Elastic N.V. (ESTC) Earnings Call Transcript & Summary
May 30, 2024
Earnings Call Speaker Segments
Anthony Luscri
executiveThanks. Good afternoon, everyone, and we appreciate you being with us for this conference call about Elastic's financial results for the fourth quarter of fiscal 2024. Joining us are Ash Kulkarni, our Chief Executive Officer, along with Janesh Moorjani, who serves as Chief Financial Officer and Chief Operating Officer. Once they finish their prepared comments, we'll open things up for questions. We put out our press release earlier today after the market closed, and you can find it on our website. You can also locate slides that supplement this call on the Elastic Investor Relations website, which is at ir.elastic.co. During our discussion, we'll be making forward-looking statements, and those might contain predictions, estimates, or expectations about demand for our products and solutions, our future revenue, and other matters. Such forward-looking statements rest on factors we currently know, are valid only as of today's call, and carry risks and uncertainties that might lead actual results to differ materially. We take no responsibility for updating or revising any forward-looking statements except where the law requires it. Please review the risks and uncertainties laid out in the press release we issued earlier today, including the slides posted on the Investor Relations website, as well as those described more thoroughly in our filings with the Securities and Exchange Commission. We will also cover certain non-GAAP financial measures, and you can find disclosures about non-GAAP measures, including reconciliations to the most comparable GAAP measures, in the press release and slides. A webcast replay of this call will be made available on our company website through the Investor Relations link. Our quiet period for the first quarter of fiscal 2025 starts when business closes on Wednesday, July 17, 2024. We'll be attending the Bank of America Global Technology Conference and the Baird Global Consumer Technology and Services Conference during the week of June 3, plus the Mizuho Technology Conference and the Rosenblatt Technology Summit during the week of June 10; with that, I'll hand things over to Ash.
Ashutosh Kulkarni
executiveThanks, Anthony, and thanks to all of you for being on the call today. Elastic turned in another robust quarter and a terrific close to the fiscal year. Once again we beat the guidance we had given on revenue as well as profitability metrics; during Q4, revenue was up 20%, cloud revenue was up 32%, and our operating margin came in at 9%. We kept executing as well on our land, expand and consolidate motion, and the count of customers who spend more than $100,000 with us rose to over 1,330. Twelve months ago, we began FY '24 with the release of our Elasticsearch Relevance Engine, which included our vector database along with our Sparse EncodeR model, ELSER. All through the year just ended, we kept reinforcing our standing as the platform of choice for creating real-time generative AI, or GenAI, applications. And I'm glad to tell all of you that we now have well over 1,000 separate paying customers who use our vector database and retrieval-augmented generation, or RAG, capabilities to build GenAI applications. I take particular pleasure in the robust GenAI adoption we see among our biggest customers, with more than 145 of those accounts holding annual contract values above 100,000 already putting our GenAI capabilities to use. The count of GenAI customers within that category has more than tripled over the last year, and that marks the quickest rate of adoption we have witnessed for any major new capability we have launched previously; over time, we believe every organization, big and small, will harness the power of AI to transform their businesses. For us this is a sizable market opportunity that will unfold across the long term. Each customer interaction I had during Q4 involved customers actively looking to explore and apply GenAI for business benefit. Our customers are applying GenAI to improve every kind of business process and to transform customer and employee experiences. That includes customers ranging from newer digital native companies such as Roboflow, to more mature enterprises, among them some of the biggest companies in the Fortune 100. Roboflow is using the Elastic platform at enormous scale to build innovative GenAI applications. Hundreds of thousands of engineers use the Roboflow platform to create data sets, train models and deploy computer vision models in production. Roboflow relies on the Elastic vector database to store and search billions of vector embeddings at extreme speed. During Q4, within the enterprise segment, one of the world's largest financial services institutions expanded with a multiyear 8-figure deal with Elastic. That company uses Elastic for search and observability in a center of excellence with multiple use cases across the business. The company is now using Elastic's vector search capabilities in a strategic new application to provide highly accurate AI-based real-time recommendations for wealth managers to support the organization's most valuable clients. Clients will not only receive better service from wealth managers but the managers will be able to double the number of clients that they can support. For this fresh generative AI application, the company picked Elastic instead of other vendors, citing our groundbreaking technology, faster returns, and the capacity to leverage their own internal tech know-how to roll out this advanced interactive tool rapidly. Take another instance: a sizable multinational electronics firm based in Asia broadened its current reliance on Elastic and selected us for vector search plus retrieval-augmented generation, or RAG, to drive their in-house employee support system. For the semiconductor division, this new LLM-based Q&A system should boost both worker productivity and product quality. Yet another major worldwide financial services organization committed to a bigger deal during this quarter, adopting Elastic's GenAI features for searching bank policies, which assists every employee in locating and comprehending the company's intricate, interlinked regulatory rules and processes. This ought to drive greater efficiency and output throughout the enterprise, while also enhancing the general regulatory stance. As the search AI company, Elastic holds a distinct advantage in enabling our clients to seize the revolutionary opportunities offered by generative AI. We merge the accuracy of search with the intellect of AI, enabling customers to create pioneering applications. The robust and ongoing uptake we observe for our GenAI features, along with our steady stream of innovation that widens our competitive edge here, strengthens our belief in our capacity to gain long-term from the enormous surge of business change driven by GenAI. When it comes to observability and security, we assist clients in tapping search AI's potential to make their organizations more robust by injecting greater intelligence and automation into their observability and security tools. During Q4, our AI prowess, encompassing our AI assistance for observability and security, kept helping us push out established observability and security providers and bring customers together on the Elastic platform. As clients aim to gather onto fewer platforms to cut expenses while speeding up their innovation with AI, Elastic has remained a winner. This stems from our unwavering dedication to innovation and our fixation on putting customers first. For instance, a top global supplier of insurance, annuities and employee benefits inked a fresh agreement for Elastic Cloud on Azure. This firm brought several tools together onto Elastic as part of its AIOps approach, enabling teams to operate quicker, more effectively and with greater precision. Elastic won out against rivals due to our capacity to take in any data type, unify and link all data on a single platform, and automate observability through the Elastic AI assistant; now shifting to products. In Q4, our team kept rolling out some eagerly awaited features to further widen our competitive edge across search, GenAI, observability and security; we introduced an unprecedented Search AI Lake and revealed the technical preview of our brand-new Elastic Cloud Serverless product, which sits atop the Search AI Lake architecture. In the past, data lakes provided inexpensive, long-lasting storage for data, yet speed suffered, along with the capacity to query all information in the lake in real time; because of that, real-time applications have historically been unable to take advantage of data lakes. Addressing that requirement is where Elasticsearch AI Lake comes in as a major step forward. It's an exabyte-scale, cloud-native architecture that comes with search and vector database capabilities built in. Real-time, low-latency use cases—among them RAG, observability and security—are what it's tuned for. Right now, AWS offers our Search AI Lake along with the technical preview of our serverless offerings. Our plan is to broaden this offering to all 3 major cloud hyperscalers and throughout the major geographical regions over the next couple of quarters, and at that point we will make it generally available. For Elastic Cloud, we view this as a thrilling new phase, since it dramatically streamlines the user experience overall, and we anticipate it becoming a growth driver for cloud in the years ahead; when it comes to search and GenAI, we rolled out a number of new capabilities meant to set our offering further apart, beginning with substantial improvements to the Elasticsearch open inference API. Integration with Cohere's vector embedding and Reranking APIs, plus the embedding API for OpenAI on Microsoft Azure, are among these additions.
