Elastic N.V. (ESTC) Earnings Call Transcript & Summary

Welcome to our first session. I'm really happy to have the team from Elastic here. Ash, Eric, thanks for joining us.

Maybe let's start like to get everyone grounded, like Ash, you reported some really good numbers last week. From -- a couple of weeks ago, yes. What was the highlight from your perspective? Like just -- what you saw?

So Q2 for us, probably the most interesting thing about Q2 was very strong commitments from customers. The way our business is set up is fundamentally our sales-driven motion is the primary motion for our go-to-market. And customers make commitments, whether it's on cloud or self-managed, doesn't really matter from our perspective. And those commitments then turn into revenue over the next 12 months as consumption happens. And we saw really strong commitments across the board. Some of the largest $1 million-plus deals that we've ever done, 2 deals that were over $20 million in total contract value, 5 total that were over $10 million and 30-plus deals that were $1 million deals. So really good momentum in the sales organization. In terms of the business areas that we continue to see strength in, AI is our fastest-growing part of our business. We saw that continue to play out well. And then 2 of the largest deals were security deals. So we are now doing very meaningful displacements of incumbents. Both the $20-plus million deals were security. The largest was the one that we publicly talked about. This was at CISA, which is the government agency in the U.S. responsible for infrastructure and cybersecurity for all civilian agencies. And then the other was a large chemicals manufacturer that basically picked us for XDR. So that includes not just SIEM, but also endpoint protection. And we won that deal against just about every incumbent, every endpoint security and XDR player out there. So we are seeing good momentum on areas -- on those 2 areas. Observability continue to do well. 2 of the top 10 deals -- sorry, $10-plus million deals were observability deals. So broad strength in the business. Consumption was strong. The only one thing that affected us a little bit was in the public sector in the U.S. because we had the shutdown that affected the last month of Q2 for us. And so some renewals where the customers are continuing to use our product basically did not close in Q2. And there were situations where they are still using it. So the renewals are just going to happen in Q3, but there was literally nobody there to process the orders. And so that just shifts some revenue from Q2 to Q3. And we talked about it in the quarter, roughly, it would have added maybe 1 percentage point of revenue growth, but that was the magnitude. It doesn't affect the full year. It just shifts it from Q2 to Q3. But overall, momentum was strong. That gave us the confidence to raise the guide for the full year. And we feel really good about how we are tracking to the overall model that we laid out at Financial Analyst Day.

Yes, yes. And then the -- how does -- it does look like it's for you guys kind of coming together really nicely. How does it feel in terms of end demand in general though? Because like we had at the beginning of the year, the tariffs and all the uncertainties, et cetera. When you talk to customers, where are they in the head in terms of thinking -- in terms of spending, sorry, yes.

No, no, no, absolutely. So across the board, we are seeing continued strength in demand. And even with the tariffs, like we had the same kind of questions of how is this going to affect buying behavior? And what we saw was it hasn't really changed buying behavior. If anything, some of the geopolitical happenings has resulted in public sector purchasing in Europe pick up. And that is interesting because we do a lot of work in public sector, not just here in the U.S., but globally. We get used for both search, observability and security, all 3 use cases. And given our roots as a Dutch company, we are seeing good success in Europe as well because of that reason. So overall, demand remains very strong. Even in U.S. public sector, in spite of the shutdown, what we saw was just movement of stuff because of renewal timing and so on. But other than that, like the demand just continued to be very good.

On that, that was my next question on U.S. federal. Obviously, at the beginning of the year, you had [ DoD, ] but like it's all about efficiencies, doing stuff better, getting -- doing it more cost effectively as well. And you guys have a good solution, but you've always had like a good pricing model. Like does that, in theory, mean like there's more discussions going on with them?

