Datadog, Inc. (DDOG) Earnings Call Transcript & Summary

Thanks, everybody, for coming here, and we're continuing the momentum with Datadog. David Obstler, I've known you for -- when was the IPO?

I was -- I think 5.5 years ago, I've been at Datadog for 7 years. My anniversary was November 1, 2018.

It's -- time flies. Hope you are having fun.

Yes. It's been unbelievable, yes.

Well, I was just talking to Yuka and this last quarter was certainly an inflection from a numbers perspective. And I think the point is you guys -- in some regards, it's just business as usual. You guys continue to do what you're doing. But maybe from your perspective, obviously, the non-AIPs, obviously that and your largest customer were kind of the 2 hallmark parts of the quarter. But just walk us through sort of like the strength in the quarter, and I guess, specifically the non-AI.

Yes, definitely. We can go through this. There's 3 real factors. One, we've been successfully expanding quota capacity to cover more accounts. We've -- that's translated into some successive quarters of more new logos and more dollars of new logos, larger new logos. So this has been building. Next, the platform and the investments in the product and the continued ability to have clients adopt the products. We have a number of metrics we did in our earnings. That is the 3 pillars, plus I think we said digital experience, got over $300 million, and we also talked about some of the successes we're having in Cloud SIEM. And then the last part is that we don't have the headwind as in weighted average intensity of optimization. That doesn't mean there's no optimization. But when you look across the customer base, we have customers who have done a good job and I think we've helped them in terms of the platform, the product expansion, the way we've dealt with customers so that we have -- we don't have that kind of pushback in optimization that we had. So it's a more conducive environment. That's what's happening and that's been building. This didn't happen like overnight on a certain day. This has been building for a number of quarters. And hopefully, we've communicated that we are seeing those trends over time.

Yes. I mean, I guess I wanted to double click on the non-AI piece, the non-AI customer cohorts. I mean what -- like is there any way to put your finger on what -- and it's probably hard to answer what changed because it doesn't -- the answer is like it's hard to say. But like I think a lot of us felt the AI natives cohort was leading the way towards implying that as customers adopt more AI, more workload, more monitoring, is that -- are you seeing this as the sustainability of this non-AI piece is like the flip is switched that more AI workloads are going to production from this non-AI native cohort because it just -- it feels like there could be some durability to that element?

Yes. We think so, but I think it's a boring answer. But remember, a very large percentage, 70%, 75% of workloads are legacy workloads that haven't been modernized, rewritten and they're on-premise or legacy. So this probably AI is causing customers to focus on the fact that if you want to get utility out of applications, you have to work on rewriting them. We also believe that some of the coding tools and other AI is going to accelerate the creation of new applications. And I think that's what we're seeing. We're seeing more companies focus on the modernization of their tech stack. And as our platform has increased in sort of value, we're seeing our market share increase in the amount of ARR that's being put on our platform relative to other choices.

So the answer is it's not one thing. It could be -- trying to think differently about on-premise workload, it could certainly be AI workload, it could be Cloud SIEM. There's a variety of things there.

So in the portfolio, when you get to our position, it's -- the performance in the quarter is sort of like you have a number of different opportunities and the more of them that are going in a certain direction, they're going to add up to this because we're large enough geographically, diversified enough geographically by customer and all of that and product that you basically have a lot of efforts. And if they are heading in the same direction, they amplify themselves and produce what happened.

Yes. We don't have the answers of -- one of the questions that we get about AI is in an AI bubble, and that's not for us to sort of like think through. But like when you -- how important is AI workload growth within that non-AI cohort? I mean are we at an inflection point where like banks, insurance companies are all of a sudden like -- the answer, like I know there's more than just that driving results. Are we at a tipping point now where AI on the broader economy is starting to...

I don't think we're in a position to call that. We're sort of -- as you know, we're a follower, meaning we set ourselves up to monitor workloads. So we're setting ourselves to monitor workloads whatever they are. What we can control, you know from -- we can talk about it more. We're putting more AI in our platform, but it's still a small percentage of the value that we're able -- that we're conveying to customers. So I think we're still early on. I don't think this is a -- in the application of software, I don't think this is like we said, like inflection flip the switch. I think it's part of the overall modernization of the tech stack and a positive effect of that on that.

