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

Thanks, everyone, for joining us. My name is Sterling Auty, software technology analyst here at JPMorgan. Very happy to have with us John Van Siclen, who is the CEO of Dynatrace for our next session. Before we get started for participants, if you'd like to ask a question, go ahead and hit the Q&A button at the bottom or to the side depending on your device, enter in your question, I'll go ahead and include it as part of our session as we get forward. So just to get started. John, thanks for joining us. We really appreciate it.

Thank you very much, Sterling.

I was interested to see the Dynatrace name rather than the Stanley Cup that I normally see in your Zoom sessions.

Always wishful thinking.

Exactly. All right. So for those that are maybe not as familiar with Dynatrace, can you give us just a brief overview of exactly who is Dynatrace and where in the market do you compete?

So Dynatrace is a leader in advanced observability for dynamic, multi-cloud environments. And that includes, for us, application, the application layer, infrastructure, network, logging and user experience. And from a macro trend standpoint, the 2 big trends driving our opportunity, the first one is software eating the world as applications drive more and more business, customer interaction, the way the whole world works. And the second one is that behind every digital transformation are cloud projects, cloud platforms for agility and efficiency. So we're a pure-play in both. We've been public now for almost a year, 4 earnings calls as of 2 days ago, and growing well, doing well and trade on the New York Stock Exchange.

So one of the things that I was most surprised about when I was really digging in, getting to know Dynatrace over the last couple of years is, so many software companies fall victim to the innovator's dilemma. I mean Dynatrace has been around for a while. And the observability and monitoring space is not something that's new in general. On-premise monitoring has been around for many, many years. And Dynatrace really got its start there. But the fact that you are able to make the migration into the cloud was so surprising. Can you walk us through what exactly did you decide to do? When did you do it? And why did you do it?

Well, it was about 6 years ago now. We saw the cloud coming, and how disruptive it would be to the entire monitoring space. The traditional monitoring of the data center was very structured and layered. Companies had many tools, we happen to play in what was called the application layer, at the time, built a category leader there. But everything was quite fragmented and discrete. We saw that with the cloud, everything was going to collapse. The cloud's software, network is software, infrastructure is software, the application layer, user experience, it's all a collection of processes and services that have to work together at mass scale. And that was going to turn the monitoring business from this fragmented set of discrete tools to much more of a unified platform approach. So what we did was, as you pointed out, the innovator's dilemma approach where we put a team of our best engineers off working on what was next, didn't use a lick of the same code. And here we are 5 years later. And the startup has eaten up the parent, 92% of our ARR as of the end of our last quarter, the end of March, is now on the new platform. So it's been a rapid, not only has the platform been very successful and well received, but has swept through our customer base over the last few years and really is the -- it is the business going forward.

So in terms of the idea of being born in the cloud, I want to make sure that I put a fine point on, you said not a lick of influence. So was there any code that you just kind of ported from the old platform to the new?

No code was ported. We reinvented every piece from how we instrument and self-discover. We put an AI engine at the core of the platform, the analytics were all rebuilt. Obviously, we have a core set of expertise and knowledge in the area that accelerates and allows us to advance sort of the concept of monitoring or observability at scale. But the problem set was so dramatically different. I mean, just take, for example, a dynamic environment, Kubernetes environment, which 80% or more of our customers are either aspiring to or already implementing. Those workloads and that containerization changes ephemerally, scaling up and down based on load and different application characteristics. If you try to instrument the way we used to in the old days where you had to configure, you had to know the environment and configure it, it's impossible. It's too dynamic. It's ever-changing. If it isn't automatic, self-discovered all the time, continuously, then you don't have a chance to properly monitor. Way too many blind spots, way too many challenges for those who sort of depend on that level of visibility or situational awareness across a broad cloud that's running multiple -- run the business applications on it.

So if you started in 2013, with that development, when did you finish and what does that actually say about the freshness of your coded in platform vis-à-vis the other competitors in the space?