Ashutosh Kulkarni
executiveWe likewise rolled out enhancements to our main vector database, such as concurrent multisegment graph search plus native code optimizations, which drove major gains in how quickly vector search queries run. On Google Cloud Platform as well as Microsoft Azure, we introduced instance types tuned for vector search, which finished the work of enabling optimized hardware profiles across all 3 hyperscalers. Native support for custom learn to rank models also became available, letting customers put trained models into Elasticsearch directly so that search relevance improves based on user behavioral data. Beyond that, Microsoft shared that Azure OpenAI service on your data now lists Elasticsearch as an officially supported vector store and retrieval-augmented search technology. In a similar vein, Red Hat revealed its backing for Elasticsearch as an officially supported vector store within OpenShift. Turning to observability, Elastic has turned into the platform of choice for customers standardizing on open telemetry. During FY '24, earlier on, we gave the Elastic Common Schema, or ECS, to the Cloud Native Computing Foundation, or CNCF's open telemetry project, as its standard schema for logs; and at this point we've made our profiling agent open source under the Apache 2 License and contributed it to CNCF, subject to their approval. We also added support for AWS Bedrock and the Anthropic Claude 3 model within our observability AI assistant, which sharpened the overall AI assisted experience and gave our customers more options; when it comes to security, we introduced our latest innovation, Attack Discovery, powered by search AI, at the RSA Security Conference. Using the distinctive RAG capabilities of our platform, Attack Discovery correlates, enriches, sifts through and prioritizes actionable intelligence from a flood of alerts, greatly cutting the effort SOC analysts must expend on the alert triage and investigation process; it gives the user an attack view centered on the few relevant attacks that matter rather than the underlying alerts, which can carry far more noise than signal. Typically, SOC teams get thousands of alerts every day and spend many hours triaging those alerts by hand, which lets a significant portion of potentially critical threats slip < >. Through Elastic's AI assistance, Attack Discovery and other AI-powered methods to remediate or prevent issues, we let security practitioners carry out their jobs more efficiently, while assisting our customers in staying resilient with less cost and less effort. According to our customers, they regard this patent-pending innovation as a genuine game changer for the SOC. The SIEM space is shifting once more, this time toward an AI-driven security analytics platform, and we are leading the charge in this evolution. Finally, on the security side, Elastic Security Labs also delivered our LLM safety assessment, together with detection rules that help customers protect themselves from prompt injection attacks and other threats to the secure adoption of large language models and other GenAI technologies. To wrap up, I couldn't be more proud of our team and the consistent way we have executed on our strategy, while managing the business with discipline throughout FY '24. We keep reinforcing our standing as the go-to platform for creating real-time GenAI apps, and momentum is picking up as buyers push out existing vendors and bring more and more workloads together on Elastic. The people who matter most to us — our customers, partners, and developer communities — are the force behind our priorities and the energy behind our creativity. As the search AI company, Elastic sits in an outstanding spot to capitalize on the radical technological change that generative AI is bringing about. On that note, I'll hand things over to Janesh so he can walk through our financial results in greater depth.
