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

Hello, everyone. Welcome to Elastic's Inaugural Financial Analyst Meeting. Thank you for joining us today. My name is Anthony Luscri, and I'm the Vice President of Investor Relations at Elastic. Over the next few hours, we plan to share details on Elastic's business strategy, market opportunity, product vision and continued investments to drive future growth. Before we get started, I'd like to take care of a few housekeeping items. First, let me share our safe harbor statement with you. I won't read through it in detail, but to summarize, we will be making forward-looking statements today that reflect our views as of October 14, 2020, and may include predictions, estimates or expectations regarding the impact of the COVID-19 pandemic and other information. These forward-looking statements are based on factors currently known to us, speak only as of the date of this call and are subject to risks and uncertainties that could cause actual results to differ materially. We disclaim any obligation to update or revise these forward-looking statements. Please refer to the risks and uncertainties included in the slide deck and those more fully described in our filings with the Securities and Exchange Commission. As a reminder, we will also discuss certain non-GAAP financial measures. Disclosures regarding these non-GAAP measures, including reconciliations with the most comparable GAAP measures, can be found in the appendix of the slide deck. In this meeting, we will not be commenting on our fiscal Q2 performance, which ends on October 31 or on our fiscal 2021 guidance. This program will be recorded, and both the recording and the slides will be available to view online after today's presentation on our Investor Relations website at ir.elastic.co. We have a great lineup of topics and speakers for you today. Our program will be roughly 2.5 hours in duration. Our Founder and CEO, Shay Banon, will kick it off by covering the business, our competitive advantages and product vision. Next up, Justin Hoffman, SVP of Sales, will cover our go to market and will host a customer fireside chat with the CIO at Ancestry. We'll then have a 10-minute break to give you a chance to stretch your legs. Finally, our CFO, Janesh Moorjani, will provide you with a review of our investments to drive future growth as well as a broader financial review. Finally, we'll wrap up with a 45-minute moderated Q&A session. Shay, Janesh, Justin, as well as Paul Appleby, President of Worldwide Field Operations, will be available to answer your questions. We will be using the built-in Q&A feature. So if you have questions, please submit them there any time throughout the presentation, and I will moderate the Q&A. So with all of that out of the way, let me go ahead and hand it over to our Founder and CEO, Shay Banon. Shay?

Thanks, Anthony. Hi, everyone. Welcome to our first financial analyst meeting. It's been 2 years since we rang the bell on the New York Stock Exchange. Over this time, it's been great getting to know many of you through our earnings calls and investor meetings. I wish that we could be together in person, and I hope that you are all staying healthy and safe during these times. If Elastic is new to you, I am excited about the opportunity for you to learn about our company. This week is also our Virtual ElasticON Conference. We're hosting more than 20,000 attendees from all across the world. We will have 100-plus sessions, bringing together our community, customers, partners and employees in a distributed way. The same distributed way we have been building Elastic from the start. We like to say at Elastic, we are distributed by design. Since the beginning, we've always embraced a distributed company culture. And today, we have more than 2,000 employees who live and work across more than 35 countries. So despite the challenges that many organizations are having when moving to a remote working environment, at Elastic, working remote, virtual and distributed has always been a first principle. And as such, I know that all of our teams from engineering to sales to marketing, finance, legal and HR, have not lost a beat and have continued to create innovations, which today, we're excited to share with you. I want to start briefly talking about a systematic and repetitive framework for making fast and accurate decisions that I learned about this summer. It's called the OODA loop. It was developed by John Boyd, a fighter pilot and one of the 20th century most influential strategist and theorist. It has a lot of applicability to what we do at Elastic. The loop starts with observe, the collection of data using sensors. The loop goes to orient, the analysis of data to form a perspective. Decision is the next phase in the loop, the determination of a course of action based on set perspective. And the final step is act, the physical playing out of decisions. And it is a continuous feedback loop. By orienting, deciding and acting the environment changes, and it must be taken into account. I really like this concept. Our singular goal at Elastic is to build products that give our users a continuous competitive advantage to provide more value, both within the organization and outside of it. In order to provide this value, organizations must be able to collect data, find ways to generate insights out of it and then act on these insights with decisions that lead to outcomes. And importantly, this loop must be done continuously in real time, and it needs to be placed at the hands of as many people within the organization to maximize its impact. When thinking about this loop, the best organizations are ones that not only form this group but can go through it fast. Speed is the most important factor in the loop. The quicker one is able to process data, gather insights and act, the better the outcomes are. At Elastic, speed is part of our DNA. When I first built Elasticsearch as a search engine, I had to go over to enable users to get insights from their data within a blink of an eye in milliseconds or less. Still today, speed is a core foundational value across all of our products. Organizations should not be limited by scale. The ability to collect as much data as needed while still being able to drive insights and action out of it is critical. We've always shared in the belief that users should have infinite scale. They should not have to throw their data away or move their data from one system to another. And together with speed, scale provides the ability to analyze petabytes of data with the potential to drive better outcomes. But let's say, you have all the data you can dream of. Not all of it is created equal. There is such a thing as noisy data. This is where relevance comes in. Relevance ensures that you can generate the best possible insights from the data and the most relevant actions. It is the sum of speed, scale and relevance, together, not individually, that drives value for our users. It ensures that decisions are based on using all possible data, generating relevant insights with speed and using those insights to lead an organization to the best outcomes and actions continuously. This is why we've built Elastic. We are a search company. We'll bring the power of search with speed, scale and relevance across all of your data to enable insights that can be acted upon, fast. For us, it all starts with a search box. Everything we do starts with a prism of search. It is the most natural feeling in the world. When you look at a search box, it is simple and inviting. It hides behind it an ocean of data to observe. It is so natural to just go and start typing, to orient. The wealth of information is curated to a decision when a query is entered and relevant results appear fast and the best action can be taken in the context of these results. As searches happen and actions are taken, we learn. We use that to feed the next cycle of search. And we can go through this loop fast. Our goal is for us to take this search experience and give it to everybody to solve any use case. And over the years, we found that search is much more than a search box. It is an experience. Search helps organizations be more productive and happy by finding insights across their enterprise apps and content repositories to make better decisions faster. Search helps organization serve their customers better by finding insights across their infrastructure and applications, making sure they are up and running. Search helps organizations be protected by finding insights across their various systems, from network to cloud to endpoints and block attacks. In our ElasticON keynote, we spoke about our various powers that we can bring to organizations. I would love to expand on it and talk about them more here. Everything begins with our developer community. We have millions of developers who have adopted our products. Our community is strong and has people who love our products. They use them at work. They use them for personal projects. They use them to solve problems and deliver value. They recommend them to their friends and teammates. Developer mindshare is special. It is not often that a company manages to capture it at such scale. We do not take it for granted. We feel it is something we earn every single day. It drives grassroot adoption within enterprises. Nothing is ever forced. A developer simply takes our product and uses them to realize value. This value then leads to new champions within a team or a department and eventually to widespread usage across the organization. Our community also provides continuous feedback on our products and help us stay ahead of the curve in terms of market trends. As an example, our most common use cases have become our solutions, and we were well ahead of the curve in identifying the convergence of logs, metrics and APM and the trend of observability. We believe that innovation starts with a single technology stack, the Elastic Stack. The power of a single stack has many force multipliers. It can be used across many use cases. It has economics of scale. By making our stack faster, more scalable, more relevant, it benefits all users and use cases. For organizations, by having their users learn to use 1 tool, they can focus on solving their business problems and deliver value better and faster. For us, a single stack is important from an engineering perspective. It gives us a unified foundation for all of our solutions, accelerating our ability to deliver value to our users. We announced several such upcoming innovations during the conference. I wanted to use this time to show how impactful this single stack concept is through them. Elastic Agent provides a single unified place to collect data with one click. It spans data collection across both observability and security. And we are taking it a step further and are integrating endpoints protection directly into it. While you observe, why not protect. Kibana Lens is a powerful, intuitive drag-and-drop visualization insight tool. It gives a security analyst a new way to do threat hunting or for a DevOps team to conduct ad hoc analysis. Actionable Kibana transforms from being a place to visualize and drive insights out of your data to a place where one can also act upon it. It allows a threat hunter to directly react to an attack or a DevOps team to collaborate during a production incident. Searchable snapshots allow us to store data directly and seamlessly on S3 or similar low-cost storage options. It balances speed with cost. It transforms the way in which users can investigate security threat, search across nearly limitless data sets or provide trend analysis over extended periods of time, all in a cost-effective manner. Schema on read provides the ability to search on any data immediately while transitioning with one click to the best performance when ready with our famous schema on write. And we have taken this stack and all of these innovations around it and have specialized it in 3 solutions. Our solutions are built on top of actually into this single Elastic Stack. This is powerful. We've doubled down on simplicity through specialization. Take a search experience, a search box and put it on a website, application or workplace. And we have Elastic Enterprise Search. Take a search box and put it on your infrastructure to observe it, and we have Elastic Observability. And take this search box and put it on your company to protect it, and we have Elastic Security. The power of solutions means we can focus on creating simple, contact specific and best-in-class experiences within each solution. The magical 3-second experiences by specializing in the most common usages of the stack, we simplify the most expected experiences of it and become the best in each one. All of that, while delivering a unified experience across all of our solutions, to benefit our users. Let's talk about our first solution, Enterprise Search. Our vision for Enterprise Search is to let users search everything anywhere. It brings search to websites, applications and the workplace. We believe that Elastic Enterprise Search is critical for every business, especially as businesses move to become online and virtual. Search brings many benefits to this movement. It clarifies the user experience that drives clicks, conversions and checkout processes. It facilitates new monetization models. Think of a dating app, swipe left, swipe right. What about search placing based on demand. It drives employee productivity and happiness, letting workers find what they need to do their jobs in today's virtual and distributed workplace. Our evolution in Enterprise Search goes all the way to our beginnings. We have been doing this for years. Enterprise Search was the very first use case when I released Elasticsearch. Search a website, search within an app. Over the years, we believe we'll become the de facto search engine used on most websites and within apps today. And chances are that if you are on your favorite e-commerce site or mobile app, the search, filters and recommendations are powered by Elastic. In the early days to build the search experience, it was largely do it yourself. And developers would use our APIs to build these search experiences. That all changed in 2018 when we joined forces with Swiftype, a company who had built turnkey SaaS enterprise search solutions on top of Elasticsearch. Since then, we've been working hard to fold Swiftype's