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

All right. Terrific. Thank you for tuning into our fireside chat during Needham's Tech week, I'm Mike Cikos covering the Infrastructure and Analytics Software Base at Needham. I'm pleased to be hosting the CFO, David Obstler from Datadog here today. Just for some quick logistics. I have a list of questions I prepared on my side. But if clients at any point, have questions, please submit them, and I'll be sure to get those in front of David, while we have him here. With that out of the way, David, thank you very much for joining us at our conference today. We really do appreciate it.

Thank you for having us. We appreciate it as well.

Absolutely. I wanted to jump right into it if we could. I think most people, I would hope are relatively familiar with the Datadog model today. But one of the things that struck me -- I know we're still -- earnings is not that far, though, even though it feels like it was many months ago at this point. In the current environment, though, it feels like Datadog is seeing maybe less pressure on growth versus the earnings prints we received out of the hyperscalers. And I don't -- I think a decoupling might be a bit extreme, but it does feel at the margin like you guys are faring better versus what the hyperscalers are delivering. And I just wanted to see why is that the case today versus maybe where we were during COVID-19 and the pandemic. What's the difference? Or what's taking place under the hood that's causing that at the margin again?

Yes. I agree also at the margin, and it's too early to know how all of this is going to pan out. But I think there's a couple of things that are potentially worth pointing out. One is, I think we're indexed to the highest growth segment of cloud workloads. In the cloud numbers are a lot of different use cases and our use cases is a DevOps modern development, monitoring of client-facing applications in real time. And when I say that, I mean that's the business. So that's the way at least in that part of the business, the clients delivering their product to their clients. And so it's a high priority. I think that's manifested in the gross retention rates that have been quite high. And in terms of -- we're probably more indexed to the movement of critical applications to the cloud than the weighted average hyperscalers. So that's one thing. The second would be that our platform has expanded quite a bit in the years since we -- it seems like a long time ago, but I guess it's 2.5 years ago since the beginning of COVID where we had, as you know, started with infrastructure had expanded. But now we have expanded the SKUs. We have very strong penetration of the platform, adoption of logs, APM, Synthetics, RUM, some of the newer products. So that provides probably some cushion against a cloud optimization or host optimization and also provides a diversification of the way the clients can use the platform so that as they're looking at using the various pieces, they can move from one to the other in a frictionless way and that may also provide some protection. Again, caution, it's still too early to see how all of this plays out, but those are some of our observations.

Right. I appreciate you calling that out. Like one of the things that we've highlighted to our clients, but the idea that Datadog really over the long term looks at R&D is like the source of strategic differentiation. And so I'm happy that you're calling out, let's say, this platform expansion. Maybe let's focus on infrastructure and more geared towards some of these newer areas like logs, APMs, Synthetic and RUM, focused on those mission-critical workloads that we're talking about as well. I did want to also talk about those more, I guess, critical workloads. Again, is it fair to think that maybe some of that insulation from the broader headwinds that the hyperscalers are facing, like the cloud opportunity is so large. And correct me if I'm wrong, but I don't want to mischaracterize, I think about the investment in observability is almost like a second decision. So if I move workload to the cloud, I'm only investing in observability behind that. And so again, because you have this large greenfield opportunity to attack, that should also provide a little bit more of a sticking point for your durable growth over the longer term.

Yes. I think we have -- that's sort of at the very core of the anchor. So that type of migration. We also have a couple of other things we have the platform itself. So essentially when over time, when a client does land and it does, and we give statistics on this, and it does expand. That provides a growth driver as I think you might call it, consolidation, we would call it better use of the platform. So that's one. And we've also expanded over time, sort of the use cases. And we also see growth of the DevOps function. Security is a perfect example in DevSecOps, we can talk about at some point where our customer base has more needs over time as they -- as the world is seeing that the monitoring in real time of the applications is quite critical, and there's more than monitoring the infrastructure. As we've all learned, in order to do this well.

