Elastic N.V. (ESTC) Earnings Call Transcript & Summary
February 29, 2024
Earnings Call Speaker Segments
Anthony Luscri
executiveThanks. Hello everyone, and thanks for joining us on this call to discuss Elastic's Q3 FY2024 earnings. With us today are Ash Kulkarni, our CEO, and Janesh Moorjani, our CFO and COO. After their opening comments, we'll take your questions. Our press release went out after the close today and can be found on our site. Supplemental slides are also posted on the Elastic Investor Relations page at ir.elastic.co. On this call, we may offer forward-looking statements covering things like predictions, estimates, or expectations regarding demand for our products and solutions, as well as our future revenue and other information; these statements are based on factors we know right now, reflect our views only as of today, and come with risk and uncertainties that might make actual results differ materially. We take no responsibility to update or revise any forward-looking statements unless the law requires it. For more on the risks and uncertainties, please see the press release we put out earlier today, the slides available on the Investor Relations website, and the fuller descriptions in our filings with the Securities and Exchange Commission. We're going to cover some non-GAAP financial measures as well. You can find disclosures about those non-GAAP measures, including reconciliations to the closest comparable GAAP measures, in both the press release and the slides. A replay of this webcast will be available on our company website through the Investor Relations link. Our quiet period for the fourth quarter of fiscal 2024 starts when business closes on Tuesday, April 16, 2024. We'll be participating in the Morgan Stanley Technology Meeting and Telecom Conference on March 6, along with the Wells Fargo Software Symposium on April 10. At this point, I'll turn it over to Ash.
Ash Kulkarni
executiveThanks, Anthony. Hello, everyone, and I appreciate you being with us on this call today. Elastic has posted yet another solid quarter. What we achieved shows we keep executing well, can bring differentiated product innovations to market, stay fiercely committed to our customers, and run the business carefully. During Q3, the patterns we noticed across the first half of this fiscal year went on fueling momentum in our operations. To start, demand from customers for generative AI is still robust, and as they gain a better understanding of what building generative AI applications involves, they grow more capable of seeing what our platform offers compared with vendors focused solely on vector databases. Next, we observed ongoing momentum in platform consolidation, since customers picked the Elasticsearch platform to replace incumbent solutions across several use cases, especially log analytics and SIEM. Finally, cloud consumption patterns stayed stable as customers consumed more than their earlier commitments. Such patterns strengthen how confident we are about the business and about future growth, given that a growing number of companies select our search analytics platform as a central element of their IT infrastructure stack for log analytics, SIEM, and Gen AI applications. In a few minutes, I'll discuss these patterns further and offer some customer success stories that are fueling our growth. Let me begin, though, with a look at our financial performance. Our third quarter results came from strong momentum and execution, and I'm happy with them. Revenue in Q3 rose 19% year-over-year, while Elastic Cloud increased 29% year-over-year, fueled by ongoing customer traction in the cloud and consolidation onto the Elasticsearch Platform. We also beat our profitability guidance, posting a non-GAAP operating margin of 13%. As for the broader patterns from this quarter, in meetings with customers worldwide, I noticed a powerful wish to use AI for better business processes and stronger customer experiences. Meanwhile, companies keep concentrating on ways to cut costs; in a way, the goal is achieving efficiencies without giving up innovation. That's exactly where Elastic excels. Through the Elasticsearch platform, businesses can tap the full potential of all their data to obtain insights that fuel innovation, improve customer experiences, and guide strategic choices. When it comes to generative AI, customers are beginning to push forward their abilities in using retrieval augmented generation, or rag patterns, to create GenAI applications with existing and emerging language models. That pushes them to thoroughly investigate the whole set of functionalities required for building enterprise class GenAI applications. Sophisticated features such as hybrid search, personalization, and re-ranking with reciprocal rank fusion, document level permissions, security, and a great deal more. All these capabilities, alongside our top-tier vector search functionality, plus the reality that we already serve as the incumbent data platform for tens of thousands of customers, mean we can greatly reduce the barrier for businesses seeking to create their own generative AI applications. That represents a distinct edge nobody else has, and with Generative AI plus RAG now more widely grasped, that edge stands out even more clearly. During Q3, we brought on several hundred additional customers adopting the Elasticsearch Relevance Engine, or ESRE, once again. With ESRE, building Generative AI applications happens fast, and there's no need for complex, costly model training — which explains a great deal about why people come to us for Generative AI. Take Consensus, a customer of ours for a long time: it's an AI-driven search engine pulling together and condensing findings from over 200 million peer-reviewed papers housed in the semantic scholar database. For advanced artificial intelligence along with semantic and text search, they rely on Elasticcloud. ELSER provides the engine behind Consensus 2.0. Better accuracy and relevance in search results have come to their users, along with fresh GenAI capabilities, including summarization of the top 10 studies, made possible through integration with OpenAI's GPT-4 model. During Q3, we grew our relationship with Stack Overflow as well — the biggest and most trusted developer community online — which swapped out its prior vector database in favor of Elastic. Scalability, differentiated functionality, and ecosystem-wide integration drove the company's pick of Elasticcloud. To give developers a question-and-answer experience that feels more human and is driven by generative AI through its overflow AI product, Stack Overflow draws on Elastic's vector and semantic search capabilities. Customers keep picking our platform also because they anticipate relying on our ESRE functionality