product, App Search, into our Enterprise Search solution and our Stack, delivering it both on-premises and in the cloud, focusing on ease of use and simple 3-second experiences to add a search box to any application, one that scales. We also launched Workplace Search recently. Workplace Search is a new innovation, and there is lots of excitement around it, especially given our current work environment. Today, IT teams must equip their workplace to work from home and workers likely have 10 to 20 different apps that they use. A single Workplace Search simplifies that experience with a single unified search across all of the apps an employee uses. It has 12-plus out-of-the-box connectors to popular cloud applications with more coming soon. And we know that organizations have legacy content sources. So we've built a custom API to connect to them to. I am most excited about the opportunity to bring a simple consumer-like experience to this market. It is much need of reinvention. Our second solution is Observability. As organizations become more and more digital, online and distributed, they have more and more infrastructure and applications to observe. Our vision is to give organizations the ability to have unified observability across all of their digital ecosystem. So they can focus on constant OVI zone and seamless delivery of value to their customers. We believe that good observability is critical for the availability of organizations today. Organizations must keep their systems up and running around the clock. And with lots of different systems and applications to monitor, come lots of dependencies, lot of force positive and, yes, some finger pointing. The criticality of downtime is lost revenues, unhappy customers and a loss of productivity. Search brings many benefits to these issues. It enables IT and DevOps to correlate logs, metrics and APM trace data. These are critical data streams to keep systems up and performing. By having this data in a single place, the nightmare of stitching data together from multiple sources magically disappears. It streamlines the process of monitoring the health of all the systems and applications. So this leads to improving IT efficiencies and ensuring teams can reduce MTTR by fixing issues fast or before they even happen. It provides actionable insights to meet business SLAs, such as is my system up or down? Why is there a spike in slow performance? Why is there abandonment on our checkout? For many years, we've been the most popular open source logging product out there. It's our users who trusted us with -- for their log data, then metrics, then APM. It is without any curated experiences, do it yourself. They did this because of foundational elements of the Elastic Stack, powerful open APIs, flexible visualizations with speed, scale and relevance. And we have invested heavily in this space, both organically and through acquisitions. We joined forces with Prelert to bring machine learning to all of our users before AIOps was a thing. We joined forces with Opbeat to bring APM to all of our logging users before observability was a term. We worked hard to natively integrate it into our single technology stack because we believe that when conversion happens, products become features. We launched a unified observability solution to double down on this vision, focusing on ease of use and turnkey components, including a common schema, a common UI framework, a 100-plus prebuilt integrations, out-of-the-box monitoring for Kubernetes or containers and many rich APM capabilities from APM agents for Java and .NET, distributed tracing and service mapping. And we plan to continue our investments in providing an end-to-end observability solution. This includes mobile APM, synthetic monitoring and user experience monitoring. We also believe there is significant value to provide all of our logging users' access to the full set of observability features, including APM. And we also believe that through our single unified stack, we can bring observability to our other solutions. A search box in an application can easily instrument it. APM traces can provide significant value to a threat hunter as an overlay to security logs. After all, every ops log is also a security event. And this leads me to our third solution, Security. As organizations become more and more digital, online and distributed, the urgency and need to protect them become critical. Our vision is to protect organizations' data from attack, keeping them safe, online and available in the face of adversity. It has been humbling to hear about how our technology is used to prevent, detect and protect organizations from attack. Stopping attacks is no easy feat, especially at scale. And worse of all, if an attack is not discovered and stopped, if there is a bridge and leak, it can lead that organization down as debt spiral with lost revenues and damaged reputation and compromise data. Search brings many benefits to security. It facilitates real-time detection and protection from endpoints to the data center. It's no longer about securing a single system, but many holistically. It allows for real-time holistic visibility for the entire SecOps team. Think about a SOC with a bunch of big monitors and screens. With Search, SOC analysts and threat hunters can monitor and analyze data in real time. It enables insight and action. It allows for insight to be put in an operational state, detect something, block it. It reduces dwell times to minimize or avoid damage before something bad happens. There is a need for new best-of-breed enterprise-grade protection system that combines SIEM and Endpoint Security into a unified experience. This is what we are building. Like with Enterprise Search and Observability, our users have been using Elastic for security analytics for some time. There are big banks who have built their cybersecurity platforms on Elastic. There are modern tech companies who declined to buy a legacy SIEM and used Elastic as they see the future. In such cases, like with our early Enterprise Search and Observability users, it was do it yourself. They did this, thanks to foundational elements of the Elastic Stack, powerful open APIs, flexible visualizations, with speed, scale and relevance. Over the last couple of years, we made a few big bets in the security space. First, we started to build our own SIEM directly into the Elastic Stack, focusing on a curated user experience. Second, a year ago, we joined forces with Endgame, an endpoint protection company. Endgame was built out of providing an endpoint security platform for the military, enabling protections, both on the water in submarines, took a phase above ground. They've built one of the best endpoint technologies and won recognition from Gartner and other security analysts. Today, with Elastic Security launched, we have a SIEM that is GA, and we released the first milestone of integrating endpoint protection into the Elastic Stack. One of the other aspects that happened when Endgame joined forces with us is that it brought a 100-plus security experts into Elastic engineering. This has radically increased our ability to move fast in building lots of new capabilities in security. Over the past few months, we've released 200-plus open protections, a 100-plus integrations, automated workflow and alerting capabilities and the ability to do one click malware protection. Our road map includes features such as advanced correlations, threat intelligence capabilities and integration and ability to do host isolation and building new antivirus certifications from Windows and macOS. We're also working hard for full integration of endpoint protections into Elastic Security and our Stack. We're really excited about our future in security. It's new. It's fresh. It is early. It is what we think the industry needs, giving the industry a fresh look at how security systems should be built. Bottom-up security practitioner first and open. There is so much power in being open. It means our products are developed in the open. It means that we build the connection with our users. It means that we form bonds to organizations that use us that extends just mere usage of our products. We use it to react to our organization's needs both today and tomorrow, with speed. The power of open drives a faster feedback loop reflected in our products. It allows for us to collaborate with all of our customers and partners. It brings transparency, faster development and innovation. I'd like to share a bit about our history, including how we think about open in the context of our company. When we started the company, all of our products and features were open source. We licensed our products under the permissive Apache 2 license. We were and still are the sole committer of our code base. In essence, we were the benevolent dictator of our road map. We controlled it and we still do it today. This has allowed us to move fast, build a lot of features and get rapid adoption of our products, known to many as ELK, or Elasticsearch, Logstash and Kibana. Between the years 2014 through 2018, we began to release our first set of proprietary commercial features. We initially released monitoring, security and alerting. We then added machine learning, reporting and graph analytics. And with our 5.0 release, we decided to fold all of these proprietary features into a single plug-in, we call X-Pack. It was sold under our paid subscription license. In 2018, we doubled down on both open and proprietary. We open the code of X-Pack and we merge its code into one single code base. These changes were made to give the users visibility into our proprietary code, giving them the same benefits they have with our open source products and enable our engineering team to accelerate innovation, building new features fast. It also simplified the conversion process for customers, eliminating the need to install any additional software. We created a new tier, Basic, which is open and free, yet proprietary. It is only available on Elastic Cloud for as-a-service usage. And we made our default distribution include all of our features with the free features owned by default. We have changed our focus to develop most new innovations into our new proprietary tier. Today, our strategy is paying off. More than 90% of our downloads are the default distribution. This means that our Basic features and solutions are being adopted by on-premises users. And when they decide to move to the cloud, they are likely going to want to stick with us. And our investments created a significant value to our users in both our Basic and Gold+ tiers. Just with our recently announced innovations, you can see that Elastic Agent, Kibana Lens and schema on read are Basic+, searchable snapshots is under our enterprise tier. And on cloud, everything is paid. And the power of cloud and hybrid is about being there for our users wherever they are. It is about powering our users with optionality, and it is about powering our users with ease of use and peace of mind. Over the last 2 years, we've made significant investments in this space. Cloud is our fastest-growing business segment. We've made focus -- we've been focused on ease of use, building a turnkey SaaS product and making it as easy as possible for users to onboard new data across our solutions and drive value, insight and action out of it. We've increased our cloud regions 3x over these 2 years. We now have 41 regions across the world within AWS, GCP and Azure as well as native billing and marketplace integration, and we continue to invest in more regions and more native experiences across all cloud providers. Organizations need to deal with the reality of data gravity. Our customers can deploy next to or in the same location as the system they need to search, observe or protect. Even more, we support seamless ability to serve to cross locations using the same unified APIs or UI. Our effort to be where customers are also allows us to support data privacy and compliance regulations, an aspect that is becoming more and more prominent in today's organizations' needs. We prefer to have our customers run on cloud. It allows us to provide a simpler and streamlined user experience and allows our customers to focus even more on their business needs versus having to manage infrastructure. And on the cloud, our customers pay only for the data they actually use. We have a single pricing approach across all our use cases and solutions. This is powerful. It gives our customers peace of mind and predictability. It means they can move fast, easily consume more and more data and easily deploy us across more and more use cases and solutions. They pay only for the data they actually use, and they love it. It means we provide clear and predictable pricing, not only for existing usage, but future ones as well. We have had the same pricing model since we started. We believe the power is in the data. We believe the value is being able to search it and drive insights and action out of it. It is important to us that our pricing model aligns with our customer value. And it is humbling to witness the amazing adoption of our products. It is reflected in our customers. We have built products from our common Elastic Stack to our various solutions that apply to all organizations, regardless of their size, industry or vertical. Two years ago, we estimated our TAM at $45 billion. Today, we estimate it to be $78 billion by 2021. It is based on secular tailwinds within data, our 3 solutions and our expansion into new markets. As I mentioned, we have focused on building significant innovations at our core stack level capabilities, innovations that apply to many different use cases that are powered by data and the ability to search it. And we made significant investments and progress in our Enterprise Search solution and our new Observability and Security solutions. And it all starts with the power of search. Search is a part of everything we do. We will always be searching for something. It is how we start our day at home. It is a part of our everyday work. It is how we seek answers to problems. It is what gives us insights to make decisions. It is a constant and timeless experience, and it all starts with a search box. The magic that happens from it, the smile, the sense of satisfaction, the wonder, all made possible because you were able to find something. Justin, handing it over to you to cover our go-to market. Justin?