It's a good -- it's a great question. I mean, 5 years ago was when we brought out the initial platform and it took us about 18 months before we had matured it to be enterprise ready. And so that was about 2018 is when we really leaned into our customer base to start to take this new platform and move it in and across that customer base, what we call our conversion program. And it took 2 years to move from very low percentages of our ARR being on the new platform to now 92%, as I said. But the platform is such a leap ahead, especially for the modern cloud workloads that all of our customers are utilizing sort of to underpin their digital transformations. That it's moved extremely quickly through. We're very pleased. It's a little ahead of schedule for us from what we thought, but it's been extremely well received and a great -- it's not just that we're selling or providing the same technology to the old data center teams, we're actually leapfrogged to the cloud teams where all the investment and ongoing innovation is, with this platform as well. So it really sets us up well for growth in the future.

So if I look at some of the names in the space, the New Relics, the Datadogs, et cetera, New Relic, obviously, an APM, so application performance monitoring from its core, then years later added some infrastructure capability and it's just got into logging. Datadog started with infrastructure as its core value proposition, added APM and then added logging. What was kind of the approach that you took in terms of what you focused on first? Where do you think your deepest set of capabilities are and what else have you added?

So we actually looked at the cloud project as sort of requiring an integrated whole. So we built in from day 1, the notion that we needed to understand the virtual infrastructure, the virtual network and all the logging. At the same time, we needed to understand the application layer and the user experience layers. So we've really brought that all together in a unified whole. And that actually is what we've been selling from day 1. A full stack view, with all of those attributes and more. So it really was conceived as an observability platform. We didn't talk about it like that, that's sort of a more modern term for what these characteristics are, but that's the way we envisioned it, built it, because we knew that you needed this entire scope of observability to really understand how any part was working in this sort of holistic whole new software stack we called these enterprise or multi-clouds. So we've had it since day 1. What we've done is we've actually now started to generalize the platform into some additional adjacencies. So there's a number of hosts in a cloud that don't require application workloads: directory services, load balancers, communication layers, et cetera. We've now provide an infrastructure-only component that extends to all of these hosts that do not need applications so that you can bring a much wider, broader cloud footprint in under the topology mapping, the real-time topology mapping we do and the AI engine that watches the entire stack 24/7, providing -- understanding normal behavior, understanding when there's degradations or anomalies, being able to alert people to exactly the precise thing they need to go do, to take care of whatever that issue is or bump in the road that might be happening. So it's -- we've extended that way. We've extended functionality out to the edge, the user experience environment. We group multiple different capabilities there and provide sort of the outside-in view from the edge to make sure all cloud facilities are working properly, whether it's delivering transactional performance via an API or whether it's user experience for, say, a mobile application. So we've been extending and generalizing the platform now beyond just the core APM capabilities. But we always envision that you needed a full stack view of this entire collection of metrics, logs, traces, code, topology, et cetera, all collected into 1 environment processed by a common, unified data platform below.

So when you look at your customer base, what portion of them are utilizing kind of that trifecta of APM infrastructure, logging versus using your APM with perhaps logging from a Splunk, an Elastic, a Datadog or infrastructure from someone else?

So the way we look at it is a set of use cases, and so the real-time application monitoring use cases, we have our customers normally think of us as providing that full stack view for all their run the business application environments. So because it all comes together as a unified piece, when we talk about full stack application monitoring, they'll use us across all those. They'll still might use Splunk for their data analytic platform that's more sort of companywide. They might have some other infrastructure areas that they might use some other technology for. Every once in a while, we find certain application loads, normally more sort of legacy kind of application loads, Java stack loads on some other platform maybe. But when you get to that cloud environment and it's a full stack view and every -- and all these processes and services are dependent on one another, you really need a single platform view vertically and horizontally to be able to understand and have the situational awareness required to make sure that those applications work flawlessly all the time.

So we're getting a couple of questions from the audience around the competitive landscape. And I think you've done a great job in the past, kind of articulating who are the vendors that you run into the most? And relative to them, what do you feel is the competitive differentiation that you offer?