down the road. Put differently, they view us as the correct pick for what they need now and what they'll need later. For instance, a legal-sector cloud document management firm signed a multi-year agreement with Elastic during this quarter. Possessing more than 6 billion documents that can be searched, this company picked Elastic owing to our innovations, our performance, and our capacity to drive search throughout their whole platform. Having previously relied on a competitor's offering, they switched to Elastic because of our market leadership and to bring the search experience up to date, matching their customers' expectations by drawing on Elastic's vector search capability to improve the user experience. Past generative AI, we observed that customers keep displacing incumbents and consolidating onto the Elastic Platform for observability and security — the second trend I mentioned before. Built atop our Search Analytics Platform, with Elasticsearch serving as the foundation, sit the Elastic observability and security solutions. Quickly ingesting every kind of data, indexing it at extreme speed and petabyte scale, making it searchable, and then enabling every sort of advanced correlation, AI-powered search, and ad hoc analytics — our platform does this with unmatched capability; this lets customers centralize onto one platform for every use case demanding this sort of advanced search and analytics. That explains a lot about why we succeed in displacing incumbent solutions, particularly in log analytics and SIEM, or security analytics. Our latest innovations, among them our AI assistance for observability and security, make the move to our platform even more compelling for customers. During this period, we finalized a number of agreements worth multiple millions in which we took the place of established providers in observability and security. The worth of what we offer, along with our creativity, is connecting with organizations in many sectors, since we keep assisting clients in reducing what they spend on IT overall, while also enabling them to extract still more benefit from the new things we develop. To illustrate, a major bank headquartered in the United States picked Elastic as its preferred SIM platform, opting for us over its previous supplier. Factors that led to this decision were the speed of the platform, how it performs, its ability to scale, and the extensive security functionality Elastic provides. Another reason they selected Elastic was our novel searchable snapshots feature for storage tiering, which lets the bank keep critical security logs for extended regulatory retention windows while cutting storage expenses significantly. Additionally, we grew our relationship with the U.S. state agency for Elastic Security. That agency had first come to Elastic for search, and then during Q3 picked Elastic to take over from its old security provider for SIEM, which let it boost assistance for its security operations center. The breadth of Elastic's ingestion abilities, the pace of its search, how well the platform works, its reliability, and what it can do for reporting let the agency obtain and keep visibility throughout its environment and across numerous agency branches, and connect alerts from a wide array of security tools. Now, shifting to products and innovation, we keep putting resources into features that let customers move without difficulty from established offerings over to Elastic.
Ash Kulkarni
executiveDuring our previous call, I brought up the introduction of our robust new piped query language, Elasticsearch Query Language, or ESQL. While we keep working to improve this fresh functionality, the response from our clientele has been enormous—roughly 1,000 customers have been testing it since the November debut. Our January release, Elastic 8.12, featured multiple important upgrades, among them scalar quantization, a vector search optimization method that boosts performance on a number of vital metrics; it greatly enhances scalability for enterprises, enabling clients to hold 4 times the vectors in RAM, while simultaneously speeding up indexing and cutting query latency by as much as 2 times. Further improvements to search concurrency produced faster results across an extremely wide range of query categories, such as vector similarity and full text search, plus a 40% drop in latency for analytical queries. Within Elastic Observability, we moved a number of features to general availability, among them the improved Elastic AI assistant for observability, SLO monitoring, and mobile APM support built on open telemetry; in Elastic Security, we keep improving our industry-leading Elastic AI Assistant for security, rolling out real-time alerts with natural language interactions that transform how security analysts handle alert triaging; this yields a security operation that is more efficient and effective, skilled at steering through the intricate cybersecurity landscape. In cloud security, we introduced support for Microsoft Azure, alongside our prior backing for AWS and Google Cloud regarding cloud security posture management. Regarding go-to-market efforts, we maintained momentum in Q3 by taking part in AWS re:Invent and our ElasticON conference events, which let us engage with thousands of customers, prospects, and users globally; we have finished six such ElasticON events so far and plan to host six more during our fiscal Q4. As a last yet significant quarter highlight, I was delighted to share that Mark Dodds came aboard Elastic as our fresh Chief Revenue Officer. Mark oversees all customer-facing duties at Elastic, covering global sales, customer success, solutions architecture, professional services, ecosystem and partnerships, and sales operations. Mark delivers deep go-to-market expertise to our leadership group as we keep building momentum in Generative AI and expanding throughout every segment of our business. I'm eager about all that Mark will accomplish for us in the years ahead, as we strive to seize the complete opportunity before us as an organization. To wrap up, I'm quite happy with our performance this quarter. The breakthroughs we're bringing to our Search Analytics Platform, the momentum we're building in Generative AI, and the traction we're witnessing as customers consolidate onto our platform continue to boost our confidence in what lies ahead. We envision a future in which our Search Analytics Platform turns into an essential component of every IT infrastructure stack for extracting insights from all data, particularly amid generative AI. My thanks go out to the Elastic team for staying so resolutely focused, and to the whole Elastic community for their commitment that never wavers. So with that said, I'll now hand things over to Janesh.