Janesh Moorjani
executiveThanks, Ash. The fourth quarter provided a strong close to the year, extending the momentum we've built through solid execution. It's gratifying that we beat the top end of our guidance on every metric once again. Just like in earlier quarters, we keep witnessing many customers consolidating their operations onto the Elastic platform in pursuit of substantial business value and innovation, and we enjoyed another quarter of robust cloud consumption. Fourth-quarter total revenue reached $335 million, a 20% increase year-over-year on both a reported and constant-currency basis. Subscription revenue for the fourth quarter came to $311 million, rising 21% year-over-year as reported and in constant currency. Inside subscriptions, Elastic Cloud revenue amounted to $148 million, up 32% year-over-year as reported and in constant currency. Elastic Cloud accounted for 44% of total revenue during the quarter. Consumption trends remain comparable to what we observed in the previous quarter. Although our customers aren't concentrating on optimization to the degree they were some quarters back, cost awareness and operational efficiency continue to be significant themes for them, which remains a key reason customers select Elastic as their platform of choice for consolidating workloads. Revenue from Elastic Cloud under month-to-month arrangements made up 14% of total revenue. As we've noted previously, Elastic Cloud revenue under month-to-month arrangements is primarily fueled by a self-service motion within the SMB segment, which is still facing challenges. Professional services revenue in the fourth quarter was $24 million, up 1% year-over-year as reported and in constant currency. Even though professional services can swing between quarters depending on the timing of services delivery, we don't anticipate significant variation in its mix over time. To give a bit more color on deal flow for the quarter, we observed a healthy balance across our solutions and kept a similar solution mix in annual contract values compared with the prior quarter. Strength in the quarter was also balanced geographically, with EMEA growing fastest, then the Americas and APJ. Customers keep making robust multiyear commitments to us, which mirrors their preference for Elastic as they evaluate platform consolidation and reflects our growing importance to their business. We also recorded a greater volume of early renewals than is typical for Q4, which mirrors ongoing customer confidence and dedication to our platform. As a reminder, early renewals have no effect on the timing of revenue recognition, so the quarter saw no revenue benefit from this. Our approach of concentrating on customers with a greater propensity for growth is paying off, as shown in our customer metrics. We closed the fourth quarter with more than 1,330 customers holding annual contract values above $100,000. We're happy with this quarter's customer additions in the greater than $100,000 category, since these larger customers offer a solid foundation for our land-and-expand motion as we build a multibillion-dollar company over time. This momentum is likewise evident in our count of customers spending more than $1 million in annual contract value. At the close of fiscal '24, that figure stood above 165, versus more than 140 at the end of the previous year. Just to repeat, this metric — customers exceeding $1 million in ACV — is something we typically share on an annual basis. Turning to a wider view of customer additions, we finished the quarter with more than 4,370 customers whose ACV exceeds $10,000, along with roughly 21,000 total subscription customers. Our net expansion rate came in at about 110%, matching what we had anticipated for the quarter. As you're aware, this metric covers a trailing 12-month period, and the impact of the consumption optimization headwinds felt in certain earlier quarters has now mostly faded. Now I'll move to profitability and cash flow, where I'll cover non-GAAP measures. Gross margin for the quarter reached 76.6%, right in line with what we expected. Operating margin for the quarter was 8.6%, surpassing our expectations thanks to our revenue outperformance. Diluted earnings per share for the fourth quarter came to $0.21. On an adjusted basis, free cash flow margin was 18%, equal to roughly $60 million, in the fourth quarter. We closed out the year with adjusted free cash flow of $169 million, a considerable gain over the previous year and in keeping with the overall improvement in our operating profitability. Last, while cash flow isn't something we formally guide to, we anticipate that adjusted free cash flow margin for fiscal '25 will come in slightly higher than the non-GAAP operating margin for fiscal '25 — much like what we saw in fiscal '24. Because of timing factors and seasonality, cash flow will vary from quarter to quarter, so our practice remains to view it chiefly on a full-year basis. Now on to guidance. As we head into fiscal '25, our focus stays on execution, and we believe we're well positioned to deliver long-term growth and profitability. Our approach to guidance is the same as it was in fiscal '24. We stay prudent about the near term and assume today's business conditions will hold steady. Our sales strategy keeps its emphasis on the enterprise and commercial segments, where we still believe that letting customers choose between on-premise and cloud deployment gives us a competitive edge. With that in mind, we expect ongoing momentum across both self-managed and annual cloud subscriptions, depending on what each customer prefers. As Ash noted, we're seeing robust customer interest in generative AI use cases, and we still believe this will serve as a major growth driver for us over the long term. This market opportunity is a big one for us, though customers remain in the early phases of the adoption cycle. Even as we expect continued growth in fiscal '25, we aren't modeling meaningful revenue contribution from GenAI this year. As we've said previously, fiscal '25 will be a year of focused investment for us, as we plow some of our natural operating leverage back into the business — especially toward the GenAI opportunity. The money we're putting to work will go toward engineering as well as sales and marketing efforts. So with all of that as context, looking at Q1 of fiscal '25, our projection for total revenue is somewhere between $343 million and $345 million. At the midpoint, that works out to 17% growth compared to a year ago, and that holds true whether you look at as-reported figures or constant currency. For non-GAAP operating margin in Q1 of fiscal '25, we're anticipating a range of 9.2% to 9.4%, while non-GAAP diluted earnings per share should land between $0.24 and $0.26, based on 105.5 million to 106.5 million diluted weighted average ordinary shares outstanding. Turning to the entire fiscal '25 year, our total revenue expectation falls in the range of $1.468 billion to $1.480 billion. That equals 16% year-over-year growth at the midpoint, again both on an as-reported basis and in constant currency. Non-GAAP operating margin for the whole of fiscal '25 should come in between 11.7% and 12.3%, and non-GAAP diluted earnings per share ought to be in the range of $1.35 to $1.47, using 107 million to 109 million diluted weighted average ordinary shares outstanding. To wrap up, we have confidence that our growth journey remains in its early stages. Fiscal '24 delivered strong results that please us, and we feel good about our outlook for both the first quarter and the full fiscal year; so with that, let's move on and take some questions. Operator?
Matt Hedberg
analystNice work on those numbers. Of course, conditions for selling are difficult right now, yet it's genuinely encouraging to watch Elastic Cloud speed up. And I suppose this goes to either of you — lately there have been plenty of < > data points regarding consumption-oriented businesses. Just this evening, a database vendor posted somewhat mixed outcomes. You folks, though, are turning in robust Elastic Cloud figures. It doesn't — nor does it appear that GenAI provides you with much of a boost. So I'm somewhat interested: might you give us additional detail about the kinds of consumption patterns you're observing that look somewhat better than what peers face? And perhaps explain why that affects your model in a way that differs from how it affects theirs?