Hello, everyone, and thanks for the opportunity, Shay, to share my thoughts on Elastic's go-to-market and the incredible opportunity we have in front of us. My name is Justin Hoffman, and I have the pleasure to lead Sales at Elastic. I joined Elastic in April of 2012 after having spent over a decade at VMware, helping grow their business from $30 million in 2002 to many billions. I saw something very similar to Elastic, passionate and motivated founders focused on solving hard problems, a disruptive technology and a viral distribution model that could immediately demonstrate value. We focused early on building trust, asking questions, listening. We understood that long-term sustainable growth was achievable with the strong foundation of supportive and diverse practitioners across many communities. We endeavored to find new ways to access these communities to help them leverage speed, scale and relevance in solving their hardest data problems. Now I know firsthand the value of winning the hearts and minds of doers. And in turn, earning the right to engage with budget holders, decision makers and executives. As a result, our customer partnerships are built on a foundation of trust and value, not spend or confusion. Let me start off today with why I think we win. As data volumes grow in size and complexity, Elastic customers are looking for a way to unite a sea of disparate data sources in stores. In a world where seconds matter, Elastic helps our customers find answers to drive action, speed to market and improved engagement. Search enables customers to iterate at the speed of thought to understand what data is valuable and provide a perspective into data that aligns to business requirements. Elastic customers seek simplicity, that is versatile enough to address the most complex data challenges. Unifying data through search provides massive economics of scale by leveraging a single stack to solve a host of challenges. While most of our customers start with a single-use case, they quickly springboard into 2, then 3. While Elastic deployment drives value, the power play is when you stop thinking in silos and start navigating across all of your data and then look at all that data through whatever lens you feel appropriate to do your job. Providing customers a single transparent pricing model lets them control what data they keep, how they store it and the speed in which they access it. We believe customers should have the power to decide, not be forced down a never-ending path where cost vary based on ingestion or are further complicated by a vendor predetermining the value of one data source versus another. Now every quarter, we close thousands of transactions and add hundreds of new customers. We continue to deliver more efficient ways to trial and purchase Elastic, helping further improve our overall cost of sale. We continued our rapid pace of customer acquisition growing from more than 8,100 total customers in FY '19 to more than 11,300 in FY '20. In our last quarter, we reported a total more than 12,100 customers. We've also seen customers increase their spend with us. In fiscal year '20, more than 610 customers had annual contract, or ACV, value greater than $100,000, up from 440 the prior year. We've also had customers who spend more than $1 million in ACV increase to more than 50, up from 30 the prior year. This continual growth of 6- and 7-figure customers is both a vote of confidence and a strong indicator that our strategic importance to their business continues to grow. We are seeing great success with customers, having now penetrated 46% of the Fortune 500 and 31% of the Forbes 2000. We have experienced a high adoption rate by many of our biggest customers across major industry sectors. While I'm excited to see the high penetration across verticals and geographies, I'm also thankful for all the great work our teams do to execute long-term contract vehicles. Each of these customers represent expansion opportunities for all of our solutions. So here's a snapshot of some of our many customers by these industry sectors. Many of these customers have been with us for a long time. Starting with a single-use case, born from an individual developer. Fast forward to today, many of these customers have dozens of business challenges where Elastic has become the de facto solution. These organizations are starting to consolidate their Elastic projects in an effort to centralize their expertise, streamline their operations and supercharge the value to the business. As I mentioned before, the customer journey is bottom up. We have millions of developers who use Elastic, all around the world. We have hundreds and thousands of developers who participate in our community events like meetups, webinars and conferences. They're eager to learn, eager to get started and eager to share what they've accomplished with Elastic. This word of mouth creates high volume, frictionless user engagement model focused on letting users get value from our technology first either through a cloud trial or download. As users engage with us and begin to adopt our technology, they find interest in our proprietary features as they are immediately exposed to all of Elastic's capabilities. It is a daily occurrence where we identify users who have a working prototype, have deployed our technology in production and have self-selected into engaging with us. This leads to a warm touch, a value-focused discussion with an informed user. As such, we're able to advance through many steps in the classic technology sales cycle. Our sales development, solutions architects, sales, customer success and partners get involved to listen, educate and onboard. We work hard to identify opportunities to deliver the greatest business value and ultimately earn their business. We then seek to identify new data sources to enrich or complement the original use case. We empower these customers to share their story with other departments to rinse and repeat. We leverage these relationships to introduce new solutions. We believe this enables a continuous cycle of success bringing success. Because of our cyclical relationship with developers and practitioners, our bottom-up model has provided us the opportunity to bring a message of value to executives across countless functional areas of the business. There's no better feeling than to have an executive tell you that they were excited to have a conversation as so many of their team cannot stop talking about Elastic. It's so validating to have a developer introduce you to their leadership for no other reason other than to see Elastic be successful and then speak up in that meeting and wholeheartedly support your message. By investing in our technical and executive relationships, we're being brought into conversations now with their line of business owners. By more deeply understanding the needs of the business and the associated cost benefit, we're successfully architecting Elastic into the technical solution. Now our go-to-market structure is mapped to the customer journey. This ensures that we engage and help our users adopt our technology in a frictionless way as well as onboard and expand with them over time. We've anchored on a cloud-first strategy, where we are intentionally showing the cloud as the best way to on-ramp into Elastic and our solutions. No need to find a server, no need to click to deploy and start a new solution and no need to find operational resources to manage it. Within the field sales organization, we're focused on curating and converting the demand created by marketing into new customers and expansion deals, both cloud and self-managed. Our direct sales organization has 3 segments: commercial, enterprise and public sector. We are not building and do not intend to build a specialist sales team. We have sales reps who are enabled to talk about the value of search across all 3 of our solutions. We have solutions architects who are deep in technical aspects of our stack and our solutions. We have invested in specialist capacity within our solutions architecture team. While they are experts in maybe areas like security, they are also very capable in observability and enterprise search. Again, the power play is when customers understand that search can unify across many data and functional silos. Our customers appreciate that their account team is knowledgeable about how their organization or others like them are using Elastic to solve similar and/or different problems. One such accelerating trend is that operators and SecOps are coming together to leverage all of their respective data to collaborate, troubleshoot and solve issues. We are finding new ways to drive efficiency into our go-to-market model. Our inside sales team drives the long tail of our transactional business, much of which these days is on Elastic Cloud, continually finding more efficient ways to engage with our community to drive the velocity business that is so critical to our present and future growth objectives. We also have a renewals team. This team allows our direct sales reps to focus on new business acquisition as well as larger renewals that have expansion opportunity. I'd like to briefly touch upon services. Our services work has been mostly focused on expert services, architectural best practices and migrations. We have a small team that can do this work, but largely, we are focused on building out regional and global consulting partners to help in this area. Geographically, each of our go-to-market segments are organized by regions across the Americas, Europe and Asia Pacific. This has allowed us to have a coverage model that aligns to our demand and to make the right investments to drive top line growth across our existing customers and spectacular community of users. 43% of our revenue in fiscal '20 came from outside of the United States. We will continue to invest in key global markets to maintain this competitive advantage. The U.S. public sector business, as many of you know, has always been a strength for us, and we have taken that strength and have deployed reps in key public sector markets, like the U.K., Germany, Australia, Singapore and Japan, as well as growing our state, local and education business here in the U.S. Now you've heard a lot about our users and our customer journey today, but I want to make sure I shared my thoughts about how we monetize our amazing distribution model. The sales team remains focused on delivering a strong reoccurring revenue stream via our subscription model. Our ability to execute a 90-plus percent subscription mix of our business has and will continue to be focused on the value of our proprietary features, our new and expanding cloud business and world-class support. While every customer has a different path to monetization, all customers start with converting from either cloud trials or the use of our free and open stack. We commonly convert these users to customers based on many factors, but the overwhelming demand is driven by the need to have vendor support for production systems. The mission criticality of these services are applications. The need for advanced proprietary capabilities. The need to orchestrate and operate an expanding footprint of Elastic instances or the customers' desire to have experts deliver a service that they can consume with far less infrastructure or management overhead. Once we establish the customers' preferred consumption model, cloud or self-managed and the driver for our commercial relationship, as noted before, we can map them to the appropriate subscription level, gold, platinum or enterprise. Each subscription level is differentiated by a support SLA and the packaging of various proprietary commercial features at the core of all of our subscription levels is the free and open version of Elasticsearch. While we pride ourselves on delivering choice to our users, there are still reasons why data may need to remain inside a customer's data center. However, we expect their long-term strategy will be to consume cloud services and optimize their resources to engage with their data rather than manage it. While we do support our customers at any stage of their cloud journey, my team leads with cloud. We fundamentally believe that Elastic Cloud provides our users and customers the best possible experience across all solutions. Customers who deploy on Elastic Cloud find the getting started experience to be intuitive. We provide the depth of control that users would expect of any market-leading PaaS while having the flexibility and intuitiveness to just pick a solution and get started. As business requirements change, though, we can optimize an environment. For speed, you have the flexibility to drive down overall TCO. This shift in cloud focus is evident in the increasing percentage of SaaS revenue in our overall business. Customers are increasingly choosing cloud due to the flexibility it provides. Our continued investment in cloud across all functions at Elastic has continued to lift our SaaS percentage of total revenue from 17% in FY '19 to 22% in FY '20. We're seeing a few key trends emerge with cloud, data residency and localization matter. This is why we continue to invest in strategic partnerships, expansions into new regions and markets. Customers are expanding with us faster in the cloud when we unlock the constraints of traditional IT. And increasingly, customers are choosing more than one infrastructure provider and need a cloud solution that can operate agnostic from one particular provider. With our cloud-first strategy, we expect that Elastic Cloud will continue to be the fastest growth segment in our business. Now continuing on the theme of choice, we are where our customers want to be. We are thrilled to have an integrated marketplace experience across all of our cloud partnerships. Elastic sales partners with their peers and all of the major providers. Elastic customers are able to leverage pre-committed spend with their infrastructure provider to deploy on Elastic Cloud. They can deploy closer to their source data or at the edge across any number of our 41 global regions. Now reference stories at Elastic are critical to our go-to-market. With any technology that is so broadly applicable to a variety of use cases, our sales teams must be able to help our users and customers understand the art of the possible. One of my favorite stories is that of a major U.S. financial services company. This customer started their journey with Elastic in 2017. They had a multi-tenanted application that served their wholesale clients. They needed a more effective way to search countless of transactional records fast and reliably, but do it in a secured way. Of course, they need a client isolation, but for regulatory reasons, they also needed to provide role-based access control for each client. While open source Elasticsearch with a rich set of APIs provided the necessary enterprise search capabilities they required for this use case, the access control became the commercial driver. After a few short months, this single-use case received industry recognition and became a catalyst for the relationship. From there, the same application owner was looking for an alternative way to monitor the health, uptime and customer experience. They have autonomy to act independently from the corporate logging standard, which at the time was Splunk. But their assessment was that there was minimal difference between log lines and transactional records. The team had acquired such significant knowledge around Elasticsearch, they decided to make Elastic their observability standard for logging. An unexpected customer outcome was the discovery that searched across transactional data and operational data combined gave them new insights to improve customer experience, reduce time to resolution and drive engagement. Now on the heels of these 2 projects and through word of mouth, new use cases for Elastic began to surface. New teams across data science and operations jumped in. By 2018, Elastic had become so wildly popular that a centralized platform team decided to elevate Elastic as a standard offering, where teams could acquire Elastic and all of our commercial capabilities as a centralized service. That service exploded as the classic challenge of pain for ingest with incumbent vendors became an obvious driver. They saw continued investment, but they came to face skill challenges of managing single-tenant environments in an operationally effective way. It was around that time that we launched our Enterprise subscription tier. Our Enterprise subscription provided centralized orchestration to drive efficiencies. They became so efficient that the team decided to open up their offering to new data sets beyond logging and launched their first Enterprise Search service. This centralized team as well as so many other independent teams have become champions across the enterprise, hosting internal community events as well as hackathons. We have only scratched the surface here, and I believe there is so much more in the coming years. Now we jump across the pond to a major European telecommunications provider. It's no secret that telecom produces large volumes of data with strict regulatory requirements. Our earliest commercial engagements with this customer revolved around observability, ingesting log and metrics data from across their networks. Observability expanded into customer-facing mobile applications where IT and DevOps were able to monitor the health of applications in real-time across millions of their customers to ensure optimal user experience. Their own analysis had shown that improved customer experience could directly be correlated to improved satisfaction and ultimately higher retention. While IT and ops were driving customer retention with Elastic, marketing was building a rich customer analytics system through Enterprise Search to deliver personalized content to their customers, all while the security team was preventing fraud with Elastic. This fraud use case became a springboard into doing security monitoring for their launch of a new home television streaming service just earlier this year. Now if I move on to our next customer, we've learned about Elastic's adoption at this customer, which is a large U.S. retailer, through their purchases of training. A single developer was so excited coming out of training that they immediately began prototyping their first enterprise level deployment of Elastic for observability. This was a new business-critical system that had no health or uptime monitoring. This project was broadly considered a success and became a jumping off point. After several observability projects later for in-store retail applications, a new project was initiated to replace the legacy SIEM. The leaders of this project were previous Elastic users and champions who had joined the retailer from outside the industry. They wanted to take a fresh look at how security practitioners could engage with data in real-time because they were generally unhappy with all the traditional SIEM offerings due to price, performance or effectiveness. After a short POC, the internal team was able to show query response time across billions of records in seconds versus 30 minutes with alternatives. Analysts can now traverse large volumes of data, iterate on this data and close gaps before they became exploits. So in conclusion, I'm just so proud of our team, our community and our customers, who have all played a critical role in our success over the last 7 years. I am more excited today than ever before about our future. With that, I'm pleased to introduce Sriram Thiagarajan, CIO and SVP of Platforms at Ancestry, a long-time customer of Elastic. Sriram, are you there?

