It's top of the hour, we can get started here. Hello, everybody. My name is Koji Ikeda. I am one of the software analysts on the enterprise software team at Bank of America. We are super thrilled to be hosting a fireside chat with Elastic CFO and COO, Janesh Moorjani. Thank you so much for doing this.
Thanks for hosting us, Koji.
Of course, of course! Just from a high level, you know, for those in the room that are, and on the webcast, that are not familiar with Elastic, maybe just spend 30 seconds to a minute on what Elastic is, and a little background on yourself, too.
Yeah, happy to. You know, Elastic is fundamentally a company that focuses on search-powered solutions, and you can think of us as a data analytics platform to power search-powered solutions at scale. We help companies and customers get insights and information and results from their data in real time, at scale. We help people get answers that matter, right? If you fundamentally think of what a search analytics platform can do, that's what Elastic does.
Yeah.
In terms of my own background, I've been at the company, this is now my sixth year, I feel like I've been there a long time. The company itself was founded in 2012, I joined in the summer of 2017. I joined when it was a pretty small company, now we've crossed the billion-dollar mark, obviously, this past year, prior to that, I spent time in the infrastructure, software, and hardware space at a couple of other companies.
Got it. No, thank you for that. Very topical, you guys reported results last Thursday.
Yeah.
you know, maybe can you give us the key takeaways from the earnings results? What were you most excited about coming out of that?
Yeah, it was a good quarter for us. We were actually very pleased with the way the quarter turned out. Turned out exactly as we had expected and anticipated entering the quarter when we first laid out our outlook. It was also the last quarter of our fiscal year, so we wrapped up fiscal 2023, and there was a lot that we accomplished in fiscal 2023.
As I mentioned, we crossed the billion-dollar mark in revenue, and so that was a big accomplishment. The overall revenue for the year was 28% growth in constant currency terms. We also set ourselves on a path to balancing greater profitability as we continue to grow and scale the business.
As I think about what we achieved here in Q4, beyond beating the expected guidance that we had provided for both the top line and the bottom line, we set ourselves up really nicely for for the upcoming fiscal year and set a goal of 10% operating margin on on a non-GAAP basis, which is exactly what we had said we would do. It was another quarter of real steady execution from us and a macro environment that was stable, and I'm proud of the way the team delivered against what we said we would do.
Got it. I'm asking every management team a couple of set questions. One on the macro, and then, of course, AI, generative AI. We'll get into that.
Yeah.
On the macro side first, you just mentioned macro. I wanted to dig into it a little bit more on what you're seeing out there. You know, how would you categorize the macro environment today? Y ou know, June 2023, versus maybe entering the year? Does it feel more or less the same? A ny sort of help there. How does it feel from, you know, June 2023 versus June 2022?
Yeah, it's a really good way to frame it, Koji, and I would describe it as stable, certainly compared to, say, six months ago. It's a little bit harder and more challenging than it was, say, a year ago. That's the way I would characterize it, and if I think about our business and the impacts that we saw, we really started to see an impact from the broader macro slowing only in October of 2022. That's mainly because of the nature of the use cases for which we are used, which tend to be core operational use cases that tend to be much more mission-critical in nature. We started to see more of a macro impact in October.
If I compare it to the year ago period, it's definitely, you know, more stressed, but it has been very stable in, say, our fiscal Q4 compared to our fiscal Q3, where we didn't see much of a change at all. Customers still take time to make their decisions. They want to partner with Elastic. They are making big multimillion-dollar commitments, multiyear commitments to us as they anticipate moving more workloads over time to us. One of the themes has also been around the pace at which that consumption actually happens, which again, was very stable in Q4 compared to Q3. As I said, the quarter played out like we thought it would, and we did well.
Moving on to the topical generative AI question.
Why would that be of interest to anyone these days?
I'm gonna ask it to you in three different flavors. First question, first flavor. How does your company think about leveraging AI, generative AI, within your offering? Maybe talk a little bit about the Elasticsearch Relevance Engine.
Yeah, I'm happy to. Fundamentally, as a data analytics platform at scale, generative AI is something that is of, you know, immense interest to us and to our customers and ways in which we can add value. We've been investing in machine learning and AI-related tools for quite some time now. If I think about machine learning, we actually started to build machine learning capabilities and embed them into the Elasticsearch platform many years ago. Several releases ago, we announced the introduction of vector search into Elastic.
If you think about how enterprises can get the benefit of generative AI, and if you think maybe you know, if you think of your own experiences as a consumer, on some of these large language models, as you know, they're trained only on information that's available on the public internet. They aren't trained on the most recent information that's available.