Janesh Moorjani
executiveThanks, Ash. We're happy to report a solid quarter, thanks to how consistently we've stayed focused on execution. Once again, we finished above the top end of our guidance for the quarter, on both revenue and earnings. Q3 mostly unfolded the way we anticipated. Total revenue for the third quarter grew 19% year-over-year, and Elastic Cloud grew 29% year-over-year. We kept our attention on profitability, posting a non-GAAP operating margin of 13%. What we achieved this quarter shows both our marketplace momentum and our ongoing discipline around investment. Strong execution allowed us to generate solid growth in self-managed revenue as well as Elastic Cloud revenue. Just like in earlier quarters, several customers consolidated onto the Elastic Platform to reduce overall spending while still pursuing innovation. We also observed solid growth in cloud consumption, as customers kept drawing down against their contractual commitments. As Ash noted, customer interest remains strong regarding Generative AI use cases. Customers tell us they're eager to use ESRE for building Gen.AI applications, because it offers a comprehensive, enterprise ready platform for grounding large language models in their private business context. Although it will take time before Gen.AI spending turns into a major revenue driver for us, our expectation remains that Gen.AI will represent a meaningful revenue opportunity over the longer term. Now let's turn to Q3 results and our outlook. Third quarter total revenue came to $328 million, rising 19% year-over-year, or 18% year-over-year in constant currency terms. Subscription revenue for the third quarter was $308 million, an increase of 20% year-over-year, or 19% year-over-year in constant currency, making up 94% of total revenue. Inside subscriptions, Elastic Cloud revenue reached $143 million, up 29% year-over-year on both an as reported and constant currency basis, which mirrors the trends I described earlier. Elastic Cloud accounted for 44% of total revenue this quarter, compared with 40% a year ago. Revenue from Elastic Cloud tied to month-to-month arrangements made up 14% of total revenue. Professional services revenue for the third quarter was $20 million, up 7% year-over-year as reported and 6% year-over-year in constant currency. As we've noted previously, professional services revenue can swing between quarters depending on when services are delivered, and we don't anticipate its mix shifting meaningfully over time. For additional context on overall deal flow, EMEA posted the fastest growth this quarter, with the Americas and APJ behind it. We still see a healthy balance throughout the business across geography, solutions, and verticals, and that diversification speaks to the breadth and popularity of our platform. Turning to customer metrics: we closed the quarter with more than 1,270 customers whose annual contract values exceed $100,000. If we zoom out on how many customers we're bringing in, we closed out the quarter with more than 4,290 clients paying above $10,000 in ACV, plus roughly 20,800 subscription customers in total; the additions we saw this quarter in the greater than 100K bracket make us happy, since they show how much we prioritize the highest value customers, and those customers are what really fuels the business. Our net expansion rate — a trailing 12-month lagging indicator, as you're aware — came in at about 109%, matching Q2 and landing right where we thought it would for the quarter. Let me shift to profitability, where I'll go over non-GAAP measures. Among other items, our non-GAAP results for the third quarter leave out a discrete income tax benefit of $207 million tied to the release of a valuation allowance for GAAP purposes; none of our operating results, non-GAAP earnings per share, adjusted free cash flow, or cash and cash equivalents were affected by this. Gross margin for the quarter reached 77%, compared with 76.8% in the prior quarter, on a slightly higher subscription mix. Operating margin for the quarter hit 13%, a better result than we had expected. That strong operating margin came from revenue outperformance along with our ongoing emphasis on managing expenses while we invest with discipline to drive future growth. Diluted earnings per share for the third quarter was $0.36. Adjusted free cash flow for the quarter was $63 million, or a 19% adjusted free cash flow margin. For the full fiscal year, our prior outlook is unchanged, and we still anticipate that adjusted free cash flow margin for fiscal ‘24 will come in slightly above the non-GAAP operating margin for fiscal ‘24. Now to guidance. Though we're quite happy with how we've executed and outperformed so far in FY ‘24, our guidance philosophy hasn't changed; we remain prudent in the near-term. Given the many moving parts in the broader macro climate, we expect business conditions to stay largely the same. We also think it's appropriate to expect that consumption patterns could fluctuate in the near-term. With that as context, for the fourth quarter of fiscal ‘24, we expect total revenue between $328 million and $330 million, which represents 18% year-over-year growth at the midpoint on an as-reported and constant currency basis; for the fourth quarter of fiscal ‘24, we expect non-GAAP operating margin between 7.4% and 7.8%, and non-GAAP earnings per share between $0.18 and $0.20, using between 105.5 million and 106.5 million diluted weighted average ordinary shares outstanding. For full fiscal ‘24, we're raising our outlook and now expect total revenue between $1.260 billion and $1.262 billion, which represents 18% year-over-year growth at the midpoint, or 17% on a constant currency basis. For full fiscal ‘24, we expect non-GAAP operating margin of approximately 11%, and non-GAAP earnings per share between $1.15 and $1.18, using between 103 million and 104 million diluted weighted average ordinary shares outstanding. Turning now to fiscal ’25, our planning work is currently underway. It’s still too soon to get into details, particularly where the revenue outlook is concerned, but I can offer our thinking on how we intend to approach the coming year from an investment standpoint; we’ve already shown that inherent operating leverage exists within our business model, and during this year we grew non-GAAP operating margin at a quicker pace than we originally anticipated. Throughout this planning effort, our aim is to strike a balance between putting money into long-term growth and delivering near-term profitability. Because of the sizable opportunity we perceive in Generative AI, our plan is to step up investments in fiscal ’25 so we can capitalize on this market, which is both large and growing. Those investments should be broad-based throughout the business, and the intent behind them is to generate substantial revenue growth over the longer term. As a result, we anticipate only a modest expansion in non-GAAP operating margin in fiscal ’25 compared with the non-GAAP operating margin guidance for fiscal ’24 that we currently have in place. We also expect that these investments will probably be weighted more toward the front end of fiscal ’25. More formal guidance for fiscal ’25 will come on our Q4 earnings call. To wrap up, we’re happy with how strongly we’ve performed thus far in fiscal ’24 and feel good about our fourth quarter outlook. With that said, let’s move on and take your questions. Operator?
Pinjalim Bora
analystOh, fantastic. Thank you all for fielding this. Nice work on the quarter. I'd like to explore the Elastic Cloud piece a bit more. Janesh, looking at this year, it appears the seasonal pattern has shifted somewhat. In Q2 there was a noticeably bigger jump over Q1, which normally runs the other way, and after that Q3 came in somewhat lower compared to Q2, which once again tends to be the reverse? I want to clarify something about Q2 versus Q3. Did we have any nonrecurring items in the second quarter that didn't show up again in the third? Or was the consumption rate just a bit softer than the prior quarter? If I'm reading it right, that 14% monthly figure suggests the unevenness acted as somewhat of a headwind. Any insight you can offer there would be useful. Also, how should we view whether that 29%, 30% growth for total cloud can be maintained?
