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

Yes. I mean, another -- not to get into too much detail about it is our pricing mechanism with our Dynatrace Platform Subscription, DPS, is another part of that. We used to price some number of years ago based on SKUs. So you had to buy a certain amount of APM and a certain amount of infrastructure monitoring or certain amount of log management, et cetera. And that is just not how our customers want to consume the platform at this point. They want to consume the platform holistically. They want to -- maybe they need more log management or less month by month. Maybe they need more or less APM month by month or full stack or whatever it might be. So the contracts that we enter into now based on Dynatrace Platform Subscription are up. You just make a commit $1 million or $10 million, whatever your commit might be. And then you can draw it down based on the consumption of the platform as you see fit. For example, if you're an e-commerce customer, you're presumably going to be consuming way more of the Dynatrace platform for observability during November and December than you probably are during February and March. And so it enables to smooth out your usage of the platform and how you're using it, where you're using it. DPS has now moved to 75% of our ARR, more than 2/3 of our customers. And so this does also provide a catalyst to new customers coming on board saying, "Hey, look, I'm just -- here is my overall commit. I'm going to draw it down over the course of time as I see fit, and as MOs need it, and that's going to enable me to really deploy and use end-to-end observability in a notable way."

Yes. now logs. Now you've expressed a lot of confidence in the logs opportunity. You guys talked about doubling this year before this quarter. So you've expressed that you had a lot of confidence, and it sounds like you have a lot of visibility. So can you just share a little bit more like why you have this confidence, this visibility? Like -- it seems like when I was at the Dynatrace conference earlier this year, like talking a lot of customers that they have logs, but they were early in their journey, and they have a lot of pipeline of incremental workloads that they need to onboard and get more users on to the platform, looking at logs. So just what does the journey look like with a customer? Is there a long runway when customers adopt logs and migrate more workloads and whatever else? Just what's driving the confidence there?

There are two enormous value propositions to using Dynatrace for logs. And logs are very well established. Everybody and our target customer set has been doing logs and doing logs for years with incumbent vendors that are sizable. So this is mostly brownfield, not greenfield type deployment. And so what they're looking at is, geez, do I shift my logs to Dynatrace, not do I bring up new workloads on Dynatrace with logs, although there's some of that as well in the AI workload space, then those might be net new workloads. But there are a lot of workloads that have existing log usage. There are two primary drivers of using Dynatrace for logs relative to that environment. One is cost. We can radically reduce your cost relative to what's on the market. I can't tell you the number of customers I've spoken to at CTO, CIO level that have come back and say, "My god, Rick, my log cost is meteoric. I mean, I've actually heard that word before, log cost is meteoric. And it is because of the volume of logs in addition to the overall cost. And there are multiple ways where we can help solve that. Number one, we can give you a lower price on logs because of the efficiency of our platform. But the second piece is that it turns out that with our Bindplane acquisition, we can now filter those logs on an inbound basis so that you don't have to ingest them all into the platform, which if you ingest and store those logs has higher cost. So by enabling a better filtering system of those logs, then that can reduce that further. And finally, it -- we would submit that you don't need all the logs if you have traces, metrics, really user data and other data because the combination of those data sets actually gives you a much better depiction of what is happening in your environment that if you just basically drive logs like crazy and that's the only mechanism you're using to do observability. And so for those reasons, we can actually reduce cost. The second piece sort of segues from my last point, which is that incorporating logs end-to-end observability gives you better analysis, better insights and better analytics. So the outcome you're getting from observability is better if you have logs as a component of it. Yet for since I've been at Dynatrace, gosh, now almost 5 years. And it never made any sense for me to have logs over here with 1 vendor and traces, metrics, really user data, all of the other observability types over here that made no sense because you get better insights and better outcomes if all of those are embedded in an end-to-end observability solution. So bringing all that back to the simple answer, simple answer is I can do it at lower cost, often substantially lower cost than what you're paying for today. And I can deliver better you outcomes in a holistic end-to-end observability environment. That then, as I said earlier, establishes the foundation for a agentic AI and agentic solutions driving autonomous operations to come.

Yes. And then lastly, on AI to come back to that. Maybe we talked about technologically like why this is important, how the world is changing, but tie it to one, just like, what are you observing within your own customer base in terms of like AI maturity? And then secondarily, kind of the financial contributions that you're starting to see, we talked about in the call, 1.5x customers that are using AI or monitoring AI are consuming 1.5x those that aren't. So what's the maturity of the customer base? And then secondarily, why are we seeing an inflection or much higher growth with those that are adopting Dynatrace for AI?