Yes, I am. Hi, Justin.

Hey, Sriram. Good to see you again.

Nice to be here.

Just, first of all, thank you. Thanks for all -- from all of Elastic. Thanks for taking the time today to share your perspective on your journey with Elastic. So welcome. Welcome to the show. So maybe we just start off here just briefly. I think you just describe for everyone on the line your role and responsibilities at Ancestry.

Yes. Thank you, and glad to be here. I'm the CIO and Senior Vice President of Platforms for Ancestry.com. I've been with the company for over 3 years. I'm leading our -- both our corporate infrastructure as well as our product platform and production infrastructure and some of the engineering functions as it relates to our cloud management platforms as well as some of the business systems and operational support systems that are used to enable our business. So excited to be here, and we've been a customer of Elastic for a long time.

That's great. Awesome. Well, maybe if you could just take a few minutes and tell us a little bit about the Ancestry and the journey that you've been on since joining the company?

Yes. Yes. Ancestry is a human-centered company. Our mission is to power journeys of personal discoveries for people to enrich their lives. The company has been around for over 30 years, and we have the largest collection of historical records from 80 different countries of origin. We are the largest DNA network, about 18 million people have gone through our DNA testing. We have 3 million plus paying subscribers on our family history business. And data is the core of our product experience. The company had spent over $300 million in all these 25-plus years of its existence to acquire content, be it historical records, be it DNA samples, et cetera. And that's what we used to offer personalized product experience as to our consumers. So it has been a great journey for the company. It started off as, believe it or not, a book publishing company on genealogy. And as technology evolved, the company itself had gone through a series of transformations establishing its presence on the Internet through Ancestry.com, then shipping CD-ROM -- software and CD-ROMs. And then later, as Internet evolved, we have a full-fledged application now we offer on our website. And the last 3 years has been an exciting journey for us in our digital transformation. We are now fully 100% on public cloud. Zero footprint in our on-prem data centers. And that's been a very exciting journey for the company. It has always transformed itself through the course of these 30 years of its existence, so.

That's awesome. How long -- just out of curiosity, how long have you been on that cloud transformation journey?

We are in the third year of the cloud transformation journey. We started in 2017 and I mean it was a big decision for the company at that time. It put all its efforts behind the cloud transformation, great buying across the executive team. And we decided to lift and shift all of our services to the public cloud. And then that was a conscious decision to do a lift and shift. And then over the past 1.5 years, we've been optimizing both in terms of cost as well as infrastructure and software. And right now, we're in the final throes of our optimization in the cloud, specifically around DevOps processes and tools, and infrastructure optimization.

And I'm curious how that's maybe changed your decision process over the years, introduction of cloud, new way of doing business, has it changed? Or -- and if so, how?

Yes, definitely. First of all, it is a -- in addition to a technology transformation, we pretty quickly realized it's also a cultural transformation, the way we operate within the product and engineering organization. Along with the cloud transformation, we also standardized agile processes at scale. Although our teams were doing agile before, it wasn't orchestrated in a standardized manner across different teams. And this cloud transformation journey provided us the opportunity to focus on what should our DevOps processes be, what should our tooling be, how should we monitor our infrastructure. And now we are -- our teams are deploying 100, 200x a day in an automated fashion whereas prior to the cloud transformation, we did 6 months, 9-month long projects that were executed serially. So tremendous efficiencies in terms of how the teams are working. At the same time, from a business perspective, we are able to learn and iterate as we throw features that our customers then inform us, whether they like it or not. So that's been -- the biggest benefit is seeing how customers adopt our features and then our product team is learning and iterating.

That's awesome. That's fantastic. Well, I'd maybe like just to shift the conversation a little bit and maybe talk about Elastic, given that I think we're roughly about 2 years into our relationship. I know the team has actually been using Elastic for much longer than that.

Yes.

But from a commercial relationship perspective, I'm curious if you could just tell us a little bit about how Ancestry became a customer of Elastic?

Yes, certainly. Our teams have been using the open source version of the stack for a long time. And they were very impressed with the real-time alerting capabilities and search capability and ability to correlate logs, et cetera. And as the company was scaling, as we were going through our cloud transformation, essentially, we also felt the need to be able to scale our Elastic infrastructure. And the paid subscription that you guys offered provided us the opportunity to be able to take advantage of, I mean, literally instant provisioning of the clusters with a predefined architecture. And that kind of speeded up deployment for us, the things which we were able to do within a week or 2, now we can do like in less than a day. And the full orchestration capabilities has been of tremendous value to us. So it's very natural for the teams to get on to the Enterprise version as they were very familiar with the open source version before and happy with it, and it was just a -- we thought of it as a natural progression in our transformation efforts.

That's great. So very much, it sounds like in line with your overall strategy of streamlining operations, speed to market instead of having paralyzed activities having parallel activities?

Absolutely.

That's great. That's awesome. Now, I don't know if you're a hands-on keyboard, CIO. But I am curious, maybe you got feedback from your team in terms of just how easy was it to get started with Elastic? How easy was it to set up, start using it and advance with it over time?

Yes. Yes. I mean, certainly, I -- personally, I don't code as much now. It's been a few years. But I certainly -- being a technology geek, I get into the details of the teams. And one thing they tell me consistently about Elastic, and I've seen that in the results that they have produced is that it's one of the tools that they are very happy that they find it a little hard to keep up with in terms of the features that you guys develop, which is a good thing that we expect from the tools that we use to manage our infrastructure or get analytics from. And in terms of ease of use, as I said, to provision and deploy an Elastic cluster now happens in few hours as opposed to other areas where we still have ways to go in terms of streamlining, deployment and efficiency that we would gain through that. Because from our perspective -- from my perspective, I want my teams to be able to do those efforts where they are directly contributing to the features on our product platform. And we'd much rather leverage the expertise that tools like Elastic or vendors provide so that we can -- as much as we stay away from that, it not only gains us efficiencies but also refocus our teams on our own product offerings to our customers.

Yes, I love that. I mean shifting resourcing to engaging with your data rather than managing it is something more certainly we take seriously, something that I'm super happy that you're appreciating that value.

Yes.

I am curious if we could double-click a little bit just on how specifically Elastic is used in the context of Ancestry? Maybe share with us a little bit about how specifically it's being applied today. And maybe generally, what are the most powerful things you've seen out of the offering?

So we do analysis on all the log data that we collect from different services and infrastructure elements from our -- both from our cloud as well as some of the application sources. And of course, as I mentioned, the real-time alerting capabilities with instant search has been of huge value for us. And in addition to the normal anomaly detection capabilities, et cetera, what we found to be extremely useful were some of the machine-learning capabilities that you guys have that really helps us get granular -- a little bit more granular into what we call some of the low and slow use cases where things may fly under the threshold of the -- or under the radar. But it's important to catch them ahead of time so that before something becomes a bigger issue or if there are analytics that we can glean in ahead of time, it provides us better information to avoid an incident that might be starting off slow. And specifically, in sort of the fraud screening area and some of the security process that we have, we found that to be immensely valuable.

Let me take a minute actually to inspect those comments just a little bit further because you brought up a couple of great points. So really, you're taking log data and using that to inform other services now that logs would -- I would traditionally associate with operational type of activities. And actually, you have security teams focused on fraud, also looking at that data as well, that's an interesting perspective.

Yes. Yes. Absolutely. Yes. We found that to be extremely useful for those teams to be able to perform analysis on that data real time. As you know, the nature of that space is that we can afford to -- I mean, yes, we can look back from an analysis perspective for a later time, but it's equally important to get that information as much real-time as possible.

And I love the point that you brought up about the anomaly detection and the machine learning specifically. You guys have been a great partner in that journey with us as we've certainly moved in that direction as a company quite heavily because your point about -- sometimes, it's not obvious. Sometimes it doesn't take a developer or an engineer looking at a screen to see an anomaly because sometimes those anomalies are so insignificant in the grand scheme of things. So threshold-based alerting isn't -- sometimes isn't good enough.

Yes. And especially from our own product perspective, as you know, we are a Big Data company before Big Data was in the world. And we use machine learning and computer vision and artificial intelligence to digitize our content assets. And we have a dedicated data science and teams that work on these capabilities, but we would much rather focus them on our product experience and our product capabilities. And that's why when we compared your capabilities, we felt as machine learning has provided us the value on our product side, the machine learning capabilities that your tool has, we wanted to leverage on the kinds of work that our teams do on the infrastructure side.

That's great. So maybe to kind of bring this to a little bit of a closure. By the way, I love chatting. This is great. I love learning and listening. This is how we do it at Elastic. I'm curious you're thinking about just the role of search more broadly in the context that maybe -- outside of maybe the traditional logging or observability area. And how you think about potentially the use of Elastic in ways of Enterprise Search or even to expand into areas like security, like you mentioned before?

Yes. I mean, certainly, I mean, search is the primary means by which we get to information these days like whether it is on the Internet for personal use or on e-mail or in any form of interaction, you would see both my little kids do the same is like anything they want, they just go in and search before I could tell them what to do or how to do things. So likewise, with the advancement in search technologies, especially the kinds of algorithms and indexing that you guys have, certainly one feature that we are doing a proof-of-concept on is the Enterprise Search capabilities that -- on the Elastic platform. So we feel that, that would be very valuable. If we are able to get the same kinds of search results and analytics from that across the different tools and applications that we have in our enterprise as we get from log data or other forms of analysis that we currently do. So certainly, the Enterprise Search use case is very appealing.

I can imagine your data volumes, that's not going to be an easy technical feat, but we look forward to partnering together on that for sure.

Absolutely. Yes.

That's great. Well, Sriram, I can't tell you how thankful I am. I really appreciate you taking the time today to share your Elastic journey with all of us. And I look forward to chatting more about what's next. So maybe we'll go ahead and pause there, and thank you for your time again.

Thank you. Glad to be here.

Awesome. With that, I'd like to pass it back over to Anthony.

I hope you've all enjoyed the presentation so far. We now will take a 10-minute break before we get into the financial portion of our presentation, followed by the Q&A session. [Break]

Ladies and gentlemen, we will resume the program in 2 minutes. We'll kick off with our CFO, then head into Q&A. [Operator Instructions] We'll queue up your questions once we get to that section. Thank you.