Janesh Moorjani
executiveHi, Pinjalim. Sure, I'll break that down for you. To begin with, stepping back, our performance across the board this quarter left us quite happy, and that includes cloud, which grew 29% compared to the same period a year ago. Overall, things unfolded pretty much the way we'd anticipated. Breaking that down, the first thing to note is that Q2 didn't contain any special or non-recurring items that failed to show up again in Q3, so we can set that consideration aside. Turning to how Q3 actually went, when we examine results by segment, our commercial and enterprise businesses both turned in strong showings — those are the areas where sales teams lead the charge and we follow up by generating usage once the contracts are locked in. Meanwhile, on the SMB front, where self-service is the primary approach, conditions stayed weak, and that softness has persisted for some time now. That too unfolded just like we anticipated, though it stayed roughly in line with prior levels. Because monthly cloud is growing at a different pace than our other cloud revenue, monthly cloud ended up at 14% of revenue, compared with 15% the quarter before and 16% a year earlier. Overall, I'd say our sales approach is performing pretty well — we're landing those first commitments and then helping those customers increase their usage. Considering the seasonal trends and each of the factors you mentioned, that pattern really just reflects how consumption naturally works, which, as we've noted, can shift up or down at times. Looking at how usage behaved throughout the quarter, I'd describe Q3 as broadly steady in every area, just as we said before. So when I take stock of all those shifting pieces, we felt pretty good about how our cloud business did, and we're eager to put up a solid Q4.
Pinjalim Bora
analystGot it. I just have one quick follow-up, and it's somewhat technical, directed at Ash. When we consider expanding token context windows in LLMs, how ought we to view what lies ahead for RAG? My thinking is this: does that latter development partly consume certain RAG use cases, or will cost essentially serve as the main force shaping how people design these solutions?
Ash Kulkarni
executiveHi, Pinjalim, and thanks for asking. Our conviction is quite strong, and what we hear from customers — the ones I talk with, the ones our team supports — confirms it: RAG, meaning Retrieval Augmented Generation, is a pattern that won't merely persist; it's a pattern that will — that is poised to become the dominant approach moving forward. The main reason, as you noted, comes down to two factors. First, cost keeps rising as you push an increasing number of tokens into those context windows. When people build real-world enterprise applications, that's certainly a concern for them, and RAG lets them keep that expense in check. Second, consider how training a model, or even refining one, actually works: at a certain point you must halt that training and move it into production; yet your internal data — your proprietary data, the basis on which you're trying to build your responses — is always shifting; it shifts by the millisecond. So to deliver responses that are truly contextual and accurate, RAG is essential regardless; those are the reasons RAG is genuinely turning into the main route for people to construct these enterprise applications, and our standing in that space feels very strong to us.
Pinjalim Bora
analystThat's quite clear, and I appreciate it a lot.
Ash Kulkarni
executiveCan you hear me?
Janesh Moorjani
executiveBrent, can you hear us?
Anthony Luscri
executiveOperator, please bring in the following participant.
Tyler Radke
analystYeah, appreciate you taking my question. I'm hoping…
Janesh Moorjani
executiveTyler, hey — your audio is cutting out. We're unable to hear you.
Tyler Radke
analystAll right, is this working at this point? Any improvement?
Janesh Moorjani
executiveYeah, yeah, that's improved. Please continue.
Tyler Radke
analystAll right, excellent. Janesh, could you lay out for us how the different components in Q4 ought to be considered? During Q3, we witnessed stronger growth in other subscription. What's the right way to think about how other subscription and Elastic Cloud grow relative to each other? Then, second, my apologies if this was covered already, is Q4's exit growth the correct top line growth figure to use as the starting point for FY '25? I realize you aren't locking in guidance, but any detail on that would be useful. Thanks.
Janesh Moorjani
executiveSure, Tyler, I'll address each of those. To begin, when I think about other subscription within the self-managed business, and the strength we observed there during the quarter — perhaps to kick things off, let me remind everyone that a customer's decision between cloud and self-manage truly comes down to their own posture and deployment format. Momentum has been good on both sides. While our preference is cloud, and we believe it's better both for the customer and for us, the choice ultimately belongs to the customer. As a result, the mix of business is typically partly driven by deal flow during the quarter. And crucially, we saw no shift whatsoever from customers attempting to move away from cloud toward self-managed. It's really more a matter of the deals we encountered across geos and segments, plus the timing of renewals and similar factors. We had already mentioned on the earlier call that we in fact anticipated a rise in self-managed revenue this quarter. So this didn't catch us off guard. Actually, in Q3 of the year ago period, a similar pattern appeared, with that self-managed business coming in a bit strong. So even looking at it through year-over-year growth, self-managed grew 13% year-over-year, I believe, versus 11% across the prior couple of quarters. So it truly wasn't a significant deviation. And so, turning to how that gets factored into our Q4 guidance — one way to look at it is that Q4 guidance sequentially sits only $1 million above what we just reported for Q3. There simply aren't many moving parts in there, so I don't anticipate much movement; obviously we don't guide to the individual pieces around cloud and self-managed, but that's approximately how I think about it. And just remember that Q4 has two fewer days, since we have 90 days rather than the 92 days we typically have in Q1 to Q3. That creates a bit of a headwind of, call it, roughly $6 million to $7 million of revenue in Q4. That's already been factored into the guide as well; and that's just something to remember when considering sequential growth rates. And then last, regarding how to think about growth for fiscal '25 — look, at this stage I think it's simply too early, right. Through 2024 up to now, the business has turned in genuinely impressive results. How we've carried things out has left us quite satisfied. Planning for 2025 remains underway at this point. A specific revenue forecast would therefore be premature; at present our attention stays on finishing Q4 strongly, and further discussion of 2025 will come during the next call.
Tyler Radke
analystThanks; and as a follow-up for Ash, that customer case where a vector database got replaced — I'm drawing a blank on which customer it was, but the one that's a coding repository company. Could you maybe walk through why — that — I mean, vectors are still quite early as a nascent industry. How often do these replacements happen? And what were the primary reasons that led them to look at another option and go with Elastic?