Sure. So the usual questions that most have on the back of their mind, one of them, let's say, New Relic. New Relic, we see maybe 20% of the time, more at the high end of their base, the low end of our base. So -- and that's been consistent for a couple of years, hasn't really moved one way or the other. They're in a different swim lane, a freemium kind of go-to-market swim lane than ourselves. We're a direct sales, Global 15,000 focused business. Datadog, since they come in off an infrastructure swim lane and again, freemium, that's a little bit different space. Very, very rarely see a Datadog. We do see Cisco AppDynamics probably the most, but that's declined over the last couple of years, 18 months, but they've taken an application-centric direct sale approach. So that's why we see them more often. When you get to competitive differentiation, it really comes down to we reinvented our platform to purpose build for dynamic multi-clouds. And no one else has done it. They're still trying to do instrumentation manually or via some kind of scripting. It just doesn't work when you get a dynamic multi-cloud environment. The scale is off the charts. They can't scale to the web scale requirements and the real-time monitoring requirements of the data center moving to the cloud. So these are just some of the characteristics, but it really comes down to -- we have a fresh new platform purpose-built for the challenges of today and the future versus a purpose-built platform for what the world looked like in 2008.

One of the other follow-up questions to that is, any sense of what your win rates are against those competitors?

Well, I share these every now and then, but we measure from the time we do what we call a proof of concept. That proof-of-concept is a test drive of Dynatrace on somebody's application set. That proof-of-concept for us takes a day, a couple of days, could be a week, some people like to test drive it longer to get more sort of fingerprints on it from different folks. But that's sort of -- we encourage it because our platform's so radically different, just -- you really have to experience it to understand how different it really is because the nomenclature from all of us is -- can be confusing, all sounds the same. So that's sort of a key piece of our program when we go forward or when we talk about win rates. So from that time, somebody's leaning in and we'll spend the time to really take a look, our win rates are in the 80% plus range and have been there consistently for quite some time. We run into politics sometimes, sometimes people buy on cost rather than on value. But for the most part, if they're open for how do we do this differently better and prepare for the next 5 years, Dynatrace is a fantastic platform.

So in terms of your sales motion, are you more of a top-down sell to the CIO? Are you more of a departmental sell to the DevOps and expand up type of approach?

Well, it's a little bit -- I'd say it's almost in the middle. We -- it's rarely from the CIO/CTO down. It's usually from a lieutenant to them, Chief Architect, whether it's a cloud architect or an application architect, could be a DevOps lead, who's responsible for that entire framework across multiple DevOps teams and their DevOps strategy. Usually is somebody who's application aware because they are the ones that actually appreciate some of the advanced technology we use for distributed tracing and code level visibility and things like that, that have always been part of our bread and butter. But it's in those kinds of folks. We don't try to do large transactions. We love land and expand. So to us, a landing zone is in the $90,000 to $100,000 range, which given our customer targets, $1 billion-plus businesses, that's a pretty modest spend to explore something that could be transformative to their IT operation. So that's where we land. Our expansion transactions are not really that much larger on average. Sure, we have some very large transactions when somebody goes all in with Dynatrace. We love them. But that isn't the core of our business. We think it makes us more resilient and more predictable from a growth standpoint. And obviously, we love it from our net expansion rate standpoint as well.

And you touched upon kind of the next area, which is the demographics of your customer base. It doesn't sound like you're an SMB provider in any form or fashion. Where is the sweet spot? And do you have any SMB exposure?

There's spillover into the SMB space, what we would call a $250 million company up to a $750 million company kind of thing, but it's very, very light because we really don't focus there. 90-plus percent of our sellers are named account territory focused on the Global 15,000. That's what we drive our entire go-to-market around. And we believe that, that base, it's partly because they spend more than the mid-market. They're more sophisticated, have larger cloud transformation projects. And we happen to be very good at supporting that level of run the business application environments for the larger customers. Great customer success team, strong services team and a fantastic enterprise-class product that works better than advertised.

Are there any adjacencies in terms of product feature functionality that you don't currently have that are just natural extensions that you see Dynatrace moving into?

