. Other ones, we'll try to come up off. And our job is to own that enterprise layer. We'll see how it all goes. But like I said, I like -- I think the dynamics are different, and I think we anticipate them well.

One of the things that you've talked about is something that the high end really consumes is the automation component. Is your automation a differentiator that -- it is a differentiator today. Is that a sustainable differentiator? Is that something that you can continue to build on? Or is that something that you automate and then you're pretty much there?

Everybody is on a journey to an autonomous cloud. Nobody wants people touching anything if they can help it. It should be self-healing, no matter what the issue and challenge. But very few are there. I mean Google is there and Amazon is there and Netflix is there, but then you go to a very short list beyond that, it's -- there's a lot of work. And we view one of our opportunities as helping to accelerate that movement for that Global 15,000 toward autonomous cloud operation. At the same time, we also have an analytics engine, an AI engine that helps with the understanding and predictability of what's going on, providing action, very precise action, prioritized actions and insights into what's happening in those cloud environments. So the combination of the automation for autonomous, cloud and the analytics for intelligence and very precise actionability are 2 big levers, and we do believe that they are sustainable advantages. I mean, we've been out there for a while. You've probably -- if you've watched this industry, others have claimed, "Oh, we're going to be there, we have this or we're going to have that," and they never show up. They stop talking about them because it's hard. And you need to architect them, we believe, at the core of the platform, you can't bolt it on the side effectively. And that's a real differentiator. We have -- like I said, we had that advantage to do it out of the public eye where we really could reinvent from scratch. And it's serving us well, and I believe it will continue to serve us well. I don't think that's an easy gap for others to fill.

Okay. I think we have some questions here. So I'm going to turn it over -- Rusty, are you there?

Yes. Thank you, Eric. So I'll start off. We have a number of questions here. The first one is kind of talking about driving ARR per customer towards the $1 million mark. Can you talk a little bit, John, about the journey there and what that would involve from both an execution as well as the new product perspective to drive that ARR to $1 million per customer?

Sure. Yes. So if we've never done it before, then it would be hypothetical, but we have plenty of customers at $1 million-plus ARR. So we know that that's achievable. What we also know is that if you have 3 or more modules, you're well over $400,000 per customer average. So we have less than 1/3 of our customers on 3-plus modules. So obviously, we want to get more on 3-plus modules. We also just announced an application security module, which, over time, can easily be a dollar-for-dollar with the application module. So it's not hard to sort of see the journey and the map to -- toward $1 million ARR per customer. It's just about landing new logos and expanding them with more modules over time. And the more effective we are with that, you can build a big company. I mean, it's not -- just like what ServiceNow has done over the last 5 to 10 years. Very similar kind of model, platform, 1 plus 1 equals 3 when you add another module to it because it already has all the platform attributes. And you continue to drive ARR up per customer with a growing customer base.

Thank you for that. Now switching gears to more of a channel partnership and development strategy. This investor would be curious to know how your system integrator and partner development approach is going and a little bit of clarity on the speed at which your certified partnerships have been growing.

No, great question. Because it is a really important part of our strategy to date, but I think it takes on even a more prominent role. I started talking about it at our last Investor Day in the fall. The cloud system integrators are driving digital transformations now in a very broad set of the -- certainly of our customer base, Global 15,000. Some of them are the global system integrators, like an Accenture or Wipro or Infosys, DXC, et cetera, and some are just -- are very strong regional system integrators. We have a growing set of those. Trained, capable and helping to support Dynatrace in -- as a key component of their go-to-market for digital transformation. On the other side, we have alliances, and this is where the hyperscalers are, it's where we have ServiceNow and OpenShift in Tanzu and those kinds of folks. They're turning from technology partners, where we have very fine integrations and sort of automatic sort of integrations with them all to becoming go-to-market partners in a bigger and bigger way. And we just announced the Google partnership. I talked on the last earnings call about some of the Microsoft work we're doing. We've always had a great relationship with AWS. But they -- as I said, they are going from technology only to actually being really interesting go-to-market partners as they do more sort of private offers, capability, be it parts of their marketplace, allow their customers to retire precommits like with a Dynatrace product and then compensate their sales organization for those kinds of sell-through of the precommits. So it's a changing world out there. And we're playing both of those partner angles right now to give us more lift, drive more productivity for a growing sales organization.

In terms of the -- circling back to your comments around the technology road map, when you mentioned the orchestration with a DevOps focus, could you expand a little bit on what you are suggesting from that?

Yes. So from a -- so we've always been -- being in the application side of things, we've always thought about it as a complete software delivery life cycle. CI, continuous integration, continuous deployment out into production and sun setting of technical debt in production. So it's actually like 4 key pieces. There's a production and you shift left to the preproduction shift, right into various additional types of production, software management. So we've always been in that entire spectrum. Where we started focusing more on recently is the automation of the CICD pipelines, where we think we can bring some real sort of unique advantages of Dynatrace between that automation and the intelligence that we provide to allow CICD pipeline or the DevOps teams be more predictable and quicker with their innovation, at the same time, they're improving their code quality, and do it not just for the top DevOps team that a company has, but across their entire set of DevOps teams. So anyway, it's something we've done well in the past. We're pretty well penetrated in our customer base on that DevOps side, but we can do more. And with the new app security module, which has a really great play in the DevSecOps movement, I think we'll see more and more sort of pre-production kind of use of Dynatrace throughout our customer base.