That's great. And just while we're on the topic, I had buried in the questions at the end. You had mentioned DevSecOps, right? So I know that security is something that Datadog has spoken to as well. Can you help us think about the pulling in? And I just want to set the tone here. Like when I think about DevOps, we've been talking about that since I want to say that term was first coined, like 2008, 2009. And that still seems like it's very early in the adoption of DevOps. And now we've been talking about DevSecOps since maybe 2014, 2015. It feels even earlier, but it does feel like it's being incorporated and pulled in. So maybe what are you seeing on your customers' behalf as far as their integration of this workflow and adoption and movement towards a broader DevSecOps culture and mentality.

I think you stated it very well. It's a good analogy to DevOps when you think about the way software development was done in waterfall, et cetera. And as you got towards containerization and the speed of it, you couldn't do that anymore, and you had to go towards more real-time development and deployment of software. And so I think you're on a similar -- we think you're on a similar journey in security, meaning that it makes a lot of sense that earlier you push security factors into the development and deployment in production, the more secure the environment is the more proactive it is. And it's very similar to the other signals that we are providing. We didn't invent this. We got this from our customers who are demanding this. And so we do. We said we've had very good traction. It's early days, but we've had that group that's progressive and is using security in DevOps, pull it through. We are still early in -- just like in APM, we're still early in the -- we were in APM. We were still running the development of this product. So there's a couple of things going on. The building of it that we're doing and putting it out there and testing functionality, and we're kind of midstream on that. And then as you rightly said, the adoption of security practices and DevOps, which we're seeing trends of -- still early, but we're seeing good signals. And what that means is there is a possibility in the likelihood that security could be adopted the same way that our other tools have been adopted, which is bottoms up, real-time, distributed, democratized. We don't know and we said we don't know what the influence of the CISO will be in this frictionless adoption. And we're open to influencing whatever way it happens. But we see some good early signs that there could be roots taking hold and that this could be a long-term positive accelerating trend.

Awesome. Awesome. And there was another comment, I guess, independent of this, separately on the earnings call, which we spoke to the growth across products being a more homogeneous during Q3. And I just wanted to make sure I fully understood the comment, but can you provide a flavor for why we're seeing. Is it more a function of customers are adopting Datadog as this platform rather than maybe a specific product, and that's what's causing this homogeneous impact? Any color there would be [indiscernible].

Yes, a couple of things. One, first of all, I think what we said where it was, we said that the rationalization was concentrated, not to say everyone doesn't watch costs, but was concentrated in the cloud native, fully scaled. And then in certain sectors, we can get into this. And we said an example was -- and I know you have a question on this was consumer discretionary. So -- and we also saw, on the other hand, we saw new workloads, whether they were created in the quarter, new customers or customers who had adopted Datadog in previous quarters. We saw them continue to ramp quite significantly. And so that led us to the conclusion that the pushing -- putting the priority workloads hasn't stopped. It's still a big focus. And -- but if you've done it and you are using more pieces of the platform and you're more fully scaled, you might have the opportunity to rationalize. Now how is that done? Yes, I think you said it quite well. We have a platform, and our clients buy credits or capacity. And they can move around within the platform on what they spend. And I don't think we're seeing that most customers are reducing their spend. It's that they're slowing down their expansion. And so they can do that in a number of ways. They can do that by changing the dollars on logs. We pointed that out last time. They can do that in terms of what data they store and they can also do that in the rate of growth of host counts of what some of our products are based on. And what we saw was we saw a little bit more towards they're doing that not as focused on the sort of the logs and we said a log like part of APM, but they're doing that not as intensely there because that was probably the lowest hanging fruit to look at first for the most affected customers, but they're watching it in the platform use. So we saw a little bit of that in Infra nothing significant. And we just pointed that out because on the previous call, we had said sort of the easiest parts of it in some of the logs and 1 part of APM had been the focus in that quarter.

Great. Great color there. And I did -- this is, again, for my own understanding, but like if I go back 2 quarters ago, we did cite more specific impact on maybe logs or some of the components within that APM suite. Why is that considered or viewed as maybe an easier area to optimize or lower hanging fruit as an example?

Right.

Where -- why I optimize there first? Is it just easier to modulate that? Or, any color.