If you logged into one of these tools today, and you said, "How many users have used you today?" Chances are the tool would not be able to give you a reply, either because it doesn't know, because it's not being trained on that, or because it's proprietary information of the company that powered that tool, and they would say, "Hey, I'm only trained on publicly available information on the internet, and I don't have information that is proprietary to my owner or to my creator," or something of that sort. People are starting to get lots of interest. Within the enterprise, people are interested in thinking about ways in which they can use that to power their businesses and to advance their businesses.
That's where Elasticsearch comes in, because we can provide the Elasticsearch Relevance Engine and the relevance that you need for a search to be meaningful to a customer. That enterprises can now say, "I wanna use the power of large language models, but apply them to my enterprise data, which I'm never gonna put out on the internet in its entirety." Large language models fundamentally can't leverage all of my enterprise data as they stand, as those large language models stand today.
It allows us to provide the relevance based on the data that exists within the enterprise to those particular outcomes. We provided on our earnings call an example of that as well, and that's just one of the ways. Enterprises have a number of other needs as well.
If you think about avoiding hallucinations, which, as you know, is made up things that appear to be factual, but in fact, are not factual. Enterprises can't afford to have hallucinations in their models. Thinking about about privacy and thinking about security needs, or even just thinking about the total cost of ownership. Things are ways in which the Elasticsearch Relevance Engine can help enterprises achieve those outcomes over time.
With the introduction or I guess, the hyper-awareness of generative AI, call it, from January of this year to June, has this really changed the way that the customers that you're selling to, thinking about the pain points that they're trying to address with Elastic or, you know, just thinking about their data stores in general? I mean, has it changed the conversations at all?
There's a lot of customer interest, a lot of customer conversation around what AI can do for their businesses. Customers are really keen to make sure that they can either stay ahead, that they can be the disruptors in their industry, rather than being the ones who are disrupted within their industries. Generative AI can be a very powerful force to reshape dynamics in a number of different situations. Forward-leaning customers know and understand that, and they are very carefully looking at this. Lots of interest in dialogue with Elastic, as I'm sure with some of the other large players as well.
Got it. last question here, kind of on generative AI and thinking about the monetization opportunity.
Yeah?
How do you guys plan on making money, I guess, you know, with this product? Is it embedded with it? Is it a premium SKU? Could it possibly create other products for you to monetize that market?
It's a great question, you know, as I mentioned, we're deeply engaged with customers together with some of the other larger players, the hyperscalers, who are our partners, specifically, and working with customers to think of ways in which customers can get the benefit of this. If I think about how this is monetized over time, the way we monetize it is fundamentally, if you think about our product offerings, we have Premium level offerings with Platinum and Enterprise subscription tiers. Our machine learning capabilities are only available in those Premium tiers. As customers continue to use technologies that leverage machine learning capabilities and generative AI capabilities, they will naturally need to be in the higher tiers.
Additionally, a lot of these use cases tend to be very compute-intensive, and they tend to be very data-intensive. That will cause all the infrastructure meters to spin a little bit faster as you're thinking about the intensity of these use cases. We see that ESRE , as we call it, is not going to be a dedicated SKU by itself, that we're going to go sell in the near term, but the monetization happens through the higher tiers, the Premium tiers, as well as higher data usage.
When you mentioned the Premium and the Platinum and it, you know, the features are there.
Yeah.
Only available there, what are some of the triggers that the customers see or, you know, usage or reaching a certain type of data ingestion or certain type of use case to all of a sudden trigger, "Oh, we need that higher tier now?" Could you help describe that?
Yeah. I think a lot of it is dependent on the specific use case, right? If they are using machine learning, machine learning is only available in Platinum and Enterprise. They will need the Platinum or Enterprise tier if they are running machine learning jobs that are powered by Elastic against their data. Similarly, if they wanna use capabilities that help them reduce their overall third-party infrastructure footprint, like, you know, lower storage and so forth, they may need capabilities like searchable snapshots and frozen storage. Searchable snapshots is only available in the Enterprise tier. As customers see the need to get more efficient with the infrastructure, it prompts them to actually move to those particular tiers.
If they wanna get certain types of outcomes, as I said, apply machine learning jobs powered by Elastic on the data, then they would need to be in the higher tiers.
Got it. Okay, you've mentioned Elasticsearch Relevance Engine differently, ES?
ESRE.
ESRE.
All right. Pronounce it ESRE.
I'm gonna call it ESRE now. Okay, got it. ESRE, one of the key features of ESRE is vector search. I struggle to try and understand what exactly does it mean? What is the differentiation? Could you just maybe spend a minute or two to really for me, you know, what is vector search? Why is it differentiated? Why is it important?