Ash Kulkarni
executiveYes. Thank you, Tyler. So as customers become better informed about all that's involved in creating Generative AI applications, what we see is that our key differentiators, the ones we've discussed previously, are truly starting not just to stand out but to be very clear to customers. And when I speak with customers, consider the examples I mentioned last quarter, whether that was DocuSign, or earlier, Cisco and so on, this time, Stack Overflow, Consensus. What stands out for these customers is, I'd say, about four items. The first is our vector database implementation built on top of the SIEM; it's simply an exceptionally strong implementation. Customers genuinely observe us performing well, particularly regarding scalability; and I've discussed the types of new innovations we've introduced there around scaler quantization that's making the system perform even better, scale even better, and be less demanding when it comes to memory requirements. So the vector performance alone is one item, the first one. The second one, which is quite powerful, is all the capabilities — the enterprise capabilities we've brought into the product, such as the ability to handle document level permissions, the security capabilities we have, the ability to leverage other facets of search like personalization and geolocation and so on, which truly matter when you're building these enterprise applications. The third one, which is now growing very popular, is hybrid search. So hybrid search is a mechanism in which you combine vector search and semantic search and lexical search and apply re-banking to arrive at the most relevant answer — that is, for customers as they grow more sophisticated in this area. They're realizing that getting the right relevance is incredibly important, to use those kinds of hybrid techniques. And then the final item is simply incumbency. We — for all of these customers, if you consider Elasticsearch and how widely we're spread throughout the ecosystem, tens of thousands of customers and then many more who know and use Elasticsearch, the fact that our data is already on that platform, their data is on our platform, just means it becomes far easier for them to take all of that data, very quickly create vector embeddings, and then use it across all their Generative AI applications. Those 4 elements are turning into increasingly powerful methods for us to stand apart; and it's my firm expectation that as this entire field keeps maturing, those advantages will grow into larger considerations enabling us to keep winning in this arena.
Tyler Radke
analystThanks.
Matt Hedberg
analystExcellent, and thanks for fielding my questions. Nice work on these results, everyone. Ash, it's clear a big player in firewalls discussed its intention to package products together, with no-cost试用 included. I'm curious whether you believe that ultimately manifests as broader market price strain, and affects areas such as logs or SIEM?
Ash Kulkarni
executiveIndeed. Appreciate that, Matt. To date, we've observed no shift whatsoever in how competitors are behaving. That said, what I can tell you is this broader notion of consolidating onto platforms. When I speak with clients, my observation is that each CIO I encounter remains highly mindful of the total budgets they manage; consequently, financial constraints persist, while at the same time there's enormous enthusiasm about pursuing more initiatives involving Generative AI. That combination is turning into something of a two-edged situation, which pushes customers toward concentrating their efforts on a select number of platforms they have confidence in, that they perceive as delivering innovation, and that they believe can assist them not just now but will also safeguard their future trajectory. And we happen to be among those who gain from this. The reality that we're able to provide exceptional innovation. We've consistently brought forward advances such as Generative AI, plus our AI assistance covering observability and security, and we do so using a pricing approach that lets customers consolidate while achieving significantly greater value relative to what they invest, and this is what allows us to prevail and push out established competitors. So that platform consolidation trend, we're actively embracing it; we've actively embraced it across multiple quarters already, and the advantages are becoming evident to us.
Matt Hedberg
analystIt's genuinely great to hear that. Then perhaps this — clearly there's plenty of discussion around vector and GenAI adoption; and it seems many of those use cases fall within enterprise search, or you know, app or site surge. What I'm wondering is, how far do you expect vector technology to reach into something like broad observability, or security search with SIEM?
Ash Kulkarni
executiveExcellent question. And while we firmly believe that Generative AI along with vector search and similar things will significantly expand the total addressable market for search over the long term, we're also noticing that our capacity to stand apart and compete more effectively in observability and security is improving; so to give you a bit of a feel for how this typically unfolds on the security side, we've now been through a number of elastic on events. And what reliably gets everyone in the audience tremendously fired up when we speak with security customers is our security AI assistant. The ability for them to now view real-time alerts and truly question the system and pose queries regarding how to handle those alerts, with the AI assistant capable of helping steer them through all of it — that is remarkably powerful, since it lowers the bar for what a security analyst must know, the amount of experience and expertise they have to bring; so this makes it far simpler for us to compete and win in these deals and push out incumbents. So we believe it's already providing us with that advantage. I anticipate that advantage will keep expanding going forward. And that is why we're so thrilled about everything we're doing in the area of Generative AI.
Matt Hedberg
analystThat's good to hear. Thank you, guys.
Raimo Lenschow
analystHey, thanks. Congratulations from my side too. I have two quick things. First, Ash, given that leadership on the go-to-market side has now shifted, is anything different showing up as you work through evolving how you go to market, and is there anything we ought to expect as we prepare for the coming year? Then I have one follow-up for Janesh.
Ash Kulkarni
executiveSure. We're really thrilled about Mark joining us. Even across the brief period he's been with us, he's simply been a fantastic partner, and it's striking to watch how deeply he's thrown himself into the role. He's collaborating quite closely with the team members, and the innate leadership he's showing is terrific. What I would say is this: the path we're on, our emphasis on the enterprise and commercial segments, our broader go-to-market approach — my expectation is that stays the same. We're enjoying a great deal of success with it. Janesh brought that up before, and we're quite enthusiastic about the know-how Mark contributes in those areas, and we recognize that as the business keeps performing well in areas such as Generative AI, when it comes to platform consolidation, Mark's abilities will only assist us in doing even better. And right now, I don't foresee any change; it's really the case that the progress we've achieved and the direction we've established — my feeling is that Mark will build on that rather than alter anything.
Raimo Lenschow
analystAll right, that's great. Yeah, that tracks. And Janesh, turning to you — I appreciate the early peek at '25, and I realize guidance isn't out yet; still, if I take those investment areas and slice them a bit finer, what's the right way to frame it? Is it sort of a blend of sales — meaning you have to build out sales capacity, and if you're thinking about recovery, you'd need additional salespeople, and it takes time before they're productive? Then on the R&D side, clearly a great deal is happening given all the shifts in the industry around AI. So is it simply uniform across everything? Or are there particular areas of emphasis worth considering? Thanks, and congratulations once more.