I wouldn't -- I don't say if you look at the logs growth and the piece of it in the platform and how logs is used to investigate problems, in no way does it mean that it's a less important product or that -- and we don't have -- our growth retention rate there isn't changed. But what it is, is that since we're distributed and were used by lots of DevOps, we don't charge by seat. It can be used -- and so it can be used in ways that can be modulated in logs that where you set maybe the retention policy a little more conservatively or liberally what to be caught or you put out the directive that you watch the amount you're indexing because this is part of everyone's days of production. And that doesn't mean that -- and that's what happens all the time, meaning that's how clients use our platform. They do a lot of development, they -- and this is not related to COVID or related to the best of times are related to where we are now. This is how clients use the platform. And then they look at use. In fact, we help them look at use. We show them used. We don't want them over using the platform. We are land and expand. So we're not a sort of a shelfware company. Most of our clients are short and they buy along the way, that removes a lot of the friction. Our clients like it very much. So we help them to say, it looks like you might have spiked here and then see what -- maybe you have the setting wrong. So that, I think, is one of the more attainable things to do when you first start looking at this cost line.

Right, right. And if I jump back up to maybe some of the growth components, but maybe from a customer level, right? One of the things that I look at is, let's say, the new customer acquisition machine that you guys have, are there pockets within the opportunity that Datadog is addressing where maybe it's easier to land new customers by vertical or specific product today versus maybe where we were 12 months ago. Like has -- have you guys noticed a change in maybe ROI or what resonates as far as attracting that first purchase by a new customer in today's environment?

Broadly, it's quite similar. We do have some changes in that. We have in advance the government part of the business by getting our medium compliance. We have HIPAA. So there -- when you talk about sectors and industries, you do have some [ several ] things. But no, I think that -- it's that -- the development of the product and the leadership, as you said, from R&D has continued to make us a compelling call as cloud workloads are migrated, and it continues. We have, over time, I think, had a broadening from the cloud native. And this is not just now, this is over the last 4 or 5 years to many other sectors, financial services, industrial products, automobile, air transportation where you see traditional enterprises. And if you look back through our transcripts, we're going through a lot of those cases all the time to illustrate this. This is not something that's changed in the last quarter. This is something that's been going on for a number of years, and we see it continuing. We think we're really early on in the traditional enterprises adoption. And so we see -- and this may be one of the things that's keeping the momentum going and new logos because we have a very strong enterprise pipeline and mid-market is that we're very early, so these are priority projects, and they've been invested in are continuing. It's too early to tell. We may have -- we do sell both in terms of the efficacy of the platform in shortening remediation time and keeping critical revenue-producing applications up but we also sell on the basis of its cheaper and better quality to buy our platform than to do it yourself and do open source. And as IT hiring, our IT headcount, gets challenged. It may be we see some of this. We've seen a little bit of it. It's too early to tell and we'll keep everybody informed if this becomes an accelerated trend, we think it's rational that it's more efficient to buy the platform than to have the IT department. And we are not in that [ observability ], create another observability platform out of the components but we understand the logic of that, and we'll have to see how that pans out.

Great. Great. And I know that you had touched on it briefly there in the comments, but like when we think about you guys versus maybe the DIY or the open source, a lot is made of competition in this market. And I just wanted to see, can you put some better guardrails around that? Like how much of the market are you -- or how many new customers are coming to you that are maybe coming from that DIY or open source background versus a more competitive either I'm going to stand up data do next to this other tool that I already had from a third-party vendor or maybe I rip it or replace. Like how do we frame that greenfield versus brownfield with the new customers you're attracting?

Yes. I think it's not a rip and replace. We don't really go for that. We do have momentum that comes more naturally to the platform once we land and are used as contracts are up. But I think if you want to combine open source and piecing together various components like using componentry I think that's what you would have in situations that are -- that would be the more common decision than ripping out a platform competitor. So we have that. So I think most of the -- as clients are getting more intense in moving workloads, they tend to evolve in the sophistication of what they're doing. And we do have -- and this has been going on for a long time. The momentum of -- I try the cloud native product or I try this, that. I see the benefit of the platform, so I'm going to move towards this over time. So that's one of the impetuses of buying Datadog and has been for quite some time.

Great. Great. And I know new logo ARR growth was strong on this most recent quarter. And I think you guys had cited it coming from new customers and existing customers. I'd be curious, though, is Datadog in any way, prioritizing the growth from existing customers? And the reason I'll let you know why I'm going with this or where I'm going with this, but the idea that maybe mining existing customers in the current environment is easier than landing that next new customer just given those existing customers have already seen the benefits of the Datadog platform. They recognize the ROI that you guys are bringing to bear. So I want to see, has there been any tactical shifts within Datadog as far as how you're going out and either landing new customers or mining those existing customers?