Yes. So there were a few impacts of that [ DoD ] happening that happened. That was more early part of the fiscal year, maybe even towards the end of the last fiscal year. What -- initially, there were 2 impacts to it. Like first was there were some agencies that were affected, right? There were some agencies whose entire budgets were decimated and so on. So that did have some effect on us, and it was relatively small. It's all public information. You can go and look at what contracts were canceled and so on. So you can see that for us, the impact wasn't very, very large, but we saw some of that. There was obviously a lot of uncertainty in those days on nobody knew what they were allowed to do effectively. But as that settled down and as the administration has fully settled in, there is a true desire to move faster in terms of modernizing, in terms of really moving to platforms that can be more efficient. Like you rightly said, Raimo, that has been one of our greatest strengths, both the capabilities that we offer, but the pricing model that we have that is very customer-friendly. And that has really opened a lot of doors. I mean the CISA deal is a perfect example of that. So with CISA, they've been a customer of ours for a long time. But this is a situation where they basically are now offering Elastic SIEM as a service to other agencies within the U.S. federal government. And so that's something that they've never done before. So rather than -- because they are like the group that defines what cybersecurity means for all civilian agencies, they don't do anything with the Department of Defense. But with the civilian agencies, like they are the ones who define the standards and everything. But now they are offering this as a service. This is pretty unique. We are very excited about this. So we'll be working with them over the next many quarters and years to make sure that we keep building this up because we feel that this is a beachhead that's going to allow us to keep growing our business in a very, very healthy way for a long time to come.

Yes. Okay. Perfect. And then I wanted to switch gears a little bit. Obviously, AI is a big topic for everyone at the moment. I remember when in the early days of then that team came up, it was all about vector and vector database and who are the vector database and who can do more and stuff like that. How has that discussion evolved?

So there are 2 things that have happened. So first is in terms of the companies that are moving fastest with building AI capabilities, the ones that are moving the fastest are the ones that are ISVs, companies, not just AI native companies, but even software businesses that have been around for a long time that are trying to -- working hard to infuse AI into their product, into their capabilities. They might be building copilots. They might be building their own agents. They might be building AI-based automation into their products. They're the ones who are moving fastest because they understand their software stack. They have more sophisticated engineering talent. And we are doing a lot of work with them. Like we've talked about the work that we've done with DocuSign and IBM and so on, but there are many, many more, right? So that is a very fertile ground for us. And we grow as their usage grows, which is a great thing for us. And then within enterprises, what we are seeing is the shift has happened from people basically just using semantic search or starting with semantic search, vector search to now trying to build more complete applications. And those can be chatbots, those can be applications that take certain simple tasks or actions. In security, we are seeing people build security automation, like what people would think of a SOAR in the past. That used to be quite static. Now people are building agents that can automate certain actions. So it just relieves the burden on their teams. So it's all about efficiency in those use cases. And for that reason, when you're building those kinds of use cases, what -- what is becoming more and more important is a complete platform to let them do that entirety of work. That is everything from bringing in that data, turning it into embeddings, using various search techniques, including vector search, but then having an interface that allows them to operate on that data, take actions on that data. And that was the genesis of Agent Builder. So Agent Builder came about that this was the capability that we announced at Financial Analyst Day. It's now out in the field. It's in technical preview. We expect it to go GA soon. But fundamentally, the idea here is you start to build agents directly on top of your data because data, the context is what gives LLMs meaning in what they need to do. And that's the secret sauce for most businesses. So that's what the trend that we are seeing now.

And remember like a year ago, there was like POCs left, right and center. Someone talks about POC graveyard as like the new term. Like where are we in terms of taking these kind of POCs and then just getting them in production? Like -- and you gave a couple of examples already, but do you see that as a trend that we just need to be aware of the time line?

I think everybody is building something. Most people have had at least 1 or 2 applications in production at this point. So we are definitely seeing production -- live production scenarios across many, many customers. And when you have over 2,000 customers like we've talked about in cloud only that are using us for these kinds of use cases, that tends to -- you can imagine that there's going to be a large number of that base that is in production, especially given that we have hundreds of them that are in our $100,000 cohort that are spending a fair bit of money with us. So these are all production use cases. I'd say, Raimo, the biggest thing that most people should appreciate is we are still early in the number of applications that most organizations have automated. So that will also grow, right? So you think about like the way I would maybe interpret the question is, think in terms of traditional applications that were being used for any kind of automation, whether it was Salesforce automation or financial ERP automation, like there are hundreds of different modules that we use in any given business. With AI, you are still at the 1 to 2, 1 to 10 stage. We haven't gotten to the point where you have dozens and dozens of these kinds of applications that have been created. So penetration into an account becomes incredibly important because that's your landing spot. Once you're in there and you are the standard, then you're going to grow with them as they grow. And that's really what we're seeing. So we are seeing contribution from the cohort. And at the Analyst Day, we talked about the fact that the cohort of customers that's using us for AI is growing roughly about 5% faster than the rest of the cohorts.