Okay. And then maybe just one last point here. You said that you felt the momentum was building. You've seen some of the early indicators previously that some of the stuff was coming around the concept of the durability of this. And are some of these trends that you're seeing now -- I mean, obviously, the macro can change in the future, but the durability of some of these trends that you're seeing...

Yes. The forward-looking things that we've communicated, one, we saw a persistence and continuation of the trends into October. We also said we have a very large and good pipeline in the fourth quarter and going into early next year. So that's what we can see. So again, we're not in a position to call durability or economic cycles, but it looks like things are continuing.

The other -- obviously, you guys renewed and expanded your largest AI customer, which was a huge talking point for a lot of investors. But what was really surprising, too, was the growth in all of the rest of the AI natives. I mean you're seeing some significant progress there with 6-figure, 7-figure customers. What are -- could you pinpoint what those customers are using you guys for like specifically? Is it monitoring known workloads, customer spacing side? Like what are...

Yes, exactly. Customers, same thing. So essentially, we've all created this AI native and non-AI native, but really, they're called modern software companies. So what do they have? They don't have a legacy infrastructure. They don't have on-premise legacy. Everything they're about their business model is a modern digital application. They're experiencing a rapid demand cycle. And like all of our other customers, we're monitoring their workloads in their client-facing applications. So what you'd see there would be metrics, traces, logs, so infrastructure, APM, logs, and then you would see digital experience, which we said we've got over $300 million, and you would see database monitoring and some security. So essentially, the difference -- the only difference here would be you have, in some cases, a faster ramp to significant scaled revenue than you might have seen in other software cycles. And they've been releasing their revenue numbers and their funding, but also their revenues. So that would be a good indication. If they're getting a lot more revenues, it's very likely they have more workloads. And if they have very -- more workloads and Datadog is the choice for these, that Datadog's revenues would be growing. That's our model. Nothing different.

One of the narratives really for the past couple of years is could AI kill software. And you guys certainly from a pricing perspective, are immune to seat-based pressure. When you look at how AI is impacting the infrastructure stack and when you think about open source technologies and things like that, I mean, how do you see AI evolving how customers think about the infrastructure or the monitoring layer in the future? And -- because there's -- you're unique and that there's no seat-based issue that you guys have...

I mean, it's also we're probably in a good position that in our infrastructure. So essentially, it really -- as long as we invest in our R&D cycle, as long as we -- and we set ourselves up to monitor whatever it is, that's what we've always done, whether it's serverless or containers or anything. So basically, we -- as long as software is going to be delivered, if it's more gentic, if it's got -- whatever it is, we're setting ourselves up to be able to meter and monitor that. And that's a good position to be in because essentially, we're not seat based, but we're also agnostic to what's being -- what's in the software. We're just going to make sure that we're able to monitor whatever the composition of that software is.

Yes. And then I guess the -- with open source alternatives or DIY, from a customer lens perspective, there's always -- and forgetting your largest customer, but what are you hearing from customers in terms of the criticality of like, oh, yes, sure, I can go do this in an open source model here or there, but like is it really -- why bother...

No, they don't want to do it. I mean that's how we got to our size. That's why we're creating versus the -- we're creating like there's other companies, we're creating those other companies in every month or quarter because they don't want to. There's tremendous value in having everything in a single pane of glass and a platform and it costs a lot less. And you're able to get the best-in-class and you don't have to put your resources that you're doing in your own business. So that's why we don't -- that's why there isn't much in-sourcing. That's why -- sorry, the outsourcing has been the trend, which is one of the things that's made Datadog. With a few exceptions, that's really what is driving the market. I don't -- it's not -- it's been moving since we went public more towards a purchase of a service or a software than in-sourcing.

Yes. And then maybe just one other kind of AI question. You sort of alluded that the AI native disclosure served the purpose. No how do you guys gave it. I forget how long ago it was first disclosed, but effectively implying that are we nearing the end of that is because you think about just the broader economy now. Is that something that we should kind of be prepared to...