All right. Let's go ahead and get started again. Welcome back from the break, everyone. I hope you've enjoyed the session so far. Let's jump into how we're investing for growth and building for the future. Over the next 20 minutes or so, I'll describe how we've built a strong foundation for growth, recap some of the points that Shay and Justin covered around our opportunity and how we will capture that growth and touch on the financial model and how we will demonstrate the efficiencies inherent in the model over time. Let's start with the foundation to drive growth. It all starts with a powerful business model that we've built. As you've heard earlier, we start with driving strong technology differentiation by looking at problems from the lens of search. Shay described this and shared why we believe that the defining elements of search, speed, scale and relevance, create a compelling technology proposition. Although search can be applied to solve a variety of business problems, we focus on 3 solutions: enterprise search, observability and security. That allows customers to realize value quickly with ease of use and compelling features in those areas, all built on a single stack, which, in turn, drives technology advantages across each of the solutions. We then drive massive adoption of our technology through our free and open tier, which includes 3 proprietary features. This distribution model has been one of the core underpinnings of our success in scaling so quickly. But we also always kept in mind that the free tier is a distribution model, and we built a compelling commercial model on top of this distribution model. In our commercial model, we offer paid subscriptions that deliver powerful proprietary features and support, all with a single resource-based pricing model. Our customers find this pricing model aligns with the value they derive from the software and puts them in control of their investments. All of this has helped us acquire thousands of new customers, and offering our technology over the cloud makes the commercial model even stronger. In the cloud, we monetize all our features. Even open source and basic features become paid since it's all offered as a service. And to round out the business model, once a customer buys from us, we make it easy for them to keep buying from us and growing their spending with us. That expansion happens in many ways. Data volumes grow, customers start many projects for the same solution or they adopt additional solutions. I'll talk about this a bit later as well. So overall, it's a powerful model with strong network effects. Adoption drives more data and more usage, which drives even greater adoption and expansion. And this is all evident in the financial results that we've delivered to date. If I compare Q1 of fiscal '21 to Q1 of fiscal '19, we've more than doubled our revenue in 2 years, and we believe we've created a healthy path to future growth with strong deferred revenue on the balance sheet, which has also more than doubled in that same period. What's equally important is that this has been primarily recurring revenue. As I noted earlier, we sell subscriptions. 92% of our revenue in fiscal '20 came from self-managed and cloud subscriptions, with the remaining 8% coming from professional services. And of the 92% subscriptions, 70% came from self-managed subscriptions, where the customer downloads the software to deploy either on-premises or in the cloud of their own. And 22% came from Elastic Cloud, our family of SaaS offerings, where we offer the customer a SaaS subscription that we host. What's even more exciting to me is that the SaaS mix has been steadily growing and our SaaS revenue more than doubled last year, growing to 22% of our total revenue. And finally, as we scale and grow the business, operating leverage inherent in the business becomes apparent. We invest upfront in building great technology, invest upfront in acquiring customers and then both recognize revenue over time and drive expansion within our customer over time. So to recap, we've built a strong foundation for future growth with 4 key elements: a powerful business model that has strong network effects; a demonstrated track record of driving growth; a high recurring revenue model, including SaaS growth; and starting to deliver operating leverage that's inherent in the business model. We believe that all these elements set us up well to address a significant market opportunity that lies ahead of us. Let's touch on that next. Shay already talked about the large and expanding TAM ahead of us. So I'll talk about how we acquire customers and then expand our relationship with them over time. There are a couple of ways in which we acquire customers. One is through our robust distribution model, where customers initially download the software. And over time, through nurture campaigns such as webinars, meet-ups, training, documentation and more, they eventually become paying customers. They can also sign up directly for our SaaS offering on the web, where after a short trial, they can pay monthly with no commitment. Customers see value in our paid features and support. And in the cloud, we monetize even open source and basic features. Customers see significant benefit in our single pricing model that puts them in control of how much they consume and pay and connects directly to the value they receive rather than being forced to pay more through some artificial punitive construct even when they didn't benefit more. All these motions are supported by an expanding sales team that helps customers connect value to business outcomes in the 3 solution areas and which in turn drives annual subscriptions. Thanks to these powerful motions, we've grown our customer base, adding thousands of customers every year. A large number of these customers are on cloud. Let's take a look next at the expansion we drive once we've initially landed a customer. Justin talked about the customers over $100,000 ACV and customers over $1 million ACV as well and how we've successfully grown these customers over time. This demonstrates the strategic commitment that customers are making to Elastic technology and how we are becoming more relevant to their businesses and to their success. Another representation of our expansion is in this cohort chart, which is likely a familiar format to many of you. It shows how we've expanded our customer relationships based on their cohort. If you look at the FY '16 cohort, for instance, it shows that in FY '20, those customers spend 3.2x their FY '16 spend. And other years also show strong expansion. Thanks to the strength of our solutions, we've maintained a net expansion rate of over 130% for a few years now. So what drives this expansion? There are 3 main levers. First, as the size of a customer's data footprint grows, they commit to more capacity. That data growth could be just because they experienced increases in data volumes over time or because they see so much value that they start to bring data in from other sources. A logging project starting to ingest APM data, for instance. Another lever is that they start more projects for the same solution. A team in division A sees their peers in division B have been successful with Elastic, and they start a project there. Often when our salespersons engage in an account, they are helping connect dots within the customers' organization to show them how Elastic is being used in different parts of their own organization. The third lever is that the customer adopts more solutions over time. That could be because a project extends beyond its initial scope. For instance, every log is also a security event. So the customer could extend the logging project into a project for our security solution or it could be because of the success they've enjoyed in one solution, the customer naturally adopts us for another solution. For instance, we could be powering a mission-critical enterprise search use case for them. And based on that, they decide to adopt us for logging or APM or security. Justin brought these expansion models to life in all the customer examples he provided. Now in all these expansion motions, a customer could also choose to move to higher tiers of subscription, to Platinum or Enterprise, as they seek more features available in higher tiers, address more complex orchestration needs or need high levels of support. And that's probably a good reminder that we only sell subscriptions by tiers. So when we discuss solutions mix and penetration, the information is directional in nature and generally self-reported by our sales team. In the past, we've shared that logging was more than 1/3 of our business. We have since then seen a greater shift towards observability, and security has seen strong traction and increase in the mix since the launch of our new SIEM solution several quarters ago and since the acquisition of Endgame. And although these solutions have increased in mix, enterprise search continues to grow quite nicely in terms of overall growth as well. So let's dig a bit deeper into the solutions penetration in our customer base. If you look at the customers with more than $100,000 in ACV, you will see that over 50% of those customers have adopted us for only one solution, either enterprise search or observability or security. And over 30% of those customers have adopted us for 2 solutions, and less than 20% of those customers have adopted us for all 3 solutions. This represents a significant growth opportunity for us to continue to drive solutions into this customer base. And we know that when we drive more solutions, those customers grow with us. If I look at our customers more than $1 million in ACV, over 45% of those customers have adopted us for all 3 solutions. Now in many cases, the footprint for the second or third solution could be small and it could be just an initial deployment. This actually represents our bottom-up sales motion in action. We start small in one solution, grow big, while extending in a small way to other solutions and then growing those over time. Both the solutions penetration shown on this chart and also the opportunity to expand the solution over time once it's been adopted show the growth opportunity ahead. One important thing to also highlight on this chart is that we define solutions at an aggregate level. Observability is only one solution for us. Under that, we treat logging, APM, monitoring and uptime just as products or features. So when a customer who's using us for logging also starts to use us for APM, we don't claim that they are using us for 2 solutions. It's still just one solution, observability. And since I mentioned APM, I'll also point out that although it's early days in our APM efforts, we have hundreds of customers starting to use APM already. The adoption here is similar to our bottom-up adoption model. People initially use paid features. And over time, we expect this will grow into larger deployments. We're already seeing deals where we win against competitors for APM. Let's quickly turn to cloud, which is another growth lever for us. We've demonstrated a very strong track record of growth there over the past several quarters. I'll unpack the revenue components of that for you. As I mentioned earlier, SaaS made up 22% of our revenue in fiscal '20. A bit more than half of that came from our prepaid model where customers commit for contracts of at least 1 year, and we bill these annually. So these appear in billings, deferred revenue and RPO. And about 10% of our total revenue came from the monthly model, which is the self-service models where customers don't have an annual contract, but they just go month-to-month and we invoice them in arrears. This stream does not have deferred revenue or RPO. I'll also point out that even though the SaaS business has both these components, I prefer to look at the SaaS revenue in the aggregate since there could be interplay between these 2. For instance, a customer with an annual contract that exceeds their contracted level in any month, the on-demand usage above that contracted rate is billed to them on a monthly basis. That's great for us because it demonstrates that the customer is continuing to expand, and it demonstrates their commitment and is likely to lead the customer to increase the size of their commitment to us. And when that happens, that reduces the on-demand monthly amount while the annual bucket increases. Our SaaS business has grown, thanks to all the investments we've made towards this: the partnerships that Justin talked about earlier, our own focus and investments in the go-to-market motions and also as we've expanded our presence around the world. We are now up to 41 regions where we offer Elastic Cloud. As a point of reference, we were in only 13 regions at the end of fiscal '19. We added 17 regions in fiscal '20. And so far this fiscal year, we've already added 11 regions so far, taking us to over 3x where we were just 6 quarters ago. We continue to invest in our SaaS business and expect that cloud will continue to be a strong growth driver for us looking ahead. To recap the opportunity on our growth levers, we have a large TAM. We have demonstrated a strong pace of new customer additions. We've also demonstrated our ability to drive expansion. We believe that we have significant more room for growth looking ahead across solutions and cloud. Let's turn to our efficient investment model. As I mentioned earlier, as we scale and grow the business, operating leverage inherent in the business model becomes apparent. As we drive revenue growth, we also continue to invest in go-to-market to drive near-term growth, into R&D to drive innovation and long-term competitive advantage and into G&A functions to support our scale. And we also demonstrate operating leverage with improving operating margin. We believe that as we grow, we can do both. As we scale and realize efficiencies, the economics become evident over time. And you've seen this in our results, with modest improvements in fiscal '20 despite the dilutive acquisition of Endgame that year and again in the commitments that we've made for fiscal '21. Let's first take a look at gross margin. The way I think about gross margin is like an enterprise software model. Given hosting costs, the gross margin of Elastic Cloud is lower, but it is accretive to gross profit dollars, of course. As Elastic Cloud has grown, we've successfully managed to maintain our gross margins in the business overall, offsetting the headwind from the mix change with operating efficiencies we have driven. While we are scaling, Elastic Cloud will continue to be a modest headwind to overall gross margins as we continue to invest in this part of the business. We will also continue to make disciplined investments in R&D. We don't use our model to outsource R&D to the community. We remain the sole committers of code, and we believe that we must continue to invest in R&D to drive competitive advantage. Last year, we acquired Endgame, which is what drove R&D as a percentage of revenue a bit higher. Looking ahead, we do expect that R&D will come down as a percentage of revenue through the operating leverage in the model. Shay touched on some of these aspects earlier, a single code base across the stack and solutions that allows us to be more efficient and features built into the stack that can be deployed across solutions, which is what we call our power plays. So R&D is another example of how we can do both, increase our total investment while delivering operating leverage. And we continue to believe that innovation need not all originate within the company. Over the years, we've actively pursued a model of building, partnering or buying where needed. Many of our important capabilities came to us through companies with which we joined forces. We remain open to M&A, especially tuck-in transactions or acquihires. As I think about larger transactions, while we wouldn't rule anything out, we believe our current solutions portfolio and road map aligns nicely with customer needs and the market opportunity that we see ahead. Go-to-market investment is again an area where we can do both, expand our capacity and capabilities while also delivering operating leverage. Our distribution model makes our go-to-market model efficient. We continue to invest globally to drive growth. And over time, we expect to realize leverage through the areas you see listed here. Justin described how we've been successful so far, and we are particularly excited about Paul's addition to the team. We look forward to driving both near-term execution and building the longer-term capabilities needed to successfully capture the full breadth of the opportunities ahead. As I've mentioned before, the leverage we've started to deliver while sustaining a high overall growth rate in the business also allows us to make great progress towards our free cash flow goals. As a milestone, we're aiming to be free cash flow positive next year. In the meantime, our operating cash position remains strong. Finally, I'll summarize our financial framework for you. As you've seen, we are a subscription business, and we continue to expect that subscription revenue will remain 90% or more of the business. We will continue to aim for Elastic Cloud to grow faster than our overall business, driven by customer preference and our investments. This should drive a gradual increase in our SaaS mix, and we'll continue to invest in the business with discipline while also demonstrating operating leverage. To that end, we expect we'll deliver modest annual operating margin gains. We believe that will help us achieve our goal of positive free cash flow next fiscal year. And finally, we recognize that many investors focus on stock-based compensation since there is a real economic cost associated with that. We agree. While the accounting expense is one yardstick, we believe that the economic cost is best reflected in the share count dilution associated with stock-based compensation. We've managed net share dilution very efficiently since our IPO, and we expect this will remain less than 5% each year. So to recap our efficient investment model, we've demonstrated the operating leverage that is inherent in our business model. We are continuing to invest in Elastic Cloud, in R&D to drive innovation and in our go-to-market to drive growth, all while continuing to drive annual gains in operating margin and targeting to be free cash flow positive next fiscal year. With that, let's move to the Q&A section of the event. I'll turn it over to Anthony to facilitate.

Thanks, Janesh, and thank you to everyone for spending time with us today. We appreciate your interest in Elastic and hope you found the event to be insightful and informative so far. We now have approximately 45 minutes remaining to answer your questions. [Operator Instructions] I plan to moderate the discussion, and we'll attempt to get to as many questions as possible in the time allotted. Joining me in the discussion today is today's presenters, Shay, Janesh, Justin as well as Paul Appleby, who serves as President of Worldwide Field Operations and is a recent addition to the team. As we continue to queue questions, maybe we start off with a question for you, Shay. Shay, obviously, a lot of good things came out of ElasticON keynote earlier today. What excites you the most about some of these ElasticON events and the ones you highlighted earlier in the presentation?