The old analogy comes to mind, I don't have to run faster than the bear, I just have to run faster than you. I don't need to be super technical myself, so I'll give you my non-technical version. The way I think about it is if you think about what vector search does, is it fundamentally takes what ordinarily would be typical search terms and encodes them in numbers or vectors. And that allows for much more powerful results around search that are context specific, and or we call it semantic search.
For example, if you think about a typical search query, if you're searching for laptops, in the absence of a Relevance Engine, in the absence of semantic search, your search query on the word laptops would actually return results that are very specific to the word laptops. If you're thinking about it in the context of semantic search, which is what vector search enables, a search query on laptops might return results on MacBooks, it might return results on ThinkPads, and things that are related to the search query, but that were not included in the search query itself. Very often, if you think about your experiences as a consumer and what you're trying to achieve with search, is you're actually having a conversation with your data.
You search on something, it returns certain results, you go down a certain path, it's what you did not expect. You come back, you go a different direction, the context-specific searches, the relevant searches, make it that much more powerful. You'll get to the results that you're looking for much faster.
Got it. That's actually super helpful.
Yeah.
Thank you for that. wanted to change the conversation just a little bit, to think about, you know, enterprise search, observability, and security, kind of the three big growth vectors that you have.
Yeah.
What's the right way to think about the growth and scale of these vectors? Which vector, in your view, has the most potential to drive upside to revenue over the next, call it six to, we'll call it 12- 18 months?
Yeah, I think, you know, all of them, because fundamentally, if you think about a data analytics platform at scale, the power of Elastic really shines when people bring multiple kinds of use cases onto Elastic. As our salespeople go out there and engage with customers, as developers adopt Elastic for different kinds of use cases, we're not focused on trying to push one solution faster than the other. We see this convergence happening quite naturally.
You know, for us today, observability is over 40% of our business. security is roughly 25% of the business, and enterprise search or search, together with the long tail of custom use cases and all the creative ways in which people use Elastic, makes up the rest of the business.
These are, you know, the annual contracts that we sign with our customers. Fundamentally, you know, that mix has not shifted much, and I see the opportunity for us to continue to grow in all of these areas. The markets are large in all of them. We're engaged deeply with customers in all of them. If I think about where Elastic really shines, it's in situations where you've got large volumes of unstructured data. If you think about logging within observability, if you think about SIEM within security or SIEM, those are areas where we will tend to land quite naturally because those play to our strengths, and then from there, we'll expand it in other ways.
You know, for example, in the world of APM, very often we will get adopted where customers want to bring business-relevant data to marry that up against the traces from the applications, and they are looking for a way in which they can do that, and Elasticsearch is really the only place that they can actually do that today.
I'm gonna ask you one more question, and then I do wanna open it up to investors out there. If you have questions, I don't know if there's a mic out there, that we can use, but I can repeat the question for the transcript, too. It's kind of the growth and profits in the light of AI tailwinds question for you. You know, you guided to 16% growth at the mid, when you look at the operating income, or I'm sorry, the operating margin target, 10%, roughly single-digit growth, right, for operating expenses.
What do you guys look at or, you know, how do you guys make sure that you're not underinvesting for maybe potential AI tailwinds over the next three to five years? You know, how do you manage that, the growth potential, you know, maybe sales capacity, R&D capacity, all that put together, where we don't come into a situation where it's like, oh, no, we're leaving product and money on the table here?
Yeah. It's a really good question. The way I think about this is across two dimensions, from a product perspective and from a go-to-market standpoint. From a product standpoint, as I mentioned, all of the investments that we've been making in AI and machine learning capabilities are things that we've been investing in for some time now. We came out with with vector search quite some time ago.
Nothing has fundamentally changed in terms of our product roadmap as a result of the recent market interest in generative AI. You know, as far as we're concerned, it's yet another data point that indicates that we were ahead of the curve, as we've been on so many different occasions in the past.
Our fundamental product priorities continue to be around the areas of serverless and some of these platform features that will enable customers to get the benefits from generative AI in the years to come. If I think about then our go-to-market capabilities, we've continued to build enterprise and commercial selling capacity for the past several quarters. We entered this year with an adequate amount of selling capacity that supports the plan that we have.
We'll continue to invest gradually in the field selling capacity that we have. When I think about our size and scale compared to the market opportunity, we've got plenty of room to grow, and we can be a much bigger company over time as we continue to invest in the business and as we continue to scale.