Ashutosh Kulkarni
executiveAsh here — I'll begin, then perhaps hand things to Janesh to expand; yet regarding the broader consumption landscape and what Q4 showed us, it was mostly the case that consumption held firm. Buyers kept consuming, ramping up their usage against the commitments they'd previously entered into; and as we've laid out before, our approach generally involves selling commitments tied to particular customer projects, with usage then climbing toward those commitments. We observed that occurring rather well. The fact that we've managed to secure competitive deals in which we replace incumbents, thanks to innovations we've been advancing — particularly around observability and security — plus the fact that in search we hold a really powerful position. With search, I'd say it matters more today than it ever has, given GenAI and all that's going on in that space; so the consumption patterns we observed stayed quite stable relative to everything we'd witnessed across Q4, not far off from what appeared in Q3. I'm unsure what competitors have disclosed. I haven't gotten the chance to review it, but we felt quite happy about the way Q4 unfolded.
Janesh Moorjani
executiveI think that sums it up nicely, Ash.
Matt Hedberg
analystLet's do a quick follow-up on that. Janesh, from what you're saying, it doesn't sound like you've factored in a lot of revenue coming from GenAI, and yet Elastic Cloud keeps speeding up. So when we look at the growth outlook for the whole of fiscal '25, how are you viewing Elastic Cloud's trajectory across this year? And when do you think GenAI tailwinds might actually show up, since it sounds like you already have 1,000 paying customers using vector and RAG?
Janesh Moorjani
executiveYeah, Matt, here's how I look at it. Looking back at fiscal '24, things unfolded very well for us, and our confidence in this year's outlook is high. First off, regarding the overall guidance framework and philosophy, nothing has changed versus prior quarters; we still guide off what we know and keep that prudent approach. As you saw, we set the initial full-year guide at 16% at the midpoint, against the 19% we just posted for the whole of fiscal '24. Breaking that down, we anticipate momentum in self-managed and in cloud alike. That comes down to what customers prefer, and in our view that actually sets us apart competitively, since we can provide that choice. On the guide as well, we aren't building in any meaningful revenue upside from GenAI, whether in self-managed or in the cloud. From a segment standpoint, our sales strategy remains centered on enterprise and commercial, where cloud annual subscriptions keep growing strongly. As for SMB, which we reach, as you know, via our self-serve monthly cloud offering, we expect it will stay under pressure from the macroeconomic side. So those are among the general puts and takes. And on GenAI, like Ash said, we keep seeing — we expect to see movement there, but no significant inflection is being modeled right now. In my view, that's a longer-term play for us.
Brent Thill
analystAsh, I want to piggyback on what Matt asked regarding the demand picture. It does feel like there's something shifting across the industry. Looking at the top players in front office and back office, and where you compete with Mongo, demand appears to have hit an air pocket. I believe everybody is puzzling over what exactly is going on. One possible reason could be that AI is triggering some sort of review or freeze, and obviously it looks like you're gaining somewhat from that. That said, if you had to put together what your customers are telling you and what you're picking up about demand, are budgets suddenly shifting? Is there a distinct signal emerging across software, as Matt mentioned, that you aren't experiencing?
Ashutosh Kulkarni
executiveYes, Brent, I appreciate the question; the way I would describe it is that demand felt steady all through Q4. Customers are genuinely concentrating on generative AI and how they can put it to work in their business. You can observe that in the sorts of use cases they're attempting to bring generative AI to. I mentioned a few of those use cases, the big Fortune 100 bank applying generative AI to enhance the experience for its high net worth individual clients. Each of these is meant to boost their efficiency, boost the customer experiences they provide. So genuine value exists there and they're working to ensure they can finance those, they are funding those initiatives. And it's certainly consuming a great deal of mind share; when it comes to the degree of optimization and such that we witnessed previously, those have, as you put it, flattened out, yet cost consciousness and operational efficiency stay significant themes for customers; and on our side, we have in fact been leaning into it. We've discussed this over the last many quarters. We lean into it since we think we can provide the innovations we deliver rooted in our generative AI capabilities, both in search for GenAI and also how we bring that to observability and security. I covered that in my prepared remarks and how we are genuinely transforming the game in security. Every one of those is helping us compete more effectively; they're helping us provide greater value to customers at a superior price. It's helping us push out incumbents, and we're gaining from some of these developments. But certainly, aside from the SMB area, which as Janesh said, has stayed flat. It's been flat for several quarters now. We are experiencing the advantage of our capacity to drive differentiation at a superior cost within our customer base.
Brent Thill
analystJanesh, one more quick question. Regarding that 16% growth figure you provided for the year, does it incorporate additional macro and uncertainty headwinds related to an SMB slowdown? Have there been any shifts in your assumptions, or added conservatism, in how you arrived at that guidance?
Janesh Moorjani
executiveYes, Brent. What I would say is that over the past couple of years, we've come to understand that whenever customer spending priorities start to move, staying very close to those customers matters a great deal. Our sales and customer success teams are doing precisely that at this moment. It's also crucial for us to remain prudent in how we see things ahead, and that caution shows up in the guidance we've given. So as we continue forward, we'll keep watching things closely. That said, right now we have a positive view of our outlook, and naturally, we'll keep you updated along the way.
Pinjalim Bora
analystNice work on the quarter, everyone. Ash, I'd like to ask you — clearly, adoption of the GenAI capabilities is going really well. Could you speak to the returns people are realizing with things like vector search? From our own research, it looks like shifting from lexical search over to vector search pushes costs up dramatically, roughly 20 to 30x; so I'm curious, based on what you hear anecdotally, what's the general story on ROI? And are customers favoring sparse encoding over dense?