No, other than increased investments. So we have a number of different sales force, and they tend to have different focus. We have an insight commercial, which is 100% focused on new logos. That's all they do. And so they're SMB. And then we have a CS or a team that focuses on upselling them. and growing the relationship product and otherwise in the SMB and somewhat in mid-market. And we have an enterprise sales team that is segmented, we have majors or large existing customers where their goal might be penetrating -- growing and penetrating additional divisions and we have other territories that are more new logo oriented. We -- every dollar is treated the same. And so when we expect, depending upon what error you're in. If you're in the commercial sales, you're not focused on growing customers as much. It's a new logo. It's a hunter group. If you're in -- it depends what areas of mid-market and enterprise. So -- no, I would say that given what we're seeing, which is that cloud projects are high priority and we'll be continuing. The new logos are the growth engine for the future. They don't really affect the next quarter. And so we're putting significant priority on that I think, if anything, over the years, we have gotten better at the segmentation about committing resources in different ways and focusing on each of them in the right way.

Great. And I know that you had touched on this earlier or I think it was the allusion to it, but the idea that in this environment, customers might be more willing to I guess, standardize their observability to a platform, right? And so I know it's early, but is it fair to think that maybe if customers are already using you guys for infrastructure, they want to standardize that more broadly, and we'll incorporate, let's say, ATM with Datadog as well. Any early indications from the market with respect to the tighter budgeting environment we're in?

That's been a huge growth driver, as you can see in our metrics. If you look at our metrics of 2, 3, 4, 5 -- that's been a huge growth driver in the company for a number of years. So that's always been -- that's been the case, meaning if you just look at our metrics on cross-sell and how important that is. And if you go and you talk to clients, you'll say -- they'll say things like, why would I ever have these point solutions when I can everything in a platform where they do it for me and it's all there. That's been a huge driver, and we expect it to be a continued driver. Whether it becomes more acute in this environment or not, I don't know. It's been so -- it's been -- you look at the graph that we show and it's been there for a long time. So I would say we expect the continuation of that given the evidence we've seen, but it's been going on for quite some time.

And maybe another question with, again, those charts and the statistics around multi-module adoption from your existing customers. Is there a way to think about, I guess, at the margin, like if customers are using 2-plus products from Datadog or 4 plus or 6 plus, that product adoption, it sounds like at the very least, it continues almost unabated as far as that adoption of your platform. But is there in any way an accelerant versus where we were, call it, a year ago just because the platform has widened and not only that, but customers are looking to consolidate the number of vendors they're working with.

I think they're looking -- that could be -- I think they're looking to consolidate, and have been, the number of vendors. I think you see that in the amount of new sales that we're doing every quarter versus alternative companies. It's the platform drive. It's the fact that we're a product-led company where we essentially have one product, the platform and the client can go in and use it. and it's all set up for them. All of that rationale means that, that has been going on. I think it's too early to tell whether that's -- like again, we don't sell rip and replace. We basically present, we basically have a constantly evolving platform that our clients go into and they adopt this in a pretty frictionless way. So we're certainly seeing that, whether that accelerates or not. It's too early to tell, but we've seen very strong trends in that for quite some time.

Great. Great. And right alongside that is the gross retention statistics. I know over the last couple of quarters, Datadog, if I go back a year, I think you guys were actually calling out mid-90s. And for the last couple of quarters, it's been in this mid- to high 90s range now.

It's the same. It hasn't moved. It's been rock solid where it's been slightly up, but I would say mid-90s, enterprise, high 90s SMB, low to mid-90s. So it's been -- it's very, very similar. We have very little churn. It hasn't moved very much in the amount for all of this through COVID. I mean this -- to any of this, it hasn't moved. So I think that broadly speaking, the mid -- we haven't put out a number, but we said mid-90s. And if you think of high 90s as 98% or 99%, it's not that it's mid-90s, and we just say that enterprise is the high side of that SMB would be the lower, but qualify that SMB still is in 90s.