It's about making sure that we drive those investments profitably, and we ensure that we are getting appropriate returns on the investments that we're making before investing further in the business, and doing that with a degree of discipline. Fundamentally, our business model is a software business model. You've got tremendous room for operating leverage in this business model. As we continue to grow the top line, we'll naturally reinvest some of that back in the business and then return some of that to the bottom line as we go.
Got it. Thank you. Any questions from the audience? Please raise your hand, and we could get the questions.
Okay, I can keep going. No problem. All right. You mentioned go-to-market. I know one of the strategies that you guys have been talking about over the past year is increasing your enterprise sales capacity in your strategy. Can you talk a little bit about how that is going? How do you think about going after the enterprise, maybe from the top down? Because I know you guys have a great bottom-up motion, too.
Yeah, it's going well, I would say. Our enterprise and commercial coverage models are fairly typical in terms of what you would see in the industry for enterprise software companies. What distinguishes us is what you said there, towards the end, which is managing the bottom up and the top down. We've had a very powerful, and viral distribution model with, and we've had tremendous adoption within the developer community. The practitioners and users of the technology, we've had billions of downloads of the product. We've got massive amount of awareness, knowledge, if you think about meetups, if you think about the community, if you think about ways in which developers engage with the product and engage with our company. That has been one of the strengths for us over the years.
Looking ahead, what's important is to make sure that we continue to work with the community of users and practitioners and ensure that we make them successful, including, you know, in this brave new world of generative AI, and make sure that we are driving that motion and pairing that with a successful selling motion, as we call further up within the Enterprise and sell further up within the Enterprise.
We've been investing in both of those for some time, and so far, I would say it's actually worked out quite nicely. We've continued to, as I mentioned, build adequate Enterprise and commercial capacity for this year and looking forward to just scaling that as we go.
When you're selling from the top down, what are typically some of the personas that you're, you know, kind of going after there? Is it head of dev? Is it head of, I don't know, security? I mean, who do you sell to?
Yeah, it varies quite a bit, actually. You know, sometimes it could be a departmental-level buyer, it could be more senior executives. You know, even when we were a much smaller company, I mentioned I'm in my sixth year at Elastic. When the very 1st year that I joined Elastic, I was in a meeting with the group CIO of one of the world's largest banks.
You know, even as a small company, we punched well above our weight when people saw the relevance that we can provide to their businesses and how important we were to their businesses. I think it varies across the board, and for, by and large, we've been more successful as we've moved further up.
I think the other piece that's really been powerful for us has been our hyperscaler partnerships with all the large hyperscalers. Those partnerships have worked very nicely. They've allowed us to prosecute market opportunities that we otherwise may not have had.
You mentioned hyperscalers. Let's talk about AWS for a little bit.
Sure.
You mentioned late May, you announced a couple partnerships, expanded partnerships with AWS. You know, tell us a little bit about those partnerships. What does it mean for Elastic plus AWS? How could this channel be a good growth driver for you in the future?
Yeah, I think it builds a lot on the history that we've already had with AWS. For those of you that are already familiar with the company and the story, you know that there was a little bit of a checkered history there, and over the past 1.5 years, 2 years, we've worked to build that into a powerful partnership. What you saw announced a few days before the earnings call, was building on that partnership even further.
There were a level of co-investments that both companies agreed to in go-to-market relationships, building out greater capacity, investments on the marketing side, greater geographic access. There were partnerships on the technical side in terms of greater investments with respect to product integration.
You know, I'll point out that these partnerships are, there's not a lot of companies at AWS and some of these other hyperscalers partner with at the level at which, a partnership that we enjoy. That's been really encouraging to see as we've continued to grow our investments and our overall relationship with them. The core elements of the AWS partnership were then building further on both the product and the go-to-market side of the equation.
Okay. I wanna kind of go back to guidance.
Yeah.
Kind of your guidance framework, and, you know, Janesh, you are the CFO, so how to think about, you know, the framework? Has it changed at all? You know, when you're looking at the guidance, you said it, the inputs. I mean, walk me through the guidance methodology and any sort of changes that could have happened.
Fundamentally, Koji, our approach to guidance has not changed, right? We've for several quarters now, maintained the view that we will guide based on what we know, and we have not assumed in our guidance any fundamental shift in the external environment, up or down. We've not assumed that things will, you know, suddenly change in terms of the way the customers interact with our business.
We looked at a number of inputs from a variety of different sources, considered all of those inputs and built our guidance framework accordingly. No real change in terms of the level of philosophy, the level of prudence that we build into the guide. You've seen us deliver modest beats over the last couple of quarters.