Well, so the way I think about it is this, we built a platform that has multiple modules, multiple monetization opportunities. We're penetrated at about 15% of our target market, 2,400 customers out of a target group of 15,000. They all have more applications and they all have more modules that they could expand into. We can build a multibillion-dollar business just on the platform that you see today. And continuing to fill out those modules and get better and better at cross-sell muscle. That said, of course, we have more coming, more modules to the platform that we're working on behind the scenes. But I don't feel that they are required to build a large software company and continue to meet/beat our goals for the next couple of years. So we'll hold what's behind the curtain for another day. But no, I don't feel the urgency or the need to have something that we don't already have that we're building out today.

All right. I'm going to lead the witness a little bit on this one. One of the questions from the audience is, is there any kind of feature or item that you believe that is unique or that you are first to market on. And for me, just watching the space as long as I have, I think you guys were really the first to market with more of the AI, DevOps type of solution. We have seen others coming to market with their flavor of it, but kind of what is the traction that you're seeing in terms of the use of your AI capability and perhaps maybe just articulate what is it that it provides for customers?

Yes. So if I were to step back and sort of put a couple of pieces together because they all are intertwined, Sterling, as you know, the first one is we had to make sure our instrumentation and how we gather data was automatic, so that it self-learns, self-discovers and continuously it's doing that. Because it's discovering logs automatically. It's discovering all the metrics available. It's discovering all of the code level transaction elements automatically, stitching them together automatically. Without that foundation, you can't build something on top that's going to actually work in dynamic multi-clouds. The second thing that we did was we realized that the cloud was going to be infinitely variable. So there's no way to predict its behavior and therefore, building a learning system was never going to work. It's too dynamic, too changing. So we had to make sure that we built a set of analytics that understood all of the interactions between all of the components, we call it a topology map or a Smartscape, that gives us a known state in real-time of exactly how everything's working in a cloud, and then we apply the AI engine against it. It can be very precise. It does not need to learn. It works immediately, driving value, very, very different. So not only were we first, but I think we've thought through it much more completely than the others who are trying to bolt something on, on the side, a correlation engine or something that is just going to have trouble being anywhere near as precise and complete.

Makes a lot of sense. I want to bring us back to -- you mentioned the conversion over to the new Dynatrace platform and off of legacy. When is that process complete? And the question that I often get is, okay, you've shown some very good growth, but is some of that artificially lifted because of the conversion process that you've gone through?

Well, so the conversion process has been a great success for us as we move folks over. The sales organization's been involved. And during the conversion process, there's expansion, really pretty natural expansion when you move across into a cloud environment. There's more applications that are interdependent. So there's an obvious opportunity to sort of increase the footprint. We've taken that piece out and put it on the side when we communicate and said, hey, there is expansion at time of conversion. It's a 4, 5 points of growth, but if you -- even if you -- it's all expansion. But if you take that out and say, okay, that's a onetime thing as they're sort of moving their base from legacy to cloud stack, it's still at 38%, 39% ARR growing business. So it's still a healthy, very healthy growth, and that's because it's made up of a combination of customers converting as well as new logos. I mean, our new logo growth last year was 600 enterprise logos. Every one of them started around $100,000, but they can each one be $1 million ARR per year, we believe. I mean, think about a $1 billion company running their business, there's really a stack of applications who wouldn't spend $1 million in order to make sure they run flawlessly, if they're dependent on it for business and business transformation. So we're -- we have a long way to go. We're only at $220,000 average ARR per customer, a lot of room to expand as well as to expand the number of customers themselves.

What happens once the conversion is complete in terms of where sales force productivity and where sales can spend more of their time? And how does that impact the business?

Well, so I think there's 2 things. First of all, when the conversion is complete, the customers expand more rapidly. That's what we found. Because it's a much more automatic platform and has more modules around it to actually extend into. So we're seeing good traction sort of on both those fronts. From a sales organization standpoint, I mean they were spending -- we estimated somewhere around 20% of their time was spent on the conversion process. And with that now behind us, as we go into our new fiscal year, which started April 1, we free up a fair amount of sales capacity. Now we're also facing, there's going to be some COVID headwind, elongated sales cycle. So I don't know how much of that is additive or how much of it just keeps us sort of level. But it is additional sales bandwidth that gives us some advantages as we go forward for sure.