Janesh Moorjani
executiveThank you, Raimo. When I look ahead to fiscal '25, obviously our continued emphasis will be on profitable growth, and more specifically, revenue will increase at a quicker pace than expenses; however, while we work through the planning process at present, our expectation is that we'll strike a balance between putting money into long-term growth and delivering near-term profitability. That's the reason I highlighted the fairly modest op margin expansion we anticipate for next year; thus, throughout this last year, we've observed a rather sizable opportunity in front of us when it comes to GenAI's growing significance and the enormous opportunity available for us to go seize. So our genuine intention is to step up our investments in that space so we can go capture that market opportunity. And my view is that these investments will in fact be broad-based throughout the business. We'll be looking to fund engineering. There will undoubtedly be certain go-to-market investments along with support capabilities so that we can unlock this opportunity. And truly perform well over the longer term. Yet our expectation, Raimo, is that such investments will underpin our growth trajectory over the longer term. That's the reason I said we'll be making the investments at this point, and that will create a slight effect on non-GAAP op margin for next year. Those investments will additionally be somewhat more front-end loaded in fiscal '25; and we'll give a bit more color along with formal guidance on this once we reach the next earnings call.
Raimo Lenschow
analystAll right, thanks.
Andrew Nowinski
analystThanks, and good afternoon, everyone. There are a couple of things I'd like to cover. The first concerns ESQL. You mentioned that one benefit of Elastic's vector search stems from the fact that a great deal of customer data already resides on your platform. Could ESQL serve as a means of attracting additional data to your platform, provided it lowers the barrier that keeps people from moving off Splunk?
Ash Kulkarni
executiveYeah, that's an excellent question, Andrew. As I said during the prepared remarks, ESQL has me really enthusiastic. Roughly 1,000 customers are already testing it, and it's only been a short while since we rolled it out. What this gives us is a boost in how developer friendly we are. As far as the way the language was built, it's extremely simple to work with. It features piping semantics, supports iterative development, and developers are big fans. Beyond that, it also lets customers shift quite easily from established legacy SIEM offerings over to our platform, since a number of those offerings already had comparable functionality built around pipe semantics and similar things; that obstacle no longer exists. Customers can adopt it without delay. And once you combine ESQL with our AI assistance, which can translate other languages into ESQL, the whole thing becomes really straightforward. That's part of why I'm so enthusiastic about our ongoing capacity to capture share and to enable customers to migrate away from incumbent solutions to Elastic, consolidate on our platform, and expand. So over the long run, I firmly think this will draw more data to our platform and make us considerably more essential to the broader infrastructure our customers rely on.
Andrew Nowinski
analystAll right, thanks. That was excellent. I just have one brief follow-up. Considering everything you've now brought to the platform, including ESRE, ELSER, ESQL, and so on, has your pricing been adjusted to reflect these new capabilities?
Ash Kulkarni
executiveYes. So Andrew, one thing, as you're aware, is that pricing works across multiple tiers. Based on which tier you're in, you unlock various types of functionality. For instance, the AI assistance sits in our highest tier, the enterprise tier, while some ESRE functionalities are in our platinum tier, and so forth. So to obtain all of these additional capabilities, customers must move up a tier. Regarding price increases, we've raised prices periodically; we've done so at times for self-managed. We've done so at times for cloud. But really — we'll do that when it makes sense, given the functionality we add. Let me check whether Janesh has anything to add.
Janesh Moorjani
executiveNo, nothing, Ash. I think you covered it quite nicely.
Andrew Nowinski
analystAll right, appreciate it, everyone.
Patrick Colville
analystThanks for letting me join this call — it's genuinely exciting to be a part of what Elastic is doing. So I'd like to dig into revenue: 29% constant currency in 3Q is honestly outstanding. You mentioned a few things driving that; Gen AI was one, along with observability and security consolidation, plus consumption returning to committed levels. So my question to you, Ash, is whether you could put those three items you mentioned in order of importance — which one mattered most in 3Q?
Ash Kulkarni
executiveYes. Appreciate that question. Here's how I'd put it: when it comes to Generative AI, it is certainly adding to revenue, though considering how large we are overall, I'd characterize that addition as still early. Regarding the consolidation we're witnessing on our platform — incumbent solutions getting displaced by customers, and customers bringing observability and security together onto our platform — that's something we've been pursuing for a number of quarters now. Because workloads typically need time after arriving to scale up, as you'd expect, and since we've been at this for a few quarters, it has contributed very strongly to business growth, and stable consumption also plays a part. So that's roughly how I'd frame it. That means consolidation is likely the largest factor, while steady cloud consumption is probably what keeps everything balanced, so to speak. But Janesh, I'm not sure whether you'd like to chime in?
Janesh Moorjani
executiveYes, I'd agree with that ranking, ash.
Patrick Colville
analystThanks, Ash. And I suppose, Janesh, my follow-up is directed at you, please; so non-GAAP operating margins for the third quarter came in at 13%, with nothing at all to strip out. From where I sit, a few investors are being picky, but look, that matches what we saw in Q2 and the guidance points to a step down in Q4. Could you just drill into what's happening when it comes to profitability? Is there anything one-time worth flagging in Q3? What's behind this decline in Q4?
Janesh Moorjani
executiveYes; so glad to address that. When I look at the shift between the third quarter and the fourth quarter, we're clearly happy with the beat we posted in Q3. But in Q4, one item we flagged previously is that, we have roughly $12 million more in seasonal expenses during Q4. I noted this on the earlier call, this ties to when employee benefits land and our engineering all-hands events. And then too, as I said a bit sooner on this call, because the quarter has 2 fewer days, we face an inherent drag of roughly $6 million to $7 million on revenue. So those two things combine, and that explains why Q4 comes in lower. The seasonal items hit only Q4. And so as we see it, we truly concentrate on steering the full-year figure; and as you observed here, we've lifted our outlook for the full year. We had lifted it even sooner in the year, and we're quite pleased that we can beat even our earlier commitments, and we anticipate closing the year out strong.