I mean, we're saying -- like I said, it's a made-up thing because it's a software company. So it was done, one, to show that -- to talk about our growth drivers. In other words, to talk about our acceleration, you'd have to talk about the growth of the AI native companies. We would say that anyway, right? That's one of the growth. And then there was the factor of the largest company that we wanted to make sure we helped everybody along. So as everything evolved, we could have a discussion about traditional enterprise, AI native and the larger company. Eventually, we believe that all of this is going to be part of the overall software cycle, and it won't be as meaningful. And so we haven't given when we stop -- we're going to stop disclosing it. We're trying to help everybody understand the drivers of the business. And so that's the utility of it, and we'll look at that in future quarters and see if there's utility.

Okay. The other thing that struck us is the acceleration in your security business, which is -- it was a long time coming as well. Can you talk -- the question is like why now? And I'm sure the answer is it's been building. But can you help us think through some of the why now piece?

Yes. I think most of the original, and it's not surprising security revenues were created in SMBs or smaller growth companies in DevSecOps where essentially security was much closer to development. And that's what we did, and that makes sense because as you're developing your product, and that's where one of our core strengths are. But I think we really didn't have the enterprise selling motion. But what happened was we got in Cloud SIEM, we got capability. We got parity and capabilities. A lot of that had to do with the work in taking logs and making sure logs could be used economically for various use cases. We already had done an unbelievable job in observability logs. And observability logs are very close to the real-time production environment. So what we did was with Flex Logs and Frozen Logs and all that, we were able to modify the product. And then we basically saw that in the competitive environment, someone like Splunk, and this has been from a number of different perspectives, there are a number of companies was -- essentially, there was a whole set of workloads where you could get better visibility at a lower cost from a Datadog product. And then we got, I think, better over time about selling in enterprises and attaching to our logs customers. And that includes essentially program management. We've been better about buyouts in exchange for long-term contracts which hasn't changed our profitability. It's just spreading the discount in a different way. And then there's a lot of other things like migration or from Splunk query language that we had to conquer. And I think we just got better programmatically. We're not there yet. We still have a lot more things to do. For instance, we set up a channel motion. That takes a long time. And so a lot of what's being created is direct. Some of it is starting to come through channel, but there's a lot more things that we have -- that we're going to be doing or we've done that have a longer investment cycle. So that's kind of what happened in security.

And SIEM seems to be a big part of that. But when Oli and the team think about the white space, the shift left mentality, DevSecOps, broadly speaking, what sort of excites the team about future opportunities in cyber?

Yes. I mean I think you have code security as well where I think we're starting to do well, looking at the security in the code as it gets put in production. I'd say that has tended to be a smaller market than the other 2 than Cloud SIEM and Cloud Security. Usually, I would say, developer-led market where you start -- you go to the developer, most of what we sell is into the production and the SRE. You just have -- you have a little bit of a different TAM size, but I think we're going to be good at it, and we're getting fully capability there. So I think that's another opportunity for us.

The thing you guys talked about it, you launched it earlier this year, and you talked about a little bit on this last earnings call, but Bits AI seems to be resonating. Talk about that opportunity? How additive is it to a customer spend? And sort of like where do you see that driving additional customer value?

So Datadog has always been an aggregator of data, the organizer, you see the picture. It is very good at diagnosing where problems are. And we've increasingly moved into what are you going to do about that? How are you going to route cases or make recommendations? And I think AI is a great -- has been and is a great opportunity for us to sort of revolutionize that. So that's part of like the service management suite where you can diagnose problems more quickly and then have -- and then route them for resolution. And right now, we're still in private beta. I think we haven't said to everybody where we -- when we're going to be in GA. But we are getting a lot of pull-through, meaning we're getting a lot of customers who are using it. This is how we launch products. And we're getting a lot of excitement about it. And we think -- and their customers is paying for it. And so we think it's a good opportunity. We don't know yet whether it's going to be monetized through increase of win rates, staving off other entrants to the market, increased workloads or through a SKU, but we know it's important to the economic -- the economics of the company.

Yes, that's great. The new engine flywheel of products has been really a hallmark of the company. And I guess the other thing that stood out in addition to some of the broader acceleration that we talked about was new logos. Logo adds were up dramatically. I mean this, I assume it's a similar answer. You've seen building momentum here. Could you put your finger on like why now and sort of like -- because obviously, there's a lot of turmoil out there...