And then maybe one for Eric. like as you think about more AI adoption, like how do we think about pricing? And I think it's different for you, like for some of the other guys, it's like token usage, but you're more of the provider for data, et cetera, for that. Like -- but how does pricing fit into that AI story?

We don't have a specific AI SKU. And so there's not going to be a price that you pay to utilize a certain amount of AI. The way that it works is as you adopt -- well, first off, as you adopt AI, you tend to move to higher levels of functionality. So you might need to use enterprise versus gold or platinum. And that in and of itself raises prices in terms of what you're consuming. And then further, as you adopt AI, you're going to bring more data onto your platform, which means that you're going to be using more data, you're going to be ingesting more data. So the way that we think about AI sort of benefiting our pricing and the dollars that we're going to bring into the ecosystem is there going to be new use cases which are going to be just net new things that we're going to be able to monetize. There's going to be more data running through use cases because you're utilizing all the Gen AI capabilities, and that's going to drive more data consumption and more data volume. And then the third thing is you're going to have to move to a higher tier of functionality in order to utilize some of the AI capabilities.

Yes. Okay. Perfect. And then at the Analyst Day, you talked about the opportunity in the different segments. The one question I get from investors a lot is like you -- in a way, if you think about your product evolution, product development, you have like the platform capabilities, but then you also have like the capabilities in the observability, security. I'm missing one search. How do you prioritize that? That's kind of the one thing for me all the time, like how do you go about that in terms of like, okay, 20% needs to go here and 20% somewhere else? Or like how do you go about that?

Yes. So I think the most important thing to first appreciate is that there is a reason why we are in these 3 precise categories in these 3 areas and not in other business areas. The common theme in all 3 of these is the data tends to be unstructured and messy. And a search platform like ours, like at the heart of our platform, what is different about us is we are a search engine. We're a search platform. And that allows us to work with messy data better than any database would. It doesn't matter if it's a SQL database or a NoSQL database. Fundamentally, those systems aren't designed -- they're designed for schema-based operations. Our system is designed for data that does not have any inherent schema in it. And that's really the big difference. So security, why are we in security? Because the data tends to be logs, the data tends to be telemetry that's coming from all kinds of devices. It doesn't have a well-defined schema. And so to be able to connect the dots between all of those different data types is a really, really hard data problem. And that's the reason why you are seeing a pretty significant turnover that's happening in legacy SIEM providers where people are moving away from them because they're having all kinds of data challenges. The same with observability. Logs tend to be very, very messy. Metrics, less so. Metrics are simpler data types, but logs and observability tend to be very, very messy. And that's the reason why we are in these 3 areas. Now getting to your question of how we think about resource allocation, the simple model is we should have the platform become capable of dealing with the fundamental problems that exist in the segments that we play in. So over 60% of our investment actually goes into the platform. And that is an incredibly high leverage model because just to give you an example, when we are -- when we build something like our vector database and we invest in that vector database and we build it in such a way that it's directly in the platform, that is now used for product components like Agent builder for you to build AI chatbots or for e-commerce search, which is more traditional, the business that Elastic kind of started with or it can be used for building things like detecting significant events directly from your log data for observability. So just from the logs without the human needing to create any alerts or rules, we can infer potential issues in the data. It is the same vector database that is being used under the covers for doing things like attack discovery, where we can look at all the alerts that you're getting, enrich that data in your SIEM and correlate it to basically figure out what are the connected patterns within those alerts that tell you, hey, this is an APT32 style attack that's happening. And all of it is being done on that same component. So that is massively powerful for us. So the actual investment that we have to put into either security or observability or search ends up being a fraction of what any other organization would because the core is that much robust. So it's a fat platform with small solution layers on top. And that has worked incredibly well for us. So we think of them as power plays, like in sports analogies like if I do one thing, it's going to help me in 3 different ways. Like that's the way we think about it. And there's like searchable snapshots was another example of that. When we built that feature, it was massively useful for observability. It was massively useful for security. It didn't help us that much in search, but that's okay. You have like one feature that we are building into the platform that's helping us in so many meaningful ways. And that's how we think about resource allocation.