Yes. And I know, again, that consistency and just to echo that, but like when I think about you guys and this broadening of the platform, I would think that if anything, since COVID -- and again, if we're trying to draw parallels and put on our hat to think about what's different this time I would think that given the expansion of your platform, probably benefits you guys as far as becoming more of a strategic partner to your existing customers today.

Yes, Definitely, I think that might manifest itself -- it could manifest in net retention, too, meaning that you -- if you want to optimize a product, you -- we do this. This is what we do with our clients all the time. We expose them and we say, "Oh, to use your credits, to use what you've done, you can do it in so many ways." So I think it could manifest itself not interesting growth, but it could manifest itself in net retention as clients have more opportunity to use various aspects of the platform. We think over time -- I mean, what's happened over time is when you think about what the metrics we've given on the newer products. And the fact that half the company or whatever is products that we didn't have 5 years ago, I think that's a testament to, as you said, the R&D product and value of the platform.

Great. And cycling back to the verticals. I know -- and I appreciate the color, too, that consumer discretionary is about a low teens percentage of your ARR as far as exposure, right? Are there other verticals that we should think about for Datadog as far as exposure where we should be mindful of maybe they're seeing similar trends or headwinds like consumer discretion or other pockets where maybe things are actually better and just kind of whole home humming along.

Yes. Yes, I think that you asked the question, is consumer discretionary, your largest -- no, no, I'm consumer discretionary. It might be like, I don't know, 10, 11, 12. It's one of our -- so one thing to think about is we are a mirror for the digital economy. And when I mean the digital economy, I don't mean the cloud natives. I mean the digital economy, banks giving portals for their customers. automobiles putting intelligence in it, plumbing supply companies distributing their wares online. Traditional grocery stores having online ordering, so we reflect that, the digital economy. So yes, in places like, I don't know, electrical utility, we may -- that may not be as big as the overall market, but we're highly diversified. So what we were doing, and I think is we always -- our strategy has been to point out the areas where maybe pro rata, there may be more of a trend going on. And of course, it makes sense that in the industry's most affected, pro rata, you would have more of a trend. But we're really diversified. So one of our large -- some of our largest sectors are financial services, media. We talked about travel. It's not one -- at time COVID, travel is doing really well. SaaS software, enterprise at software. So -- yes, I mean the economy could affect everybody, right? But what we're pointing out is the areas and it makes sense, where there might be more intense effect. But consumer discretionary is not one of our largest sectors. It's one of our many, many sectors that reflects the digital economy.

Okay. And just to put a sharper knife to it, but it sounds like, again, it's not like Datadog went out of its way to attract consumer discretionary. It's one of the many verticals that you guys are playing in. But again, you guys have this broad cut versus this digital economy that you're talking to. And so...

One of the great things about our product is we're so horizontal. So that's why if you listen to our earnings calls, shipping companies, insurance, our market is any company, any company. So it could be the most tradition-bound insurance companies headquartered in Hartford or whatever they are and the online, all of them because to compete you generally increasingly have to have a digital business. They're all our customers or potential customers. So that's why we're so distributed. We've said that in the areas where it's cloud native and they've sort of ramped, they're fully deployed and they're in an affected sector. That's where we have the most intense rationalization. We may well have some rationalization across the whole customer base. And I wouldn't -- but we're just pointing out where the weighted average because at COVID, we had everybody doing this optimization at once. So we went to the low end of the organic growth. And right now, what we said is we're not at the high end, we're in the middle, smack in the middle. And so what that means is that we have a portfolio of customers that are doing different things. Not everyone is doing the same thing. So we're trying to illustrate where it's more intense where customers are doing that rationalization.

Thank you for spilling that out. And I know again, I think we started off this fireside by talking about how Datadog does R&D is like an area of strategic differentiation. And so if I think about your R&D road map, how -- like how does the company maintain its investments in, let's say, new product launches -- I'm sorry, scale into new product launches while maintaining its investment in existing products. Like how do you guys prioritize that under the hood within the confines of data do. I'd just be curious to hear again how you guys are looking at that R&D road map longer term?