Ashutosh Kulkarni
executiveSure; Pinjalim, I appreciate the question. We observe demand for both sparse encoded models, such as our ELSER model, and dense vectors as well. What we're finding is that the initiatives receiving funding, at least at the outset, are — a large share of them involve boosting efficiency in workflows tied to support, whether that's supporting customers, or supporting internal as well as external customers. That Fortune 100 bank I brought up, the goal they're pursuing is a quicker method of delivering accurate details about developments impacting the portfolios of their high net worth individual clients as fast as possible; it lets them cut down on the — the amount of time their internal employees need in order to get that information to clients. In practice, each wealth manager can serve a larger number of clients, and the experience overall improves too; so given those sorts of dynamics, the total cost generally trends more favorably, even though you're paying more for the technology, and those are the sorts of initiatives we're observing. And we've had customers, including Cisco or CBC Bank, Stack Overflow, customers I referenced on certain — on earlier calls who have presented at our ElasticON Conference events and discussed the particular metrics and gains they've realized, and many of them saw obvious ROI. So that's the trend we're observing. Naturally, as more and more customers progress along that path of scaling up their use cases, we anticipate seeing bigger and bigger benefits, and over time that will be reflected in revenue for us. But it's genuinely thrilling, and I'm quite enthusiastic about the adoption we've witnessed, particularly within our over 100,000 customer cohort, since that cohort is essentially made up of some of our biggest customers; and as they take on our GenAI capabilities, that will be extremely, extremely significant for us over the long term.
Pinjalim Bora
analystYes, got it. One more question for Janesh. Janesh, this is kind of a hypothetical one about the guidance. If I go back and look at your figures, RPO in fiscal '23 finished the year growing 17% in constant currency. Then in fiscal '24 you delivered 18% revenue growth in constant currency. It looks like RPO is speeding up as we exit fiscal '24 now, yet revenue growth appears to be slowing down. So could you maybe address that gap? And is there some reason why you might have longer duration that's pushing RPO up perhaps. But is there any reason revenue growth won't broadly track ARPU this year?
Janesh Moorjani
executiveYes, Pinjalim, glad to discuss that. There are a couple of items to keep in mind when you look at how RPO has performed. One is, perhaps just to start. The Elastic Cloud from month-to-month arrangements that we discussed, which carries a heavier weighting toward SMB, where macro pressure remains greater. That portion of the business contains no RPO at all. So that's one element to naturally factor into the model. I think the other element I'd also point to — I mentioned this briefly in the prepared remarks — is that we saw a somewhat larger volume of early renewals in Q4 than we typically see in any other Q4. And that simply reflects ongoing customer confidence and commitment to our platform, and it came in a little more than $15 million above what we normally see. And the majority of those contracts were due in Q1. In certain cases, I think those were tied to the single-digit price increase for our self-managed products that took effect in May. Some customers wanted to lock in those lower prices and renewed ahead of time, and that's clearly something we had anticipated when we announced the price increase; and just as a reminder, I said this on the call — on the script as well — that early renewals do not affect the timing of revenue recognition. So there was no revenue benefit from that during the quarter, and it doesn't change anything for revenue for fiscal '25 either. The other thing I'd note in Q4 is that we saw very strong multiyear commitments. Many of these are new and renewing customers, and they all chose to sign multiyear contracts with us. And these multiyear commitments are obviously extremely valuable to us because they establish long-term commitments, but those don't necessarily convert into revenue in fiscal '25; so those are some of the items I'd flag as you consider RPO and how RPO connects to revenue.
Tyler Radke
analystJanesh, to piggyback on the pricing side of things, could you walk through what you've assumed for the entire year? And separately, I understand the technical preview of serverless just came out. Could you address how much serverless, along with other cloud pricing approaches, factors into the guidance?
Janesh Moorjani
executiveThank you, Tyler. Let me take the first item first, the self-managed price hike we implemented. For anyone who might not know about it, we actually increased pricing on our self-managed offerings beginning this month. The bump was a single-digit percentage, driven mainly by all the technology improvements we've rolled out in the platform. That won't be a major revenue contributor in the near term for several reasons. To begin with, it clearly applies to self-managed only, and since it's a single-digit percentage, it isn't all that large; beyond that, it only enters renewal contracts when renewal happens; and because we've signed quite a few multiyear deals, not every one of them will come up during fiscal '25. Then naturally there's the revenue recognition waterfall sitting on top of all that. So the revenue effect isn't significant, yet we keep delivering enormous value to customers who are prepared to pay for what we provide, and the price move reflected that, and in time it will show up in revenue over a longer horizon. On serverless, it won't reach GA until later in this calendar year; and remember, serverless isn't a substitute for the hosted product we offer now. It's a supplementary product; so we anticipate adoption will be gradual and will, again, feed into the model as time passes. Thus we haven't included any material upside from serverless at this stage in the revenue guide.
Ashutosh Kulkarni
executiveAbsolutely, and Tyler, to build on what was just said regarding serverless: our view is that it provides customers with greater optionality, correct? So the existing Elastic Cloud product, which we will soon be referring to as Elastic Cloud Hosted, will keep the same pricing structure that exists at present; it gives customers additional control over their Elasticsearch, the store itself, and the technology itself. When it comes to serverless, that's a more — actually an even more completely managed solution. And it operates on top of the search AI Lake that I mentioned; the pricing approach for that will likewise be consumption-based, though it would rely on metrics we've observed from customers that they want to see from us, meaning pricing tied to how much data is ingested, how much data is stored, and how much is queried. So that's additional choice for customers. It will open up more kinds of workloads. It will provide them with greater flexibility to optimize those workloads in ways that genuinely let them accomplish more and more with Elastic; so I'm extremely enthusiastic about what that will mean over the long term for Elastic Cloud. However, as Janesh noted, since it will only become generally available following the next couple of quarters, we are not modeling any benefit from it in FY '25.
Tyler Radke
analystAll right, that's really useful. If I could ask one more thing, following up on the bookings front — and Janesh, thanks for clarifying that some renewals may have come in early before the price increase, which is what made bookings look so strong. What I want to understand is how this connects to where the net expansion rate is heading. From an ARR standpoint, how do you see upsells and contracts growing, even if the timing happened a bit sooner? Given what you said about optimization pressures letting up, do you think we've reached a point where NRR can begin improving from here?