That makes sense. That's a great segue into COVID. What did you experience in the March quarter in terms of COVID impacts and what did you see in April, if there was any change?

Sure. As I said a couple of days ago on our earnings call, there was definitely a pause in the middle of March when everyone scrambled to take care of their employees, work from home and get that set up. But even with that, we had a solid close to Q4, consistent with prior year Q4s. And we've seen April and early May be pretty strong, a little bit stronger than a year ago. So we're pleasantly surprised with that. We -- gives me sort of cautious optimism that things will be -- that this will be a solid growth year. But I think we all know that when looking back 4 or 5 weeks ago, nobody was quite sure. Nobody was quite sure. I know that there's other companies who've said, feel very strongly that digital transformation may even increase going forward, accelerate. We do see a little bit of that. I think it's still a little early to claim victory on it myself. But the early signs are definitely positive.

So the 64 -- it used to be $64,000. But with inflation, it's got to be more than that, dollar question on investors' minds is, okay, you put up a good March. You've talked being cautiously optimistic about the trends you've seen in April and May, yet you took a decent cut to the guide. What was the thought process as you were building that guidance? And what were some of the kind of key underlying assumptions that you factored into it?

So our world, our new fiscal year started April 1. We didn't start in January. So we're not updating guidance. We're setting guidance for the year. And we wanted to be prudently conservative given the uncertainty because there still is plenty of uncertainty. If you think about just simple scenarios that we all think about or watch or could certainly build in our minds around a second wave, another dip in the markets. Depending on timing of that, what would that do to quarterly close rates, et cetera. We took, like I said, I think, a prudent, cautious approach figuring this is not just a calendar Q2 item, but it's a calendar Q3 item as well, and that it doesn't start to come back again until calendar Q4 and then normalize again as we enter 2021. So that's the way we looked at it. We tried to divorce our enthusiasm and confidence for the business. I'm an optimist. Try to divorce that from what's the right floor to set, that we can step up off of throughout the year.

One follow-up question that we got from the audience is that 20% of time focused on customer conversions, is that across the entire sales force? Or is that concentrated in a subsegment? And how many quarters does it take to kind of recapture that productivity?

Good question. No. So first of all, the conversions, because of the customer relationships that the sales organization has, it's not a separate group. It's the entire sales organization and some had more installed base than others. So -- and usually, those are some of your better reps actually, that have the more customers. So as those customers then convert, the sales team have been able to start to spend more time on new growth. It's not a total step function. But it still will take -- it still took us finishing 1 year where it was on their minds, to starting a new year where I took it off the list for them or we took it off the list for them, at our virtual kickoff 3 weeks ago. And now it's starting to gear up. And with a direct sales organization, it takes several months before you get the full impact of a change like that. So I think it's more of a Q -- for us, a fiscal Q2, calendar Q3 item, you start to see the lift, and then it really starts to come back strong in the December quarter.

Last question for you. In light of COVID in the overall environment, what are you doing with your own internal Dynatrace IT budget? Are you delaying any spending? Are you moving forward under original plan?

Well, we're in growth mode. So it's cautious growth mode, but it's growth mode. And the only things that we've ratcheted back a little bit are things like, well, I don't -- obviously, we don't need all the laptops we thought we needed right up front, we can stagger them through the year, things like that. But as long as the investment is a prudent investment that has a return on investment, that's in a reasonable time frame. And we feel like it's an essential item, not just something nice to have, which we don't do very often, anyway. It's pretty much essential if it gets past me and our CFO, then we're spending. So it's just a sort of a filter, maybe a little tighter filter around what's essential, but we're still, like I said, in growth mode.

No. That sounds good. With that, John, thank you so much for joining us today. We really appreciate it. Stay safe and stay healthy.

You, too. Thank you, everyone. Cheers, Sterling.