Great. And that segues nicely into the next question. So a lot of organizations have been discussing the cloud application security space, some from a DevOps perspective, while others from a more traditional security background. And now the likes of yourself and Datadog are approaching it from this observability approach, what do you view as the correct approach here? And is security and observability more of a competitive dynamic or a complementary dynamic?

So the way we -- so first of all, the security strategies are layering. There's not one -- I'm just going to go with one. I'm going to layer them. So -- but what's changing in the world, which is giving us an opportunity to step in is that the concept security ring-fencing and sort of protecting the edge is starting to break because these applications are sprawling across clouds, driven by the Kubernetes orchestration, and there's -- just really hard to ring-fence these things. So the only way to really solve that problem is actually have the applications being aware. Aware of their vulnerabilities real time, eventually aware of attacks in real time. And that creates an opportunity for us because we live in the apps. We have code level detail of every type -- we have every entry point, exit point. We have all this information. It's just having the algorithm set and connecting some dots, having some partnerships like with Snyk, et cetera, in order to be able to put together a very powerful unique security offering for where the puck's going, not just where it's been. So we like the vector of attack we have. We actually think it's unique. No matter what everybody else wants to talk about, we still think we have a unique one because of our code level detail and AI engine. And the intelligence we put into the -- removing false positives to make the Dynatrace approach much more efficient for developers and much more robust in real-time for the CISO.

Rusty, I know you've got more. I'm just going to ask 1 quick one in light of that one. Talk a little bit about the opportunity size. I know the market, you talked about being an $18 billion opportunity, which is a big portion of your overall $50 billion opportunity. But just talk about how you think that would -- that opportunity compares in size to your traditional observability. I think you've said it can be one for one. So could security be equal in size to observability overall?

Yes. When I referred to it, I was talking about more of the application modules of the observability, but still, I mean, whether it's $18 billion or $15 billion or $25 billion, it's big. And some of this is forward-looking, of course, because when do all these workloads really move to the cloud, when do they go multi-cloud, when is it truly dynamically orchestrated, et cetera. But I'll tell you that the customers I've talked to about our approach have thought that it's spot on. The thesis is absolutely right. We have to get our coverage a little more full. We have to do a little of this, the CISOs all have -- want to check it twice and 3 times before they'll let anything in there. But there is a great opportunity there, and it is, we believe, quite greenfield. Yes, there are alternatives, but the alternatives all fall short, which is why everybody is looking for something different. They're either preproduction scanning or they're periodic scans in production with lots of false positives, none of it is efficient. And everybody knows that they can't build a big enough stock in the future to deal with all the security vulnerabilities. So they're all looking for alternatives, and we have a really, really good approach at the right time, and I think it can be a really big part of our business. It will take some time. But it can be a very big part of our business.

Okay. I think we have a few more minutes. Rusty, you got a couple more questions. So let's try to fit these in, if we could.

Sure. Sounds good. So just an update on the Kubernetes and microservices as well as serverless monitoring. There hasn't really been any hack in the microservices space. So curious if you could secure the microservices environment and how you kind of think about serverless monitoring more broadly?

Yes. So it's just sort of a couple of different questions in there. So first to bat, on the security side, we targeted the Kubernetes world with cloud native. Like, where the puck is going. We really aren't trying to solve. Even though we say, yes, it supports Java and Node, we're not trying to go back into a monolithic stack security project. People have their -- whatever they're using, it's not really our approach. Our approach is to go where things are going, where it is greenfield. It's a smarter place, and everything is going to -- all those job application is going to be refactored anyway into cloud-native and we'll be there when they get refactored. So I think it's a smart strategy. It's also why we dampen expectations on ARR growth early. But by fiscal '23, that's sort of -- think calendar 2022, it will start becoming a material part of our ARR, and we'll talk some more about it at that time from a financial impact standpoint. On the other front, and I forget now the question, I should have written it down. Wasn't security...

Rusty, serverless, I think?

Serverless monitoring.

Yes. The serverless is -- more and more of our customers have some dimension of their cloud that's serverless. They're all talking about mesh. They are talking about other pieces. These are perfect areas for us. Open telemetry is going to be a key player in a lot of these new technology areas. It's an industry standard technique that we've been a member of that group because we see it as a great sort of driver of more data into our platform. But all these pieces are things that we cover today. And we're unique in our ability to stitch all of these pieces together and provide a comprehensive view from the very most modern, all the way back through hybrid clouds to data sources of record that still matter for these environments. Mainframes and mid-range systems that provide just a lot of the data of record that these apps depend on. So unique on that hybrid side from front end to back end.

I think at this point, we're going to have to wrap it up. I have lots more questions. It was very informative. Greatly appreciate the insights, John. And -- but we are going to run out of time. So I'm going to wrap it up there. John, I really want to thank you very much. That was a very interesting discussion. So I appreciate your insights.

Appreciate it, Eric. Thanks for letting me join.

All right. And for our attendees, thank you very much for your time. I hope you enjoyed that as much as I did. And with that, I'm going to say thank you and farewell.