Janesh Moorjani
executiveYes, Tyler, that's a wonderful question. When we look at the net expansion rate, it came in exactly as we anticipated. There are a few points to make regarding the net expansion rate. First, as you're aware, it's measured on a trailing 12-month basis. That means the impact of the consumption optimization headwinds has begun to fade and has mostly dissipated. Now, regarding bookings or orders, it's important to remember that in the cloud business, the net expansion rate is derived from actual consumption rather than from commitments. As a result, it doesn't influence the net expansion rate right away. For the self-managed side of things, the net expansion rate does factor in the commitments customers have signed with us. That said, on the whole, the momentum in our land-and-expand approach is coming through very well; we clearly delivered a strong close to Q4 and to the full year. So when I consider what this implies for the net expansion rate ahead, we don't give a specific outlook on it, but I can tell you we're quite confident about our expansion motions performing well, in both self-managed and cloud, and the figures we just reported reflect that. So that has us enthusiastic for the remainder of the year.
Ethan Weeks
analystEthan here, filling in for Rob. I'd like to ask about those more than 100,000 customers who are taking up < > at the moment. Could you describe the size of their commitments compared with what they spend on Elastic overall? And are you noticing any meaningful increase in spending from that particular segment of your bigger customer base...
Janesh Moorjani
executiveI can begin, and then Ash may want to chime in; we're genuinely quite happy with what we've observed regarding our GenAI capabilities as a whole within that group of accounts above 100,000. That group, as you're aware, represents our higher-value customers, and it's the main — that pool is what chiefly fuels our business. And since over 10% of those accounts are already using us for GenAI, that's among the — expansion in that category has simply been extraordinarily fast. When I consider the degree of uptake there in terms of what it signifies and in terms of dollars contributed, it remains modest. Most of our customers, as we've noted, are still at quite early stages when it comes to GenAI adoption. And set against the total opportunity we foresee over the longer term, I'd say the near-term revenue contribution is modest and will keep taking time to build, and that's how we've constructed our model. We haven't baked any significant upside from it into the guidance, as we mentioned a few minutes back; yet regarding the use cases for which those customers are choosing us, there's a range of different ones. At this point, most of them lean toward their internal enterprise rather than anything they expose to external customers just yet. Still, I think that simply reflects where the industry as a whole stands regarding GenAI adoption. So across geographies and across various kinds of use cases within that population, we've witnessed really strong traction.
Austin Dietz
analystJanesh, to get at what drove cloud growth during this past 4Q, when we look at the annual cloud business, it grew in the mid-40s during 2Q and 3Q, and it appears it accelerated somewhat further, reaching 50% in 4Q. So I suppose my question is, what has contributed most to the steadiness you're observing in the annual cloud business? And then, how do you feel about how long that annual cloud growth can be sustained as we head into fiscal '25?
Janesh Moorjani
executiveGlad to address that. At a fundamental level, when I consider our land-and-expand motion, it has been performing quite well. Over the past several quarters, we've discussed the investments we've been making to fuel that, along with the strength of the commitments and the strength of the consolidation we are witnessing onto the Elastic platform, and all of that has been unfolding quite well; clearly, we also had a strong finish here in Q4. And so at a fundamental level, those drivers keep playing out, and we are seeing strength from that selling motion in both the enterprise and commercial segments. And as I mentioned, that is offset somewhat on the monthly cloud side by some of the broader macro weakness in the SMB segment; but as we move forward into fiscal '25, we're continuing to focus our sales efforts in enterprise and commercial and continuing to drive those expansion plays, particularly with areas like GenAI and areas like platform consolidation. And I believe that is actually working quite well for us. So we've been pleased with how cloud has performed so far, and we're looking forward to Q1 and the rest of the year.
Mike Cikos
analystThis is < > filling in for Mike Cikos at Needham. You've made it clear that consumption is going to fluctuate from one quarter to the next. That said, is there anything you can point to regarding what seasonality or pattern we should expect for consumption in fiscal '25? And along those lines, do you anticipate that 2Q will be seasonally strong the way it was this past year?
Janesh Moorjani
executiveMike, I can take that one. Looking back at fiscal '24, I'm genuinely pleased with how consistently we executed throughout all 4 quarters of fiscal '24. I'm very proud of what the entire team accomplished. When I think about fiscal '25, we've clearly just introduced guidance for the full year at 16% year-over-year growth and for Q1 at 17% year-over-year growth. My expectation is that revenue will keep ramping nicely as the year progresses; and regarding the particular figures for Q2 and later, I believe it's too soon to say. So we'll give you a clearer picture on Q2 in particular once we reach the end of Q1, but our expectation is that revenue will ramp nicely as the year progresses.
Analyst
analystAll right, that's clear. You've brought up a few times that the SMB segment keeps being soft, but did anything shift from one quarter to the next? And what steps are you taking to push back against that weakness?
Janesh Moorjani
executiveYeah. As far as what we saw across the quarter, things were extremely steady. We have not observed any major underlying changes when it comes to how the SMB segment looks overall. And when I consider the operational moves we're making in that monthly cloud self-service business, it remains a key piece of our business when we look at growth. We keep putting the right investments in place; for instance, the product work we did around serverless will in the end simplify adoption and make scaling smoother for users in that segment, and that naturally reinforces a product-led growth self-service motion; we're aiming to bring our serverless offerings to all 3 hyperscalers by later this year. Beyond those product improvements, we're also concentrating on making the right demand gen investments to keep bringing users to the product; and that motion is paying off quite well for us too. So I'd say it's largely a wider macro effect on SMB that we keep navigating through; but when it comes to both product and applying the right demand gen motions, we keep investing through an efficient cost model so we can keep serving that segment profitably.