Definitely a combo of successfully ramping quota capacity. So we've been able to maintain productivity, maintain strong CAC return and ramp. And I think we're landing larger. So it's also a size because of the platform. So we're basically going into places where I think we're better at and we're getting more lands that are larger. Remember, we used to say that, we don't try for large lands, we're bottoms up. Well, we are bottoms up, but we also are increasingly getting more of a top-down and more of a here's the whole platform or more of the platform. So it's that combination, I would say, capacity and platform adoption that is causing that, which are good things because those are things that can be sustainable.

Yes. Yes. The -- I wanted to -- I guess, sort of in the same lens, you guys have been such an organic engine over time, and you've made some smart tuck-in deals. But I think you've got about $4 billion in cash now. How do you think about, broadly speaking, deploying that? I mean, could -- as you guys get bigger, could the size of M&A deals increase? And how do you kind of think about how M&A is part of the kind of the broader R&D?

Yes. I mean at the heart, M&A for us is a part of product management. It's taking a look at the products we want to create in the platform and/or the talent needed to create it. And for the most part, we've acquired teams that have some commercial adoption, but aren't going to have the platform or the go-to-market machine that we have. And that has been the majority. That could get larger. It could be that companies that are a little farther along. We're open to larger acquisitions. I think we answered some rumors by saying we're not working on anything major. The bar for anything major is really high because we essentially -- we're not a consolidator. We don't take companies that are lagging in technology. We don't want to do that. It basically takes us off what we're concentrating on. So we're open. So I think that amount of money is good to have, but it's not that much money that you can't do a couple of larger deals and not use parts of it. So -- but I think for the most part, the center of our acquisitions, whether it be slightly larger companies or larger companies is really about that product map. It's really led by product management.

And I assume part of the high bar means deeply and natively integrated into the platform. You don't want to necessarily have a Frankenstein model where there's different bits hanging off the side...

Tech leadership, so not decaying modern software. Consistent with Datadog, here's another one that's kind of hard, a team that wants to stay and be part of the leadership of Datadog. So that is a really important one. We're not interested in acquiring a company to make the founders rich and go away. We're interested in them being part of Datadog. That's not a bar.

In terms of U.S. Fed, it could certainly be more for you guys. How do you think -- when you think about FedRAMP, elevated levels there, how should we think about that as a contributor to kind of the future growth algorithm?

It's small right now. I think we get asked sometimes the government shutdown now or non-shutdown, it doesn't really affect us. I mean we don't have that much government. So it's an opportunity. It's an opportunity that I think we're going for FedRAMP high. I think that will enable us to go into Department of Defense and others. The government does not move particularly fast. And when you call -- for Datadog, you need modernized software stack and cloud. So that's slow. I think it's a long-term opportunity and one that will be -- augment our business. We're certainly under punching our weight. It has a lot to do, I think, mainly with the end market. But -- and we're investing behind it. We think eventually, the government will modernize more. We want to be there when it does with the right capabilities.

Yes. If you've been to the post office or driver's license, it certainly could use...

Could use. Yes. I mean, essentially, if there's -- we want to have it set up so that we can handle the most important government workloads, including defense, where a lot of dollars, where the budget is going and then we have to follow the trends of what they're doing in their tech stack. Yes, so that's it. So I think it's an opportunity that's small for us and could be a lot bigger, and we think it will be a lot bigger, but it takes time.

Yes. When we think about going into Q4, you guys are such a predictable subscription model that you don't -- budget flushes don't really impact you. When you think about how customers are thinking through kind of like AI budgets, I think one of the prior concerns was, is there just confusion around how customers are thinking about AI dollars and deployment. Do you get a sense that the customers are more comfortable now in kind of that AI spend envelope and that could help open up opportunities for increased use of Datadog in the future?

Yes. I mean I think we said our pipelines are strong, and they're diversified across a number of very interesting and traditional industries. Yes, I mean I think that it will. I think that there's -- when you're talking about a business and an application, it's not all -- it's not like 10 -- like no one -- it's not like everyone is waiting for whatever the AI thing is. They're still doing their business. AI will be part of the application increasingly. But I think we're communicating that we're in a good buying environment, and it's due to a number of factors.