Thanks, Anthony. And first of all, welcome, everybody, and welcome to our Virtual Investors Day. I'm happy to host all of you here and excited to answer all the questions that you have. It's always very hard to go and pick one set of innovation or features that we announced during the conference. First of all, there was just a selected few in my keynote, and then there were like keynotes that we had around the stack and the solutions. If I had to pick one with everything else that we've built, I would go with searchable snapshots, the ability to go and store data seamlessly on systems like S3. That's a major innovation that we've been working on for the past 2 or 3 years. It requires foundational changes in the stack itself. We used to be fast. And as a result of it, everything is interactive. The APIs return seamlessly. Dashboards load seamlessly and fast and things like that. We had to go and, to a degree, make our systems slow and be able to handle with slightly slower storage systems, which means that dashboards can now load asynchronously. APIs can be fire and forget, and then you can check for their status. Tons of things and investments that we've done at the core of the stack itself. And then on top of it, we've built the ability to take the existing data that we have and store it natively on systems like S3. We heard our customers coming to us and saying, we love what you do. It's great that I can store 90 days of data, 120 days of data and get millisecond-level responses. I'd love to get to a point where I can store years worth of data. And let me control what I can do and be able to go and store it efficiently. And we've been working on that, especially with the ability to go and seamlessly do that. The same dashboard, same APIs, all of that works exactly the same. One of the data stored locally on fast solid-state drives or on a cost-efficient system like S3. I'll add on top of it, which is I find it as a power play of ours, the fact that we invested so much on schema on read means that when you take the data that we store and make something like S3 searchable, it means that we can search even when data is stored in S3 extremely fast compared to other solutions. That makes me very excited. I mean if other systems will take hours to scan record by record to try to find a result on an object storage like S3, we already have the data structures in place with our schema on write to be able to go and find it much faster. And it's only the -- it's not only much faster. It's much cheaper when that query executes because you don't have to go and scan through all of that data and pay for the cost of retrieving the data. We can go and selectively pick and choose which parts of the data you want to go and get, thanks to our schema on write, and make that more efficient. So there's one of those special moments where we take a feature we implemented, and this is like years of innovation that go all the way back to our significant investments in core things like schema on write. And beyond that, I'm also very excited about the fact that this significant feature, it will be provided part of our Enterprise subscription, which is our highest paid subscription tier. It's one of those special moments where we can provide significantly more value at a higher paid tier in the case of Elastic. And our free users will have another reason to engage with us commercially when they're running on-prem or in a self-managed environment. And our cloud customers will have an opportunity to upgrade to a higher tier, while we provide significantly more value and the ability for them to reduce their total cost of ownership, right, the number of servers that they potentially need to run, the number of storage systems they need to hold, if they had to store that data somewhere else in other systems.

Great. Thank you, Shay. So our next question comes from Brent Thill at Jefferies. And I believe this best goes to Shay. Can you give us an update on your security initiatives? What are you most excited about? And where is the low-hanging fruit? Can you give us a sense over time how big this business can be versus the other categories?

Yes, of course, happy to. So I think the most obvious and straightforward, I would say, not straightforward technically but straightforward from a business perspective, for our evolution in the security market is to get into the SIEM market, the S-I-E-M market. Every log event, every IT ops log event is also a security event. And we can use that and expose it in different ways that are curated to security practitioners, allow them to define rules and machine learning algorithms that run over that data, basically that data center level data, bring all of that data from various sources, network, hosts, cloud, bring it all into a single place and being able to go and analyze it and protect your company against any type of attacks that is happening. And we've been investing in that. We've GA-ed our SIEM solution. Endgame joining us didn't only allow us to improve our ability to go and deliver endpoint security, which I will touch in a second, but also just bring all of that security DNA into Elastic, more than 100 people into our engineering team to be able to really double down and speed up our innovation when it comes to the SIEM market. And even within the SIEM market, I mean, it's built on top of the core foundations. I talked about searchable snapshots. It's going to be very exciting to be able to bring that to the hands of any security practitioner out there and not have them be bounded by storage. And beyond that, I think our openness as a company, for example, when we went and started to create security rules and detection rules that are part of any SIEM system out there that are done in the open and in a collaborative manner with the rest of the security community, which I'm super excited about. So that's the most obvious next step for us as a company. Beyond that, we're working extremely hard to fold all of endpoint protection that we got through our Endgame -- to us joining forces with Endgame and folding it into the Elastic Stack. This ties into our single-agent technology, while you observe, why not protect. And we started that just recently with our latest release, where we folded antimalware protection into our unified agent technology, but we have quite a few protections left in the checkmarks to go and add when it comes to endpoint security. Our goal is to provide the best endpoint security or endpoint protection product out there that is part of our security suite, if you will. That's more long-term for us as a company. And obviously, the opportunity over the long term to not only fold endpoint security together with SIEM to make it very, very seamless to a user observing your data, going to insights and going to action and actually protecting the system, stopping threats, but also the ability to fold security and observability markets together because many times we're talking about the same data sets.

Great. Thank you, Shay. So next question is for Justin, comes from Ittai Kidron at Oppenheimer. What's the threshold for inside sales engagement? Are partners used just for fulfillment? If not, how much of business originates from partners?

Great. Thanks, Anthony. Yes, some very different questions. So let me actually -- let me break it down into a couple of answers. So let's talk about partners first. So partners with respect to Elastic are -- certainly, there are fulfillment partners. Fulfillment partners play a role in terms of the actual end state of a sales cycle. And certainly, Elastic has those fulfillment partners. We look strategically on partners though in many different categories. Our cloud partnerships, for example, across Google and Microsoft, in particular, Tencent, Alibaba, strategically important to us, emerging with AWS as well through their marketplace. We view our cloud partnerships as incredibly important to our go to market. So certainly, they can act as a fulfillment arm, but they're also very involved in that front part of the sales cycle in terms of sizing, scoping, architecting a proper solution for a customer, identifying the business challenge and the pain point that goes with it. So we also have categories of partners that fall into MSP or OEM-type categories. We also have local GSIs and solution-oriented partners. They bring Elastic to market and have built it in as part of their discussions that they have with their customers every day. So our partners are playing an increasingly critical role in their business as we're helping them create business momentum for themselves through either consulting services, training and education that they provide, other add-on technologies in which they're able to sell in conjunction with Elastic, whether that be infrastructure or cloud-related services. So we're really excited about the ecosystem that we've been able to build around our partnerships and the leverage that, that gives us in terms of just continued growth. Now with respect to inside sales, so inside sales is -- actually, I'm incredibly excited about the investment that Shay and the entire leadership team has given me to be able to invest and build out an inside sales team. Our velocity business, as I talked a little bit earlier around our go-to-market around the land and expand opportunity that we have, our velocity business across all of our segments is just so critically important to our success. And our inside sales team is the tip of the spear in a lot of ways. We have a very, very advanced sales development team that actually helps us qualify and quantify the vast number of leads that are coming inbound to us, helping us build a prospect team and in-account-type discussions and webinars and seminars that we do with our customers on a daily basis to drive continued demand, to drive continued awareness within our account base. But we don't have a threshold per se. I presume the question is around a dollar-based threshold. I don't think about it in terms of a dollar-based threshold about what is an inside deal versus what is an outside deal. I think collectively, we talk a lot internally about how are we going to solve the customers' problems, how do we align resources to best answer those customers' questions. And we drive through that as our lens of success as opposed to establishing some artificial dollar-based threshold in terms of who engages.

Great. Thank you, Justin. Next question is for Janesh. It comes from Mark Murphy at JPMorgan. Janesh, given the comment from Justin about leading with the cloud, do you have a long-term target maturity -- or maturity target for your cloud revenue mix?

Thanks, Anthony, and thanks, Mark, for the question. So we do lead with cloud, as Justin said. And at the end of the day, we believe it's better for the customer and better for Elastic for those workloads to be deployed in the cloud. And as we also said, we ultimately remain agnostic to the customer's final preference. We are where our customers are and where they eventually choose to deploy applications. Many customers are just at different points in terms of their journey to the cloud. And so we believe it's important for us to be there wherever they are, and some customers are much further along and some much earlier. And some are more ambitious than others. And they have big plans, but then very often, their plans go slower than they expected that they would. And you see this reflected in our revenue mix as well. So it does reflect the move of those workloads to the cloud from a customer lens. But we remain agnostic. So we expect the mix to only shift gradually over time as it has done between fiscal '19 and fiscal '20, for example. I will point out though that this actually plays to our advantage because of the primary mode of deployment for customers today on cloud is a hybrid cloud approach. And so when we think about our ability to be where customers are and to offer them our capabilities, both in a self-managed format as well as in a fully hosted format on the cloud, that plays to our advantage compared to what you might see with a few others out there in the industry. At the end of the day, the data has a lot of gravity to it. And as we talked about, people will deploy us where the data resides. And so we're pleased that we're able to actually be in all of those different locations. And this also underlies the reason that we continue to invest so heavily in cloud as we continue to build out our presence across different geographies in different regions around the world and with different cloud partners.

Great. Thanks, Janesh. Next question is for Shay, and it comes from DJ Hynes at Canaccord Genuity. Can you talk a bit about the road map for your ML and AI capabilities? What makes sense for Elastic to own? And where are you better suited to just be the data engine that powers third-party tools? Where does that line get drawn?

Yes, of course. So first of all, the way that I think about it is not necessarily by drawing lines but by thinking about it in the context of layers. So I'll start with our basic layer where we provide a place to store data now for infinite period of time through our searchable snapshots capability that we're going to roll over the next few quarters and open APIs. That means that any tool can go and integrate with us, especially giving our free and open distribution when it comes to our self-managed customers. And we've seen a whole community of integrations happening across the board from older-school, Python-based libraries to newer ones like TensorFlow and OpenAI tools. So I'm super excited about the ability to integrate with all of the significant innovations that is happening across the whole community. We can't implement all of it, and we are building an open platform around data and around search that allows for any type of machine learning or AI to integrate with it and integrate with the workflows and the decisions that various companies do when it comes to their investments. When it comes with what we decide to bring into the stack and provide out of the box on top of that, that tends to be driven specifically by the needs of our solutions. If you think about it from a solution perspective, many of our customers expect to have certain behaviors out of -- all capabilities out of the box. It goes all the way back to several years ago where we joined forces with a company called Prelert because our observability users asked for better ways to find anomalies in their data and easier ways to do automatic classification of their data. And that's something that we've built into the stack and then exposed to our observability users and now our security users and all the way to, for example, our Endgame acquisition, where there's a significant amount of investments that go into our data science team on the AI/ML level of capabilities that go into the endpoint and into the rules and detection rules that we have to drive the ability to go and protect your company. So we're thinking about it in the context of layers. We're happy to be open and integrate with any system out there that supports ML, and we see that by the investments of our community. And we're working on bringing and folding in ML capabilities that drive out-of-the-box experiences through our solutions.

Great. Thanks, Shay. So coming back to you again, Shay. This comes from Raimo Lenschow at Barclays. You seem to be talking more about search again as your product evolution enabled you to differentiate more again versus -- against -- versus AWS Elastic. And please explain.