Raimo Lenschow
analystCongratulations from my side too — those are incredible results and outlook. Ash, when you consider how customers perceive the direction you're heading in the future, and when you consider all those other companies out there claiming, well, I can do this too, and so on — what are you hearing in your discussions with customers regarding their grasp of what you provide compared with what rivals provide? And how far along do you feel you are on that adoption path?
Ashutosh Kulkarni
executiveRaimo, I appreciate that question. Well, to begin with, I'd say that each and every discussion I've had over the past three months — even where customers were mainly relying on Elastic for observability and security — every one of them touched on generative AI. This is being raised by customers themselves, since each and every client I've talked with appears to be considering ways to apply generative AI broadly to improve their products, their systems, and their processes — making them better, more automated, more efficient, and so forth. Regarding how aware people are of what we can do, I'd say we've succeeded in helping our customers grasp all the vector database capabilities that come built into our platform. We take real pride in the fact that since we made this part of our platform from the start, rather than a bolt-on or a side feature, our customers gain the advantage of it right away, provided they're on the correct usage tier. That means they already possess the skills, their data already resides on Elastic, and now, without needing to pick up anything new, they can get going quickly, leveraging us for these sorts of generative AI applications — whether for dense vector usage or sparse encoded vectors — plus the sheer range of functionality, the integrations we continue to roll out with other companies, and the capacity to tap into all those large language models available today. Each of these has proven to be an enormous strength and benefit. The past twelve months have been a learning process for customers as they come to grips with this changing landscape. What I'd say is that we remain in the early stages of the broader GenAI journey, yet customer awareness — both about what sets us apart and about why they need something beyond a mere vector database, something far more comprehensive featuring hybrid search, featuring reranking capabilities, and so forth, along with those integrations I mentioned — all of that is growing increasingly evident. Consequently, I'm finding myself in fewer situations where I must spell things out, and in more conversations where customers want us to collaborate with them to assist not just in building these applications but truly in discovering what's achievable. So things are progressing quite well, and I'm genuinely enthusiastic about it.
Raimo Lenschow
analystThen, Janesh, here's one for you. In principle, if you consider it, when a customer was relying on search, this ought to be a really straightforward upsell — you simply walk in and tell them, hey, I can offer you hybrid search combined with vector < >, and so on. So could you remind us what share of the various workloads search accounts for? I realize you've never provided exact figures, but does it represent a significant portion of what you did? How does it slot into your broader picture?
Janesh Moorjani
executiveSure, Raimo. In Q4, our solutions mix spanning observability, security, and search struck me as quite aligned with what we've typically observed throughout the prior year when it comes to ACV. Each of the solutions is expanding well for us. On the figures side, observability remained slightly above 40% of the mix measured by ACV, security stayed at approximately 25% or somewhat more, and search then accounts for the remainder. Looking ahead, we believe search stands to gain from GenAI, though we clearly also recognize the chance to generate growth in security and observability through those consolidation motions we've been witnessing. So our overall solutions mix leaves us feeling quite positive.
Ashutosh Kulkarni
executiveOne more point on that: search has become more compelling to customers than at any prior time, and generative AI is the reason. I'm noticing a genuinely intriguing shift in conversations around how ready customers are to put money into it. Beyond that, folks want to explore how Elastic and search AI could improve their observability and security offerings. That explains why all of it appears to be gaining from what we've built in this truly foundational search domain.
Ittai Kidron
analystGreat numbers, everyone. Ash, I was hoping you could share a bit more detail about how the ramp is shaping up among the GenAI clients. To start, congratulations on hitting 1,000 paying customers. Still, if my memory serves, you already had at least several hundred who'd been with you beyond 2 quarters. Is there a consistent pattern when it comes to behavior, adoption, and growth? Could you — is there — anything you can tell us about what we should anticipate for usage growth? What are you hearing from customers regarding how quickly they bring it in? And conversely, how quickly those applications end up generating additional demand?
Ashutosh Kulkarni
executiveYes, that's an excellent question; when you look at it through the lens of the customer journey, adoption is nearly always step one, correct? Which technology platform are they picking, and that's really where the 1,000-plus customer figure comes in — that's the way to interpret it. So it serves as the best leading indicator of traction, along with the 145-plus customers in the 100,000-plus category, since those are the ones that tend to — really, they're our biggest customers; so that's where we anticipate the largest revenue opportunity going forward; now, after customers select us, they move into the implementation stages. And we've had numerous customers put everything into production, with their applications scaling up. And a lot of those customers have come and spoken at our ElasticON Conferences. I named a few earlier, like Cisco or Stack Overflow and so forth. But other customers I brought up during this earnings call, the bank, the electronics company, etc., they're moving their technology into production; and as time goes on, it's going to ramp. One thing you need to keep in mind with generative AI is that within an enterprise context, enterprises typically show far less willingness to tolerate hallucinations than consumers do when using something like a ChatGPT. So the typical length of time it takes customers to run testing and so forth and become comfortable with deploying these tends to be longer. So that's among the reasons you're observing this somewhat longer curve. But once it takes hold, since there's obvious ROI tied to it and the customers who've spoken at conferences have discussed the ROI they've experienced, in our view, this is a question of when. And it's simply a question of timing, which is why we haven't factored it into our FY '25 guide, but we believe very strongly that this will be — it's going to expand our total addressable market in a major and meaningful manner over the long term.
Ittai Kidron
analystExcellent. That's wonderful. Now, for my second question, let's talk about the SIEM opportunity. For years, this has clearly been a major success area for you, with Splunk as the primary rival; however, during the past 3 months, plenty has been shifting in this space. LogRhythm and Exabeam are joining forces, IBM is divesting its business to Palo Alto, and naturally, Cisco is completing its acquisition of Splunk. My question, I suppose, is how do you view the competitive landscape in SIEM as it relates specifically to your company? And given all these developments occurring in the market, in what ways are you adjusting your product, your go-to-market pricing, or anything internally to seize what might be a massive displacement opportunity across the coming 12, 24 months?