Yes. So we essentially plus or minus 50-50 in the existing platform and expanding the products and 50% new, roughly speaking. We have a very large platform investment. There's so many things in the platform, and you saw that in DASH when you think about more automation, more intelligence, workflow processes something called power pack, which is the automation of creating the dashboards, online collaboration, ITSM. There's so many things in the platform, which are will create greater value because it's all based on what our customers are doing. And then there's the new products and the new functionality that we've talked about. And so we -- every -- we have a longer-term sort of plan. And then every month, we meet in OKRs. And we essentially review the plan and make sure that we're hiring to that. And so we refresh that every quarter, as we see success, often we're launching a product, and we're doing it -- and because we have this frictionless way and our customers are seeing the platform, we get a lot of feedback. And often, we put something out there, clients use it. We learn. And there's -- and I think this is one of the great things about Datadog in that we have this automatic feedback loop and allows us to refine the product, refine the pricing, et cetera. So we're constantly doing that. I think we said all along, we think there's a very, very large opportunity and continued cloud workloads in the penetration of more of the full platform within our existing customers. Most of our customers do not use both logs and APM today. It's a huge penetration. We have the -- expanding of the functionality, we have the enriching of the platform. We have the security opportunity. We have the shift left. All of these are growth drivers that we think are exciting opportunities for us. all in the core of the very long opportunity of putting more mission-critical applications in the cloud.

And I want to marry that R&D investment that we're talking to with the sales and marketing, right? So I know, again, you guys have this frictionless go-to market where the customers are using the product, getting up to speed in a bottoms-up approach, right? But how do you ensure as the number of Datadog or the modules like percolate and bubble up that your sales and marketing organization is being brought along for that track and then effectively communicating that ROI to maybe their customers' constituents, who's sitting around at that decision-making table?

Yes, I think that's a great question. First of all, one thing is because our greatest salespeople are customers, and we basically have set this up so they can use it pretty instantly. Our customers are our greatest sales people. So they see how things -- so that's really good. That really helps. And then we have invested in sales engineers of presales and some increasing specialization in sales and engineering so that we have people trained up. We have a very large investment in technical account management and in customer success. We've increased our investment in more dedication. Meaning, you can have -- if you have a very large instance or usage, you can have some dedicated resources that can help you use it. And then in product management and product marketing, we're investing in telling that story takes security. That means going to security conferences. That means getting your word out. So all of this is what we're doing. So, so far, we have not needed, and I know this is a related question to create a specialized sales force. It's worked really well for us. Why? Because in the end of the day for our clients, as we talked about, we have 1 product, the platform. So as long as we have the same buying persona for the most part, using the platform every day, this has been working, and we don't have any plans to change it. We said there may be some places where we reach a different persona that's not in the platform every day like it could be parts of security, and then we may have to think about overlay or expertise and all that, which we're experimenting with, but don't -- aren't far enough along to be able to announce that we're doing this or that.

Great. Great. And I think we probably only have time for one more item here before we go. But I just wanted to talk about the long-term model here for you guys, right? So you're already delivering very strong operating margins and free cash flow generation. As we think revenue and ARR continues to scale based on these growth drivers we've been talking about, are there specific categories where Datadog is looking to drive more efficiency? How do you frame out that model over the next couple of years?

Well, I think we've been -- when we're not public, we gave a long-term target of 20% to 25% operating margins. And a lot of times, companies don't evidence the ability to get there. We evidenced the ability maybe quicker than how long-term target. And this gives us tons of degrees of freedom. What this shows us is that there's lots of levers to pull. And we've proven, as you've said, that we are profitable, there is scalability. There are lots of things if you look at how we handle gross margin. And so there's a lot of this is in our decision-making of given the economies and scalability and the model of investment decisions, and we've been a really good steward of capital. We've really not been or gone to what you see out there in the market. We stay right up in the middle and consistent. And I think this lets us make the decisions between how intensely we want to invest for the future versus do it more slowly. And we think there's a very long-term opportunity, and we think we've run the trust, and we've been able to show that this is a very scalable cash producing model that we can pull levers of investing more intensely at this time and pulling back to better control that. And that's what we watch every day in managing there.

That's great. And, of course, we're going to have to leave it there, but...

Good questions. Thank you.

Really appreciate today.

Comprehensive.

Yes. Thank you very much.

Thank you, everybody. Have a good day and a good conference.

Thank you.