Sure. I spent some time talking about it during my portion of the presentation. First of all, we view search as the foundational layer that ties every single use case out there that Elastic is being used for and specifically across our 3 solutions through our capabilities to deliver speed, fast and relevant actions and outcomes out of your data. When it comes to building differentiation, I'll mention that we -- about 2 years ago, we've started to introduce an open, free yet proprietary tier called Basic. And since then, we've made significant investments into it. Our effort was to try to create a system that still supports our big community of users that love the fact that we have our coding be open, we actually opened all of our code, love the ability to go and integrate and exchange ideas with us on that level. And they love the fact that they can download the software and run it themselves even if it's on a laptop and test it out. And all of this, we're enabling them to our free and open distribution. But we did put it under a proprietary license to make sure that these features can only be applicable or available on our Elastic Cloud. Within that, I'll just mention the set of innovations that we have announced or we touched on during the keynote and today in the presentation. Searchable snapshots, they go into our Enterprise subscription tier. That is proprietary and paid for on self-manage and a higher pay tier on cloud and others like Elastic Agent, actionable Kibana, and schema on read that goes into our free proprietary tier and above. So definitely, significant investments going to our proprietary tiers that helps drive that differentiation. I'll end with just mentioning a number that I'm very proud of. Our community and our users are very supportive of these investments. They totally understand what's going on and they support our efforts in this space. And that, I think, is reflected by the fact that more than 90% of our downloads are actually done with our default distribution that has all of the free proprietary features turned on by default. So that's a significant value that we continuously provide to our customer base and our user base. That helps drive that differentiation on cloud over time.

Thanks, Shay. Next question is for Justin, and it comes from Matt Hedberg at RBC. Can you talk more about why you win in the cloud versus peers and also about your cloud partnerships versus the peers?

Yes. Yes. Thanks for the question, Matt. I would actually start by dovetailing off what Shay just mentioned, which is our default distribution today, 90-plus percent of our downloads actually in self-managed world happen with the inclusion of our Basic license. And what we're seeing is that translates into the cloud as customers are wanting more and more the experience that we have helped build and curate within all of our products. That is inclusive of our Basic license. And as customers seek other technology providers out there that may offer open-source Elasticsearch natively, they're becoming more and more focused on the full experience that we're able to deliver to them as the creators and maintainers of the Elasticsearch code base. So we're certainly seeing customers wanting that experience and wanting that in cloud as well. That is certainly driving a lot of adoption on our end. I think flexibility of choice is a big driver. We're seeing especially across enterprises, organizations adopting a multi-cloud strategy. Organizations needing a hybrid cloud strategy that exists where they can manage their data on-premise as well as in the cloud and having a consistent common framework platform, set of technologies, irrespective of the way you run it becomes really important for our customers. So having alignment with us in that respect, I think, also drives our cloud. I think our partnerships have become strategically important as well. We can help our customers onboard into Elastic Cloud through the native marketplace experiences that they have today with their current infrastructure providers and easily on-ramp, leverage their current spend or committed spend they have with those providers and get started with Elastic Cloud right away. I think that speed to market that -- not having to sit there and figure out how you're going to operate, manage the day-to-day of your Elastic environment, we can focus customers on how do they engage with their data quickly. I think that certainly is playing a huge role as well. So I'll leave it at there. I think those are some of the top reasons. Thanks for the question.

Thanks, Justin. We'll go back to Shay with this one. It comes from Rob Owens at Piper Sandler. As the market appears very fragmented and noisy, how are customers thinking about consolidation versus specific tools for use cases? In other words, broader operational tools are also lending to SEM and business intelligence capabilities. So how do you feel about your advantages there? And where do you feel you're most advantaged?

Yes. There's multiple areas that I think that at least we're investing in to make sure that we provide the best value to our customers as markets collapse, as products become feature, as we like to say at Elastic. Definitely, when it comes to us evolving from logging to observability, for example, launching a whole new fresh look on workplace search capabilities or our entrance into the security market. First of all, it's just the technology implementation that we have, the fact that we implement everything into a single technology stack, which requires a lot of investment. This is not an easy feat to do. And it's pretty complicated to do. But the fact that we're making the investments means that products can become features much faster, both in our eyes when it comes to developing it but also in our customers' eyes. It's very natural for our customers, for example, to move from logs to APM because they're already used to the same tool and they use the same mechanisms or evolve from logs to SIEM and then through our single agent technology evolve, for example, from SIEM to endpoint security down the road. So we're building the foundations to allow for our customers to move from one use case to another and hopefully defragment a market through technology and through value. The other one is our bottom-up adoption model. We go directly to the practitioner. If we build something that doesn't provide value, it hits us right in the face immediately. So it doesn't go through C-level executives or something along those lines. We have constant connections through our open distribution and open communication channels to know exactly what needs to be built and then self-correct and move extremely fast in that loop to make sure that we're building the right things for our users. That means that it allows us to have a significant competitive advantage from my perspective in terms of not only knowing where the market exists today but where it's moving down the road and the ability to build the right tools for the actual practitioners that take these tools and use them themselves on a day-to-day basis. That also means that we're not suffering from the lack of oxygens that exist at the C-level executive suite where it's really hard and exhausting to fight over the attention of a CISO or a CIO to just give you a chance as a company to do a POC. We think that through bottom-up adoption, we empower C-level executives across the world to make decisions that are driven by value, by usage of the software within the organization, by being able to go back to their teams and say, "I am going to enable you and make you supported and successful by doubling down on existing usage that is already happening." And I think that -- and speaking as a C-level executive, I love when that happens. When bottom-up adoption happens, whether it's through features in our products like searchable snapshots or through usage of various tools across our organization because it means that it has been tested. It's developed by the people that are closest to the problem, which means that we're making the best decisions possible. And then my role in the organization is to go and enable them. So that's something that we're trying to strive for as a company in terms of our go-to-market model, this developer-first, practitioner-first approach that I think helps us consolidate markets that are on a path for consolidation, either now like observability or in the future like security and observability.

Great. Thank you, Shay. Next question goes to Janesh and maybe some others here. It comes from Tyler Radke at Citi. You talked about how you have half the -- approximately half of the Fortune 500 and 1/3 of the Global 2000, yet your TAM penetration is below 1%. And with the Fortune 500 and Global 2000 having most of the world's data, this would imply that there is significant room to grow in these accounts. How do you get $1 million customers to be $10 million or $20 million? Maybe talk about what you see at your top 5 accounts in terms of spend and use cases.

Yes, I'm happy to start that off, Anthony. So first off, it's really exciting. In fact, this past weekend, I was actually cleaning out a little cabinet at home. And I rediscovered an old Calvin and Hobbes comic book that was called There's Treasure Everywhere, and it just reminded me about the opportunity that we have here at Elastic. So as I think about it, you're absolutely right with that observation, Tyler. When I think about the enormous opportunity that we've got ahead of us and the size and scale at which we operate today, I think there's just a significant opportunity for us to grow as we drive the business forward. And I think that comes in multiple dimensions. And if I think about it from the standpoint of the customer lens and the customer spend that you talked about, Justin provided several examples of cases where customers started off very small with us, and they've grown well beyond the $1 million threshold. Now they are multimillion-dollar customers for us. And we certainly see that they've got both the business need as well as their priorities are shifting towards the areas where our solutions are very well positioned. And so we do see the opportunity to continue to grow even further in those accounts over time and drive an even higher share of wallet. And I think the levers that drive that expansion for us are the ones that I talked about. It's both the volume of data growth -- or all 3, I should say, the volume of data growth as well as more projects associated with a particular solution or a use case and then the proliferation of solutions. And so certainly, when I think about our largest accounts, whether it's the top 5 or 10, we do see that there's been this deeper and wider adoption of solutions as part of our overall bottom-up adoption model where we get adopted in one part of the organization for something. And as that project grows or the solution grows, we continue to move further up. And in the meantime, we are getting adopted in different parts of that organization for either the same solution or different solutions. So we see much deeper solutions penetration. And it's actually a big part of how we will continue to move further up into the enterprise and continue to extract even greater value. And I know Paul is here with us. He's joined us recently. I know this is a topic that he's actually really passionate about. So Paul, why don't I just turn it over to you, and maybe you can offer a couple of additional thoughts.

Yes. Thanks, Janesh. I really appreciate that. And to be frank, this is one of the reasons that I joined the company. I saw the incredible potential for us to grow and scale the business into this near $80 billion TAM. And -- but I want to start, first of all, in talking a little bit about our superpower because I think it's really important to understand the power of this free and open distribution model creates an incredible momentum in the G2000 as the creator plus the developer, the practitioner is adopting and leveraging our technology and demonstrating value, as Shay quite rightly said in the -- within their organizations. We're earning the right to have these conversations with the C-level executives. Now of course, the other thing that I wanted to rip up as well is what Justin referred to earlier. We already have these million-dollar relationships with many organizations, and the ability for us to grow and scale those relationships is huge. Now the interesting thing about engaging with the enterprise is that this is a well-trodden path. The value-based solution selling model and methodology is one that's been proven in a number of organizations. So one of the things I saw when I got here is the ability to take what we already had that was already existing and evolve the nature of those relationships to build those million-dollar, multimillion-dollar ACV relationships into $5 million, $10 million and $20 million relationships. So the important message in that is that we're really talking about an evolution of our model and nothing revolutionary here. So it really means that our ability to drive scale and repeatability will unlock an enormous amount of this potential TAM as we kind of bring these more persona-driven, solution-driven engagements with the C-suite executives.

Fantastic. All right. Thanks, Janesh. Thanks, Paul. Next question comes from Jonathan Ruykhaver at R.W. Baird, and it goes to Shay. Can you talk about the enhanced value customers can achieve by using Elastics' unified endpoint protection with SIEM? I know you talk about fast and scalable user experience. Is that enough to convince customers to move off of an existing endpoint solution that probably already has integrations with existing SIEM, network and other security tools?

Yes. So first of all, I'll start with saying that our focus in being -- is on being an open SIEM. So we do provide -- or starting to provide endpoint security solutions or capabilities natively integrated into our stack, but our SIEM is open and integrated already with all of the various endpoint providers out there. And we're happy to do it because we'll take any data that we can to make -- to help organizations protect themselves. When it comes to our investments in the endpoint security market, I'll mention a few aspects. The first one is you need to be able to have an agent running on a system to be able to collect that data -- on the host to be able to collect that data and ship that security events. And regardless of -- and the fact that we support many different agents and many different collectors that can ship data to us, to open APIs, our customers expect us to provide them with an agent that does the same. So we've been building these single unified agents and the ability to create these security events over the last few years. And it culminated in this one click, one agent technology that we released recently. Beyond that, people don't like to install, I don't like to install on my laptop too many agents running on my system. So by the fact that we're already collecting all of that security -- these security events, we're already starting to form a footprint when it comes to our agent technology, we can go and provide more value to our customers by providing them with endpoint security. So our goal, the same, is to be able to go and grow bottom-up through value and building on top of the same SIEM level integration and value that we provide to our customers through an open SIEM by being able to go and provide them with more. I would also say that one of the things that I'm excited about is the ability to really change markets, not by making -- taking products and turning them into features within a bigger product or a bigger opportunity, but it's also to consolidate pricing and packaging models. We always asymptoted towards the value of data at Elastic. We always valued this the most. We thought that that's the biggest power that we provide, bring data to our system and be able to search it and drive data insights and actions out of it. And that's why we've had a single unified packaging and pricing model. That means that if you deploy our SIEM, you don't go and pay based on ingest or based on number of users or something along those lines. You just pay for the data that you use. And even there, we're innovating all the time. So if you want to store data for longer term, again, like we're asymptoted towards the low price of long-term object storage systems with searchable snapshots. And it means that when we joined forces with Endgame, we don't charge based on the number of endpoints that are out there. You can go and deploy as many endpoints that you want. We won't be the gatekeeper. If you suddenly have another company that joined you, another group, another team and being able to take this endpoint security and not only deploy them, for example, on your laptops but also on servers across the organization. And then those endpoints need to be centrally managed. They will ship security data that you would need to store anyhow as part of your SIEM. And we have one of the best SIEM -- scalable SIEM solutions out there that is open, and we think that the value that we can provide you is through there. So that's another way for us to go and basically look at these products and turn them into features, is the fact that we have a unified pricing model that help customers reduce total cost of ownership by doubling down on our usage as a set of tools for more than one use case.