Ashutosh Kulkarni
executiveExcellent question; and what we're observing, particularly within the security market, particularly in SIEM, which is our — as you're aware, I've mentioned this before, that's our primary emphasis regarding security. The market is shifting toward the place we've always occupied. We've regarded this as a challenge rooted entirely in data. And with data volumes expanding, it's growing increasingly evident to customers that a data platform is required on the back end capable of ingesting any and every type of security data, including from other security technologies — for instance, if a customer relies on another vendor's endpoint security or cloud security offerings, we must be able to pull that data in and correlate it across everything extremely, extremely rapidly. That's an area where we excel tremendously. The next point involves leveraging AI to automate every process surrounding it, the efforts we've undertaken with our AI assistance, the Attack discovery work we've completed, the market is heading in this direction within the realm of security analytics, and we've benefited from it; we've managed to replace incumbents quite effectively, and we're witnessing the outcomes of that consolidation strategy, the land-and-expand consolidation strategy unfolding with the sorts of dynamics you mentioned. So I'm quite enthusiastic about what this signifies, and we're doubling down on the go-to-market and product fronts regarding this.
Ittai Kidron
analystCould you go deeper on the go-to-market angle? What you laid out involves plenty of product improvements and advantages. Yet what adjustments have you made in how you sell, so you can actually seize this?
Ashutosh Kulkarni
executiveSure. We wrapped our sales kickoff only 2 weeks back, and among the things our sales organizations feel most enthusiastic about is this very topic. Naturally, we've got well-defined campaigns aimed at capturing this chance. Honestly, the message we give our sales force is that there are 2 things we must ask: first, which vector database are you currently working with — that's what we ought to be putting to each customer. Second, whether you rely on any of those established security or log analytics solutions that simply haven't been providing value or pushing innovation forward. That approach is producing wins for us. So for us this represents an unmistakably major go-to-market priority. And the rigor, plus the overall raising of standards that Mark Dodds, who serves as our CRO, brings, is something that genuinely excites me.
Shrenik Kothari
analystCongratulations on the strong execution. So, Ash, you brought up the recent agreement that you signed with the global insurance provider to deploy Elastic Cloud on Azure. And you pointed out the consolidation and data ingestion, which gives you an edge. But considering all the GenAI discussions you mentioned and the fact that Elastic is the first non-Microsoft company on the Azure OpenAI Service, which by itself shows how strong your partnership and relationship are, could you go into the strategic significance of this OpenAI Service specifically and how it is and can make a difference? And then I have a quick follow-up.
Ashutosh Kulkarni
executiveYes, Shrenik. The hyperscalers see a great deal of value in working with us, because when it comes to generative AI and all the areas we operate in, the data is usually already sitting on Elasticsearch, simply because of how widespread we are in the market. We give customers plenty of options, and the reality is that a vector database alone won't cut it. You need all the other search capabilities we've developed across the past 12 years. That's among the reasons the hyperscalers have collaborated with us so closely to weave our technologies together well, so that customers have a great experience. The integration I discussed in my prepared remarks makes Elasticsearch and our vector database functionality accessible via the Azure interface for creating AI applications, which simplifies things for customers, and it will be a quicker route for us to reach the market. And we're seeing that show up in deals such as the one you're mentioning. So, great things across all 3 hyperscalers.
Shrenik Kothari
analystUnderstood. And a fast follow-up for Janesh, picking up where the previous question about go-to-market left off. It appears that you're certainly making progress displacing rivals, plus taking advantage of GenAI and broadening alliances. But regarding GenAI specifically, you noted before that it remains somewhat in an early design stage. It still largely resembles pilot programs. And the new AI workload option keeps developing too. So how are these first programs and data shaping your internal models from a financial perspective — what inning would you say this is? Is it still mostly design phase, or are things beginning to shift or curve upward?
Janesh Moorjani
executiveShrenik, given time constraints, I'll keep this short. Let me simply restate that our guidance was constructed conservatively based on what we know. We haven't factored any substantial upside from GenAI into our model. We anticipate continued growth since fiscal '24 included a minor contribution. Thus, we expect '25 to bring some contribution too, though the amount we're modeling isn't significant. Additionally, we aren't modeling any meaningful upside from serverless, which clearly arrives much later in this fiscal year when it reaches general availability. That's the method we've followed.
Jake Roberge
analystRegarding the 1,000 paying customers that you currently have, would you say those are mostly people already with you who are moving up to premium offerings, or are these new capabilities beginning to widen the top of the funnel and generate some additional new logo business for you?
Ashutosh Kulkarni
executiveJake, let me take that one. Both are happening. That's the straightforward response. It's clearly simplest to convert those who are already customers; however, we're likewise seeing fresh customers, particularly those who arrive via the monthly cloud motion and subsequently become paying customers. So it's happening everywhere.
Jake Roberge
analystAll right, that's helpful. And a quick second question about the pricing adjustments. What's the cadence for rolling those out to the self-managed customer group? Does it happen yearly? Or should we understand this as the first time in several years that you've raised prices?
Janesh Moorjani
executiveJake, this marks the first instance in a number of years — it isn't something we do each year. Typically, we link pricing adjustments to the value we provide. So whenever there are meaningful changes in terms of the value we've added across the subscription tiers, that's the moment we'll step in and make price adjustments, since that's when customers recognize the worth of what they've been receiving, and holding that conversation becomes far simpler at that stage; in other words, doing this yearly isn't our normal approach.
Ashutosh Kulkarni
executiveOkay, thanks so much, everyone, for being here on today’s call. Fiscal 2024 was excellent for us, and I feel genuinely proud of the way the team carried things out, still delivering exciting new ideas and creating value for our customers. We’re thrilled. Thanks again, and enjoy the remainder of your evening.
Operator
operatorThat brings us to the end of this call. We appreciate your joining us for today's presentation. You are free to disconnect at this time.
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.