Yes. I'm going to ask you one more and then we'll see if -- this is a big group, we'll see if there's a question out there. In terms of -- you and I have had this conversation about pricing and packaging. And I think you feel that you guys have the right framework to get you to multiple billions of dollars more than you are now. And obviously, there was some -- talk about your largest customer and some -- maybe some bespoke things going on there. But do you think as you get these larger customers over time, that the pricing and packaging has to evolve for a tier of customers that are growing at a rapid pace and consuming a lot more of the product? Or do you think, no, like this is the model that will get us to customer spending...

I think broadly speaking, it's the right package. It's volume and term discounts. I think there will -- there may be in the future, and I think we said it, different types of workloads, whether it be, let's say, on-premise workloads or that need to be priced differently. And I think we said we're working on it. And I think when you think about Frozen Logs, Flex Logs, log without limits, metrics, I think this is all what you're saying, which is it's all evolving the tech stack to be able to do differentiated pricing to address this. And I think that's what's been most important. And so we think broadly, but we're working -- we work on it all the time. We have new SKUs. We have new -- we divide those SKUs. We have cost of delivery, if you don't need the logs, it's a different price. So we're pretty disciplined on looking at the gross margin and making sure if we can lower the cost, if it's not growth pricing, if you don't need the logs that -- and we don't have to retrieve them or start whatever, then it's a different price. So we're working on that.

Yes. Great. Is there -- we may have time for one out here. If there's -- anybody has a question for David? Otherwise, we can keep rolling. We'll scan, scan, scan for logs. We're being thorough here. Yes. When you think about some of the building blocks for next year, and when you think about growth and profitability, obviously, you've done a great job balancing both as the company has scaled. What are some of the things that you would sort of like -- when you think about some of the big moving pieces without -- you're not going to guide, I understand that. But like when we -- how should we think about the evolution of this model as you continue to scale and profitability matters, but also you're seeing growth inflect at the same time?

Definitely. So it's bottoms up. So in terms of go-to-market investment, we're looking at territories and target accounts, and we're looking at accounts that could buy Datadog. And we're sort of making bottoms-up investment. We're still -- I would say, we're not saturated. We still have lots of accounts that could use additional coverage. We're looking at the go-to-market in a holistic way, whether that be enterprise marketing or channel investment and things like that. Looking at data centers, and that would be, do we have enough of a TAM in a market that needs data residency. And so we look at that. And then when it gets to R&D, we're looking at what the platform investments and product investments are and how we have to sort of staff them. So that's how we're sort of balancing all that. I think as to remind everybody, we've given a long-term EBIT target of 25%. As you know, we've been past it. We've been under it. But essentially, we've been in the low 20s to the mid-20s overall. We use that discipline to help us to calibrate the level of investment. It is definitely easier as you get scale. You can put a lot of money to work productively because, for instance, in R&D, 30% of our ARR is over $1 billion, a lot of money. And so I think it becomes increasingly -- the dollars are there and the impetus is increasingly on what to put them behind.

Yes. Maybe then just to wrap, when you and Oli and the executive team sit around and you think about some of these -- the big, big opportunities for Datadog, where you sit here and say, this is a moonshot for Datadog. I mean do you have any sort of like bold like -- not financially, but just like longer-term predictions on while this is...

Yes, our vision is to have a platform that addresses our core user group and all of those around it with increasing functionality. I think service management is an example of that. So you just don't -- you actually move towards resolution in the platform. I think security is an example of that where you get additional buyer groups or user groups so that you do what we've been doing all along, which is you expand the TAM by having more and more parties in the platform, using the platform and creating value. That -- I think service management, security, the AI, the bits AI, making the platform AI and covering AI workloads are probably at the top of the list and maybe I said security and class, a top of list of what could be breakout opportunities for us.

Cool. Well, out of time. Look forward to you hear from you. Thank you, David from all of us.

Thank you, everybody. Thanks for coming. Really appreciate it.

Thank you.

Thanks.

Thanks.