And then related to that is then like how do you think about the growth you get from that? And like -- and yes, maybe we go by the 3 groups like observability, security, search.

So I have a pretty ruthless model on this where the model is like we have GMs for each of the areas. Their job is to drive the business case and make the -- it's like -- it's a resource allocation equation. And I don't really care which one grows faster. Like my favorite child is the one that's growing fastest. And we do think forward and we look at like forward-leaning investments that we make strategically. But each group has the mindset that you need to succeed on your own. And it's not that we are going to keep funding you just because we believe that this business should grow at x because what matters most is really driving the total growth of the company towards and beyond the midterm model that we laid out at Financial Analyst Day. We unveiled that at Financial Analyst Day, but internally, we've had that kind of thinking within the company. And then the different groups, if they are able to run faster, if they're able to grow faster, they get more investment. If they are, for whatever reason, not able to because they need to build some additional features to become more competitive, then we might say, okay, well, let's get that right first before we add more fuel to the fire. But the whole, the goal is grow that like we've talked about. We've talked about Rule of 40. We've talked about getting past 20% growth as we look at sales net subscription revenue. And all of that is like the way we run the business.

And the -- as part of that on the Analyst Day, there was like -- I think it was like, if I remember correct, like 15%, close to 5% from AI. Is that just going back to that 5% point that you mentioned earlier that people that are using you are kind of just using more? Or is there other factors that we should consider?

No. So, a, we are getting new customers all the time. So one of the changes that we made in the sales organization was Q1 of last -- of FY '25, when we reorganized our teams and the segmentation, we made the segmentation changes. One pretty critical element of that was to create very focused hunting territories that were hunters that would go after greenfield accounts, accounts that we had never done any business in. And the goal for that was to go after new logo generation in a meaningful way. And this was not just SMB, which we do through our self-service cloud, but really like go after the mid-market, go after enterprises because even today, we have 40% of the Fortune 100 isn't a customer. Why? I want to get that to be a customer, right? So that's the mindset. And you have to drive that through the sales organization. So we are seeing new customers come on. AI's growth will come from that as well. But that equation that we had talked about was purely for -- from a cohort perspective, right? So if you take the existing cohorts and the rate at which they were growing, what the analysis that we've done looking back for a period of time is the cohort that is using us for AI is growing at least 5% faster than everybody else. And so it was one way to think about the over 20% model, and there are multiple paths that we are working on. So it's not just about expansion of current customers. It's about the new logos that we are signing up. It's about expansion across multiple use cases. So all of that plays in. Our model is basically a land-and-expand model.

And then you mentioned the changes on go-to-market already as well. Like one of the other thing was just to be if I remember correctly from last year, like deeper on account coverage and how many accounts per sales rep. Like last year, that caused some disruption, like now that everything is settled down, like what do you see there in terms of results?

So I think the data kind of speaks for itself. We've done more million-dollar deals, larger deals like -- and you can see that in the numbers, the $100,000 cohorts and the growth in that. And just the sizes of deals gives us a lot of confidence that we are getting deeper into accounts. We are doing more meaningful larger accounts. Like the CISA kind of deal would not have happened if we hadn't made some of those changes, right? Because to get into that situation where we get a customer that's been using us for many years, but has been using us like they were a 7-figure customer, but had never gotten to the scale that they are now. For them to not only use us in bigger ways, but champion us and build a service with us that they can take and sell to others, that contract is just going to keep growing, right? And it's a 1-plus option year contract. So the potential in that is much more than the number that we've talked about. So it's a wonderful opportunity. I think the model has settled in. We are seeing the right kind of behavior from the teams. And all of that, honestly, Raimo, fed into the confidence because our pipeline is very strong. So when we raised for the full year, it was not just based on the business that we've already closed and the contracts that we've already signed, but even what we see in the pipeline, and the signs are really good.

Yes. Let me just add one more thing to that. I think that Mark talked about this a little bit at Analyst Day, but we also saw a meaningful uptick in productivity in terms of what our sales force was able to generate on a per person basis. And we saw a really nice improvement in sales efficiency. And that's why as we look at our business today, we view the go-to-market motion 5 quarters after the issues that happened in Q1 of '25 being very investable. And that's why we've been adding to the headcount because we've seen success, we've seen repeatability. We've seen productivity and efficiency in that go-to-market. And so now we're deploying more sales capacity.