Well, we -- in terms of a maturity and AI deployment in production, we still think we're in the early innings. It really, in many ways, has only just begun. Having said that, we did provide some metrics on the call. For example, we're now observing AI workloads for more than 1,000 customers. So that's pretty sizable. We are taking agentic actions to actually deployed Dynatrace agents to take action to do things like auto remediation and triage in over 800 customers. And then lastly, you mentioned the 1.5x. We are seeing, and we can take cohorts and look at cohorts of -- AI cohorts, non-AI workload cohorts. And we see 50% higher consumption of the platform of the AI cohorts relative to the non-AI cohorts. And in part, this is simply because AI workloads are pretty chatty. They generate tons of telemetry. And managing that telemetry and then ingesting that and then taking that into a consumption model that's been doing a DPS drawdown, Dynatrace Platform Subscription drawdown, is something that we're seeing. And it turns out all of those lead to various monetization levers for us in the AI space that are incremental to the normal monetization levers we see. First of all, drives more consumption. It's great. Second of all, it drives AI observability, which are new categories or new capability of answering that question I mentioned earlier, which is it right? Or is it accurate? And thirdly, anytime you use a Dynatrace agent to take action, we'll pass through some of the charge of that because we have real cost there, too. So multiple different ways to monetize AI workloads incrementally.

Right. Now DPS. Now this has been something that's been compelling for this year, coming years. We have 3 cohorts of customers up for renewal. Can you just talk about maybe -- look, it's the smallest quarter of the year, but it's the first quarter of the year. And what activity are you seeing in terms of the fiscal '24, fiscal '25 cohort of DPS customers come up for renewal, how should we think about that opportunity for the balance of the year?

I think the primary way to stick away and just to make sure that you're all tracking that since we put DPS in place 3 years ago -- and most of the DPS contracts are 3-year contracts, even though they have annual resets -- what happens is that you're now ending the end of that 3-year period for many customers, and those customers are having to renew. And so the question is, are they renewing based on their consumption of the platform, flat, up, down how would that play out. And the good news is we talk about consumption. We mentioned on last week's call that consumption continues to grow north of 20%. And so the result of that when you have 17% ARR growth as we positioned or as we indicated last week based on the Q1 results inclusive of our Bindplane acquisition. And your consumption on the platform is north of 20%. Over the course of time, you would expect ARR growth to migrate toward your consumption growth number. And so the question is how does that happen? And is this DPS renewal mechanism that begins to kick in is one of the ways in which that could happen because you need to increase your contract commit in order to manage the consumption growth that you're seeing. Now Q1, Q2, probably not too much impact there because those are really the first couple of quarters of DPS. And so it would be a modest number. But as we get into the second half, the expectation is that those numbers increase, and the result of that should be that we would begin to see a positive move in NRR, that the expansion business or expansion metrics would begin to tick up as well. Pretty sticky metric. So it's not like 110 just goes to 115 or something like that. But we would expect to begin to see some accretion of the NRR metric as we go through the year based on these DPS cohorts becoming larger.

Right. So it's just a lag effect between the consumption activity versus the renewals and the NRR, kind of. Right?

Yes, you could put a DPS contract in today, and you could overconsume your contract in the first 3 months, and you wouldn't have to do an upgrade of that at that point if you didn't want to. You could just pay based on consumption.

Yes. Maybe last minute. The other really compelling interesting part about the observability stack is the ability to automate it to the AI kind of, I don't know, AI for observability, Dynatrace Intelligence. What's been the traction at this point? How has the customer feedback been in terms of the value that they're getting from this? And ultimately, like how does this benefit the model from an uplift perspective and monetization strategy?

Yes. I mean, I mentioned it a few different ways. We can monetize AI workloads incrementally, so I won't go back through those. But I would say that it would come as no surprise to anybody, certainly in this room or this virtual room, that every organization we would sell to is as AI very much top of mind. They're driving AI workloads, they're driving agents to create code. They're working to accelerate application delivery and beyond. And they are realizing rapidly that you simply have to have an observability solution that can oversee those workloads. And so that is -- we certainly expect that, that does drive material momentum in our business and the business opportunity we have with these customers, and we see it through the early innings of the monetization layers. And some of those metrics we talked about, the number of customers we have observing AI workloads, the number of agents that are being used across the customer base accelerate the consumption. So those are sort of the key leading indicators. DPS is the vehicle to achieve it. And then these incremental monetization models that are unique to AI workloads become additive in terms of capabilities. And candidly, Eric, we haven't really built those incremental monetization layers into even the guide. So we see those as -- we don't want to get ahead of ourselves in changing guides. But we simply don't know how rapidly that gets monetized. And so we've taken a conservative approach to that, but we do view that the AI areas here become critical. And AI observability is absolutely is a $10 billion incremental sector or category to the $80 billion or so overall observability space is going to be phenomenal, and it's growing at 40%, 50%. So this is going to be a new area for us to monetize that is mission-critical to the evolution of the observability market, which is where you started.

Yes. Super exciting. I love all the opportunities [indiscernible] ones in this space and observability is top of mind for sure.