Patrick Colville
analystOkay, glad to be involved in this, and thanks for the invitation.
Janesh Moorjani
executiveThanks.
Brent Thill
analystBrent, I realize RPO moves around quite a bit, but was there any slowdown hiding under the surface that we ought to know about?
Janesh Moorjani
executiveHi Brent. On the whole, I'd say the annual contract selling motion has us genuinely pleased, and that includes the way sales managed to deliver a really robust Q3. We've discussed certain trends we observed in the business, such as platform consolidation and similar items, and every one of those points to that success. When I think about RPO, perhaps the single item worth flagging is that contract lengths came in close to what we'd seen before in the business, at about 1.5 years, though they ran a bit shorter than the comparable period 12 months earlier. As a result, RPO may shift somewhat depending on the terms of the deals we closed in a consumption business. You also get differences as customers attempt to align the timing of their renewals and expansions with the pace at which they consume. So following it quarter to quarter can be somewhat noisy. There's nothing else I'd really flag. At the fundamental level, RPO growth matched revenue growth and every other metric. And it was broadly consistent with the rest of the business. And as we've said in the past, we treat revenue as truly the main indicator of how well our business is doing, and we're extremely pleased with how that turned out.
Brent Thill
analystThen moving to Ash regarding AI — you brought up some initial successes, though you noted it remains quite nascent. Are there any shared themes you can identify when it comes to specific industries or geographic scope? What commonalities are beginning to emerge? Is there anything you could do to accelerate awareness, given that you've watched competitors launch boot camps and run smaller-scale rollouts, which appear to be connecting well with others in the market? Is there anything you're picking up and advancing as this AI wave progresses?
Ash Kulkarni
executiveYeah. I appreciate that question. So among the trends we're certainly observing, first, it spans a wide range of sectors. I need to cite a few instances in what I prepared. Consider certain technology firms such as Stack Overflow and Cisco that I've referenced previously, plus others operating within document management — that being the one I brought up, Consensus, which focuses entirely on research, on scientific research — it's thoroughly broad-based. Many of the applications remain internal in nature, since customers keep gaining greater confidence in their capacity to handle hallucination risk, and clearly the RAG pattern, which they rely on Elastic to construct, genuinely provides them the strongest shot at cutting down hallucinations — yet that's something customers are still ensuring they mature into — so that's pattern number two. A substantial amount of internal facing work persists; and eventually that will shift, correct? Now regarding the way we're propelling this and the way we're bringing it to our customers. Look, I've stated over the past three quarters that each quarter, we've continued adding several hundred customers on an incremental basis who now use us for these vector search applications, storing dense vectors, performing vector similarity searches. I've discussed how our ELSER Sparse Encoding model has been adopted. So we're quite thrilled by the momentum we're witnessing when it comes to customer adoption. And the manner in which we're pushing this is essentially via our sales team and the marketing events we're holding — we've completed six of our 12 Elastic on Events scheduled for this fiscal year, and six additional ones remain in this fourth quarter. And GenAI occupies the central spotlight, and we're running boot camps. We're conducting GenAI Workshops in a Box alongside our customers; these have proven broadly successful. So I'm quite enthusiastic about the early momentum. As you noted, it will require time for this to develop; so we remain in the early innings; yet across the long term, I think this will deliver a very material expansion of TAM for us.
Brent Thill
analystThanks.
Andrew Sherman
analystOh, excellent. Appreciate it. Hi, everyone. Janesh, that annual cloud figure looked quite robust — holding steady with growth in the mid-40s. So is it truly just the SMB strain that's pulling the total down? And what's your assumption for that segment moving ahead?
Janesh Moorjani
executiveYeah. I'd say the calculation you're performing is fairly simple; and you're correct. I mean, at the core, the annual cloud selling motion has left us quite happy, and that's a sales-led motion, plus both the commitments we've obtained, which Ash has discussed, and the consumption we've generated toward those — toward those commitments. And regarding the SMB piece, like I said a moment ago, it's still soft. That matched what we'd seen previously. As for looking at what's ahead, we won't break out a segment-based perspective on cloud in this setting, but at the core, when I consider our cloud business as a whole, the opportunity set there keeps us genuinely enthusiastic, and we anticipate it will keep expanding quicker than the remainder of the business.
Andrew Sherman
analystSo, Ash, in the pipeline, are you noticing any signs yet that more clients might get in touch with you as a result of possible disruption at your biggest competitor in logs?
Ash Kulkarni
executiveAndrew, I appreciate that question. Honestly, nothing specific in terms of disruption or change stands out to me. What I can tell you is that the entire effort to bring customers and prospects together on our platform by pushing out the incumbent players — that's something we've been working on for quite some time. To us, this looks like a huge opening; we hold so many advantages in log analytics, SIEM, and search, and we consistently enter accounts through those — and after we're established there, that opens the door for us to grow even more broadly; and given everything happening in the market, plus what we've built, the fact that customers can now take advantage of things such as ESQL. You and I have discussed Frozen Tier before, and it delivers enormous value to customers, along with our AI work in observability and security AI assistance. Every one of those plays a major role in why customers today feel so at ease and so sure about switching to our platform, and that's genuinely thrilling.
Andrew Sherman
analystGreat, thanks guys.
Koji Akida
analystHey everyone, I appreciate you letting me ask a few things. Historically, whenever we've looked at this company through the lenses of enterprise search, security, and observability, well, previously you described approaches for understanding how fast each of those three areas might expand. So I was hoping you might drill down a bit and share your current thinking on what growth rates these three categories could see across the medium term, which would help frame your medium-term ambitions for growth?