Yes. So that should be exciting because you have now have hunters, which you didn't -- so that gives you new accounts and you have like the more coverage there on the sales. On that note, how do you think about -- how do you think about like investments then going forward? Is that -- like is it -- are you looking at the signals on productivity and that just drives where you want to go? Or do you -- is it more driven on the growth that you want to achieve the old sales force model I just throw bodies at it? Like how do you think about that kind of dynamic there?

Yes. I think it's a mix of both. I think that we want to see sustained productivity, and we want to see that as we add to the sales force, as we add capacity into the model that people are able to continue to be productive and continue to be efficient. If we see a big step backwards, that would obviously be a signal to us of something negative. But we've been seeing quite the opposite. We've been seeing a ton of strength in productivity. The business model is actually getting more efficient. And so that's why we're investing. Hopefully, as we continue to see the strength that we have in the product, where across all 3 of the solutions, we're seeing Gen AI benefit us. It's expanding our TAM in search, obviously, but it's also making us more competitive in security and in observability. As we see that continue where the product is really very sellable and as the go-to-market motion as Mark is really driving this replicable and efficient model, we're going to continue to deploy capacity.

And how long does it take -- like in theory, I would assume there's also some even more productivity gains because if I remember correctly, you kind of reduced the number of accounts per senior sales rep, but those that they lost went to someone else, but that person needs to ramp up, build pipeline, et cetera. So in theory, we should have like even the secondary effect coming at some point -- while it should be pretty much now coming...

I think the best way to think about it is, let's say that there were -- each rep had -- and I'm just making up numbers, but order of magnitude, not totally off. Let's say a rep had 25 accounts, enterprise accounts, maybe we had business in 5 or 6 of them and the others were greenfield where we had never done business. We basically kept the 5 to 6 that they had -- that we had in ARR and maybe give them 1 more or 1 or 2 more. So brought their number down to between 5 and 8. And then everything else that was greenfield, we moved to these 100 territories. The thing with 100 territories is it takes time to build that pipeline. Now the difference is now there is a person who only makes money if they close business in that new account. That is a different mindset than the prior model. In the prior model, what ends up happening is human beings tend to follow the path of least resistance. So if I have existing business in 5 to 6 of these accounts, I'm just going to try and grow that. So expansion continues to be good, but you don't get new logos. And so part of this was to explicitly -- because, look, we can't become a multibillion-dollar company if we don't make some of these fundamental changes. So that was part of the reason why we did it. But you're exactly right that even now, we are just starting to see the benefit of the hunting motion kick in. So there is more of that benefit to come. And we -- so it's not about just throwing bodies, like one thing that I am very focused on is growth is most important because in the long term, like it really shapes the business. But it's equally important to be continually improving your profitability every year. And so part of the focus is to make sure that we really get that productivity model right because that's the biggest lever. We can make every rep more productive. That means investing in better training. It means investing in better tooling. That's all part of what Mark's been driving for the last 6, 7 quarters. And it's -- we're really seeing the benefits.

And then I realize like we could have gone on for a longer. We have like a minute left, Eric, for you. On the margin side, the -- we've seen the improvements as an organization, like on the other hand, you're kind of investing into growth and want to do more stuff. Like how do you do that? It sounds almost too good to be true.

Well, the goal is to have a measured approach and to look at both profitability and growth and to weigh them against each other. At the Analyst Day, Navam talked a lot about our journey along the Rule of 40. So he emphasized that as well as this motion to get to the 20% plus sales-led subscription growth. And so every year, we look at our business, we see where we can invest and where we can drive really positive returns from that investment. We weigh that against maintaining both the growth and sustainable growth as well as continuing to increase our margin. And we've been working towards that. And if you look at the last multiple years now, you've continually seen us have a sustained increase in our margin profile while maintaining the strong growth that we've had. And that's sort of the path that we've been on and that we'll continue to be on as we get to that Rule of 40 and beyond in conjunction with this march towards the 20% plus sales-led subscription.

Perfect. That's a great summary as well. Ash, Eric, really enjoyed our conversation. Thank you...

Thank you.

Always a pleasure.

Thank you