We are fundamentally not trying to build a business where we lowball a number and then, you know, come in, meaningfully higher. Fundamentally, we just don't think that does, anyone a good service. No real change in terms of the approach that we've articulated.
Okay. Elastic Cloud, key component to growth.
Yeah.
Yeah. How should we be thinking about Elastic Cloud from here? anything we should be aware of, you know, any sort of components of the Elastic Cloud? What gets you excited most about Elastic Cloud, and maybe a little bit about the history of Elastic Cloud from, you know, where the growth has been over the past few months or actually past few quarters?
A lot of the growth in Elastic Cloud has come from a lot of the natural expansion vectors that we've enjoyed. More workloads, customers continuing to adopt Elastic Cloud for more use cases, the fundamental growth in the pace of data volumes, just good old-fashioned selling by our sales force to secure greater commitments.
You know, if I think about the self-serve motion that we have on Elastic Cloud, where people can sign up on the web, over time, that gradually increased. It used to be about 17% of revenue. For the past couple of quarters, it's been about 16% of revenue. What we're seeing fundamentally in Elastic Cloud is this motion where people have now made these large commitments.
Consumption is still taking time to ramp for all the right reasons. You know, if you think about, when we engage with a customer and we work with a customer, the customer has made a commitment to us to grow their Elastic footprint over time. At the same time, the customer is looking for ways in which they can reduce their costs in the current environment. The first thing we'll engage with them, with that customer, is helping them optimize their existing deployments, which then naturally translates into this theme of consumption optimization that's talked about.
Customers have also, in the last six months, made these commitments to us, where they know that even in the current environment, despite everything that they're aware of in the current environment, they've still made those commitments to us, and those commitments have annual minimums. The customers know that they're going to project plan and eventually migrate that data and bring more workloads onto Elastic, and we're helping them with that as well.
I do expect that that will continue to come into the numbers over the course of the coming several quarters. Fundamentally, Elastic Cloud continues to be a big growth driver for us. In terms of our selling motions, it's gotten a fair amount of engagement with customers. Our sales team is very effective at selling it now.
It's certainly top of mind for our sales force. We've even introduced a consumption element into compensation plans, for the sales force this year. Cloud will continue to be a meaningful driver of growth for us, and it should grow faster than the business overall.
Got it. You mentioned briefly optimization. As we think about our models and modeling out the business, it feels like there might be some visibility into when you guys are lapping some of the beginning of the optimization. Can you help us understand, you know, when does that happen? How do we think about optimizations as a headwind, potentially, you know, eventually tailwind for the business?
Yeah. I mean, if you think about when we started to see some of these optimizations, it was in late Q2 and entering Q3 for the most part. So if I think about just the anniversary, that's when you should start to see it. Also, as I mentioned, over the past six months, we've done really good work from a sales perspective to go secure large contracts with our customers, and these are, either annual contracts or multi-year contracts that have annual minimum commitments.
As we continue through our customer success function and our sales team to engage with those customers and translate, all of those contracts into actual usage, that will also provide a good benefit to us over the course of the year.
Got it. Last question for you, Janesh.
Yeah.
Serverless. Kind of a big deal for you guys. Walk me through what the serverless option is or product or, I guess, educate me a little bit more on. on serverless. What does it mean for the business? How could it be a growth driver for you over the next 3-5 years?
Yes. You know, a good analogy to think about in the context of serverless would be, the Lambda offering from AWS. How that compared to AWS's traditional, EC2 offerings? You know, when I think about what serverless does, it's a few things. Number 1, it unlocks additional use cases for us. If customers have really bursty, capacity and bursty, data that they want to then leverage Elasticsearch for in terms of data analytics, it allows them to do that much more effectively through the serverless offering.
The second thing it does is it actually allows customers to get operationally a lot, manage their Elasticsearch deployments a lot more easily from an operational perspective, 'cause we take on an even greater burden of, a lot of the orchestration capabilities. You know, if you think about it, today, Elastic is more of a platform offering, it's a PaaS-like experience, this takes us more towards a SaaS-like offering that from an ease-of-use perspective.
The final piece is that from the standpoint of something that's near and dear to me, from the standpoint of our overall value realization and margins, it unlocks more possibilities there that allows us to be more efficient at the back end as well, also allows us to then see that translated into better margins on the cloud side over time.
Got it.
Yeah.
Janesh, we're all out of time. Thank you very, very much for doing this. Super, super appreciate it. That was an awesome conversation.
Likewise.
Thank you, sir.
Thank you again for hosting us.
Yep.
Much appreciated.
Yep.
Thank you.