Janesh Moorjani
executiveKoji, let me try to address that, and then I'll ask Ash to chime in with anything further. When I look at how things break down among the solution areas, Q3 as a whole lined up with the pattern we've typically observed over the past year where ACV is concerned. It does fluctuate somewhat, depending on the flow of deals, renewal contracts, timing, and similar factors. Looking at the medium term, each of the 3 solution areas is expanding well for us. Should any shift in mix happen, it would happen gradually. That means we aren't making active efforts to alter the mix. Looking ahead, search ought to gain from GenAI, which, as we've mentioned previously, we see as a broadening of the TAM. That creates a major opening for our business. Beyond that, we see a sizable chance to expand in security and observability, thanks to the consolidation trend we've observed and the robust GenAI features we've embedded in the platform. Those capabilities genuinely contribute to strengthening security and observability too. Take, for instance, the AI helper we introduced and have previously discussed. That's how I view things over the medium term. Ash, is there anything you'd like to include?
Ash Kulkarni
executiveNo, I think you nailed it, Janesh. Nothing more to add.
Koji Akida
analystFantastic, everyone, I really appreciate it. Thanks a lot.
Janesh Moorjani
executiveAppreciate it, Koji.
Brad Reback
analystThank you so much. Janesh, when it comes to the investments you're describing for the coming year and past that, is it necessary to put money into COGS to enable all of this AI capability?
Janesh Moorjani
executiveBrad, when I look at how our investment profile shapes up going forward, a natural rise in COGS is going to occur. That, I'd say, stems largely from the cloud expansion we're seeing. Nothing structurally very different is what we anticipate. Having said that, as our serverless offering gets rolled out in the months ahead, there could be certain upfront investments tied to it, though we'll incorporate that into the model at the point we give guidance in a more formal way for fiscal '25. Still, nothing significant is what I'd point to right now.
Brad Reback
analystBrad, thanks a lot — that's great.
Ash Kulkarni
executiveAsh, let me jump in here for just a second to build on what you said. There's one point I really want everyone to walk away with: delivering RAG functionality doesn't require us to have things such as GPUs. Our capabilities run extremely efficiently on CPUs, and in the cloud we offer instance types tuned for exactly that — they're simply standard instance types, time memory instance types. So when you look at it from a COGS standpoint, a big part of why customers genuinely appreciate what we provide in the GenAI space comes down to the efficiency overall. I wanted to make sure I mentioned this, because if any questions come up regarding GPU usage and similar topics, I'd like there to be some clarity on the matter.
Brad Reback
analystExcellent, thanks.
Mike Cikos
analystHi, appreciate you letting me join, everyone. This has two parts. I suppose the first one is for Janesh, if that's alright, though I recall that earlier we discussed how price adjustments could happen, yet it really appears that you're leveraging the multiple tiers to push adoption upward toward a higher tier for certain new capabilities. And I'd simply like a clearer picture — when we consider the revenue growth Elastic is showing these days, should we believe a bigger portion is coming from customers moving up a tier? Does that deserve a mention at this point? Or does it remain comparatively minor in the overall picture? And let me also pose my second question while I have you, though this one's for Ash — I realize we were pointing to ESQL as genuinely assisting in smoothing out some of the migration challenges involving legacy incumbents; and I simply want to confirm I'm interpreting that correctly. Could you walk me through whether, if I'm working with ESQL, that's a route for me to begin moving across perhaps freshly created data or to power new use cases? Or does it actually involve transferring whole workloads and data estates that had been stuck with those legacy incumbent SIEM products?
Janesh Moorjani
executivePerhaps I'll take the first question and be concise given the time constraints. Regarding the higher tier and the upsell opportunity there, we've typically observed a fairly consistent rise in uptake of our higher tiers as time passes, especially the enterprise one; that's been visible for a number of quarters now, and we anticipate this will keep fueling Elastic Cloud growth going forward. So I'm not foreseeing any specific inflection point in our modeling; it simply aligns with what we've witnessed for a good while already. We regularly introduce extra features into our higher subscription tiers. Our sales team shows customers what that's worth. That's simply a component of our routine upsell motion.
Ash Kulkarni
executiveLet me also say a few words about ESQL. The point of ESQL isn’t to pull data in from incumbent vendors. That has never been difficult for us. The thing is, those incumbent systems offered query languages that let users build interactively through piped query semantics. As a result, anyone moving onto the Elastic platform faced some friction, since they were accustomed to working a particular way. Now, though, we can simply convert the rules they’ve already written into ESQL and hand them that ESQL query language—something they know very well, and a development style they know very well—and that is what lets people make the switch far more quickly.
Ash Kulkarni
executiveMuch appreciated. I appreciate everyone being on today's call. We're happy with how we did in Q3 and with the solid execution, since the concentrated selling approach we've taken with enterprise and commercial customers is producing results. We feel enthusiastic and assured regarding Q4 and how robust the business is as we move toward FY '25. Thanks once more, and enjoy the rest of your day.
Operator
operatorThat brings this call to a close. We appreciate your joining us for today's presentation, and you're welcome to hang up at this point.
Read the full transcript via the API
You're viewing the first half of this call. Get the complete Elastic N.V. transcript — plus 132,000+ transcripts from 7,000+ companies, speaker segments, AI summaries and full-text search — through the EarningsCalls.dev API.
Get the API View API docs →This call discussed
For developers and AI pipelines
Programmatic access to Elastic N.V. earnings transcripts and 132,000+ others is available through the
EarningsCalls.dev REST API. Plans from $39.99/month — full transcripts, speaker segments,
full-text search, and the recently-added /api/v1/transcripts/recent polling endpoint for ETL pipelines.
More tools for stock research
OptionsBell: an unusual options activity scanner that tracks options flow across all US stocks and emails you when large trades appear in Elastic N.V. or any ticker on your watchlist.
StockMarketScan: a stock screener with alerts: 24 pre-built screens refreshed every 5 minutes, watchlists and email notifications. Both offer API and MCP access and fair monthly pricing.