Thanks, Shay. And the follow-up on that from Jonathan goes over to Justin. And how do you see the adoption moving forward? Recognizing the success you experienced with high-volume, frictionless sales process focused on the end user for search and observability, can you replicate the success with that sales motion in security? How do you see the importance of C-level engagement relative to practitioner in that go-to-market?

Yes. Thanks. I was thinking about -- it's a great question. It's -- maybe the best way to answer that question is through an example. Maybe just a most recent example, in fact, of a customer conversation, in fact a CIO conversation that I had just on Monday with one of those Fortune 500 customers that Tyler was talking about and asking about earlier. The relationship with this particular customer started a couple of years ago, and it was exclusively around open source use. Open source have become incredibly pervasive within the organization. They had started using it to solve departmental, small projects, but it started to become the hammer for every nail. Our relationship elevated over the following kind of 6 to 9 months. We actually did our first commercial transaction with that customer after about 9 or 12 months. I got a long-term contract in place, went through the appropriate process to get that agreement done so we could work with this customer and scale over time. That relationship then started at about $40,000 a year in spend, relatively insignificant. You fast forward to last Monday, and the conversation has changed dramatically. What started as open source use dovetailing into a $40,000 observability use case for just some basic logging infrastructure has evolved now into a customer-facing, search-based application, where they're intending to roll this out to all of their user base, all of their customer base to be able to attract the advancement of molecules and how those molecules are being created and the process in which they go through and which -- how they store that information. Incredibly important for their business, incredibly important for their customers and certainly center to their revenue, center to their go-to-market. So the conversation elevated very quickly to the CIO. And she was keenly interested to understand what we're doing and how we're doing it. And most important and interesting to her was how we're doing -- going to be able to deliver this to them in the cloud. So now we've got open source use, transitioning to on-premise subscription customer, transitioning to cloud, going from observability to search, not an uncommon trajectory for us with customers. As the conversation progressed, it was going well. I did ask the question. I said, "What about security?" And I said, "It's not uncommon that we see customers viewing observability and security as the same data set but just through different lenses." And she said, "That's a really interesting question and something we've been thinking long and hard about lately. In fact, we've been considering moving off of our legacy SIEM for a lot of different reasons." It was at that moment, and unbeknownst to me in the back of the Zoom as it were, one of their platform engineering managers stepped up and actually said, "We've actually been using Elastic in the construct of security for a while, and we're actually thinking it's the system that we're going to move forward with for security as well." So this conversation went up 360 degrees, covering a breadth of use cases tied to the value as a business. And we couldn't be more excited about how the bottom-up model just can continue to proliferate and create velocity at the C-suite as well. So hopefully, that example helped shed some light on the question.

Yes. Thanks, Justin. Paul, do you have anything to add there?

Yes. I mean I think it goes back to the comment I made earlier about the superpower of our free and open distribution model, driving this kind of organic adoption. And having worked in enterprise technology for some time, the Holy Grail is to be able to couple that kind of broad-based bottoms-up adoption with the kind of top-down C-suite engagement. And that's where we're looking to drive repeatability and scale, taking that incredible organic engagement with customers. And I think that example that Justin just shared is a fantastic example, that coupling that with enabling each and every one of our sellers with the content that they need to engage with the personas that are driving these decisions at a C-suite level. And that's exactly the work that we're doing right now, and it's going to help us unlock greater value in these accounts.

Great. Thank you, Paul. So next question comes from the buy side. You said in the keynote that schema on write is still better for most cases. Could you elaborate? What's the drawback from opting to schema on read? That goes to Shay.

Yes, of course. So happy to answer that. I mentioned a bit of relevant information in the context of searchable snapshots and why schema on write is so valuable. So I'll touch on that for a beat. First of all, over the last many years, we've been famous for our schema on write capabilities. This has provided the ability to do millisecond or sometimes second-based responses, snappy dashboards. We've perfected that schema on write. We provide you also flexibility. You can throw almost -- like throw almost any type of data at us, and we'll know how to index it and make it searchable. So even the level of structure that you need to provide to that data is very, very minimal. Beyond that, we even went and created common schema efforts that is now widely adopted across the industry called Elastic Common Schema to assure that we go and even try to formalize that in an open way across systems. What we do see is that even with all of that, there's still a small place where people still would love to have some flexibility around being able to make decisions after the fact that they ingested data. So we've developed this schema on read capabilities that makes it indistinguishable when it comes -- compared to schema on write when it comes to the usage of the system. So you can take the same dashboards, the same drag and drop visualization experience, the same rules that you would write in your SIEM and so on to be able to go and use that schema on read. It will be slower by nature. It's done on read time. You need to go record by record, the ones that match and evaluate it at read time. That's going to be slow. So we could only implement that after significantly investing at the core infrastructure level to make sure that our systems can work with slow searches and slow experiences or slower experiences like being able to load dashboards asynchronously and giving you a toaster notification that it's ready and all the way to other systems that we have. And that obviously also applies to our searchable snapshots capability and the ability to search across slower storage systems as well. The value of schema on write though is still there and it's still significant. So we're also investing significantly that you can, with one click, move from schema on read once you define it to schema on write. We believe the schema on write should be the default experience. We believe that there should be a single one-click experience to move from schema on read to schema on write. And the place where it manifests itself, it's not only on the last 60 days where you can load the dashboards in milliseconds or provide search as you type and snappy results with relevance baked into it across billions of records. Funnily enough, it also applies to the longer storage like cold and frozen level storage of data, as I mentioned, because the slight additional storage that schema on write requires gets manifested tremendously when it comes to the ability to actually have same response times from long-term storage systems, add much saner cost because you don't have to go and scroll through all of that data across that long-term storage that ends up creating additional costs associated with each query. We can go and pinpoint and find exactly what you need. So we're still huge believers in schema on write. It's our default behavior, but I'm super proud of our ability to go and provide more flexibility to schema on read and one-click move from schema on read to schema on write to our customers.

Perfect. Thank you, Shay. So let's go back to Janesh with a question from Mark Murphy at JPMorgan. A few years ago, about 1/3 of self-reported use cases were in logging. So search was a bulk. How has this mix evolved? And where does it stand today?

Yes. Great question. So as you know, we sell subscriptions based on tiers. And so we don't offer specific SKUs for each of the solutions. So a lot of the information on mix that we talk about is generally self-reported by our field team. And so we consider it more directional in nature. But that said, what we have shared in the past was that logging was more than 1/3 of the business historically. And we have seen a greater shift towards observability more broadly within that. We've also seen really strong traction on the security side, which is something that we're really excited about and an increase in mix since the launch of our SIEM product several quarters ago and then also since the acquisition of Endgame. Broadly, we've been very happy with the adoption of our security solutions. And in fact, even here at ElasticON Global, many customers have talked about their success with security, and we've got some really exciting offerings here that we announced as well. So although these solutions in terms of observability and security have increased in mix, enterprise search, I'll say, continues to grow quite nicely for us as well in terms of just overall growth. And again, there, we talked about some of the more recent launches that we've done as well, which continue to be exciting for us from the standpoint of the opportunity set ahead of us. And then in terms of the solutions penetration, I provided a little bit more color earlier in the presentation. And we continue to see just room to expand across the customer base and just drive even greater and deeper solutions penetration there over a longer period of time. And finally, maybe just a quick reminder that we -- when we talk about solutions, we just define that at an aggregate level. So for example, observability is one solution for us. And under that, we will treat logging and APM and monitoring and uptime. They are just separate products or features. And so if you have a logging customer that adopts us for APM, we don't claim that they are using 2 solutions. For us, it's just one solution. It's just observability. But overall, I think that's how we've seen the mix evolve, and we continue to be really excited about the future in terms of the scope and the growth opportunity across all the 3 solution areas.

All right. Thanks, Janesh. Unfortunately, we're butting up against the end of our time. So we have time for 2 more questions. This comes from Brent Thill at Jefferies. And Justin, if you had a magic wand and could remove any constraints or roadblocks from the go-to-market, what would that be?

Well, I guess I would say first, I've got something better than a magic wand. I have a very supportive executive team in Janesh, Shay, Leah, Sally, Kevin, Paul that all supports the additional investments that we need in the field to continue to hire, build out our cloud capacity, our coverage and how we support our users and customers. They're helping us build programs to move up into the enterprise effectively with the right messaging, building programs to drive our velocity model, which we've talked a lot about. Our velocity business drives our 6-figure and 7-figure transactions. They're helping us with leading with the cloud and the ability to carry that message forward and delivering great products that allow us to represent Elastic and hold our heads high in the field. Our -- anything that's holding up our execution in the field right now, it's -- frankly, it's really just the physics of execution. Sometimes it just takes time. We have an incredibly strong distribution model built on an equally strong commercial model. I truly believe that adoption of our technology is ultimately going to drive our growth. And the fact that we can tie one to the other, I think, is a huge advantage for us as a business. Now there's always going to be external factors. COVID-19 might be a perfect example of that. Things that we can't control that affect our customers and community at large. What we can do though is we can be there for them. And the fact that we have the continued support and the support of our customers, I think, it is just an execution game at this point and the opportunity for us to go demonstrate our ability to do that. As long as we continue to keep the motions around investments in our business and we keep the same commercial model going forward, I don't think we can be beat.

All right. Thank you, Justin. And so our last question goes back to Shay, and it's from DJ Hynes at Canaccord Genuity. It would be great to hear a little more about the Google partnership and get some detail on how that's contributing from both a go-to-market and a financial perspective.

Yes. Thanks for the question. I'll take a step back and just reiterate maybe the investments that we make and our strategy around cloud. Our goal is to be there for our customers wherever they are. So it means that we have made significant investments like Janesh mentioned with our cloud footprint across the 3 major cloud providers. We deploy on various regions and add new regions all the time. And our goal is to integrate natively within each cloud provider as much as possible and provide it as a native experience like any other cloud service that, that cloud provider provides while retaining the capabilities that we have at Elastic to be able to be the only cloud provider that has our proprietary set of features, allow -- have the ability to go and search in a seamless way across both on-prem or self-manage and across clouds, using our cross-cluster search capabilities, all of the things that we've described before. Specifically, when it comes to Google, super excited about our relationship with Google. Goes back to Google conference about 1.5 years ago, where we announced a much deeper partnership together with Thomas Kurian, to hosting Thomas or T.K. today at our keynote in our virtual event. Since then, we made significant progress. We work really well across all levels, all the way to engineering, by the way. We worked closely with Google engineers, data center engineers for a few months to find a solution to a really mysterious bug that help make our system run faster on GCP but help make any customer that runs on GCP faster as it was a fix in the opening system level, all the way to our business go-to-market, our sales, marketing relationships, our cloud product integrations. We've announced our billing integration, native billing integration with Google. We now exist on Google and Google console and within the marketplace itself, and we want to provide more and more options when it comes to billing through Google marketplace and through Google billing integration and using our cloud for it. Still tons of work left to be done, the ability to connect over a private network, the ability to have more buying options when it comes to our billing integrations, but I'm super excited about it because we have a partner there with Google that we can double down on. So I'm very, very happy with the relationship that we have with Google. And it's one that we're trying to build with all cloud providers to make sure that our users are successful regardless of which cloud vendor they decide to use.

Great. Thank you, Shay. So that was the last question, and thus, the conclusion of our program today. The replay and deck for this event will be available on the Elastic IR website in the next couple of hours. Shay, I'll turn it back to you for closing remarks.

Yes. Thanks, Anthony, and thanks, everybody, and thank you. Thank you very much for joining us today. I also want to just say a special thank you to Sriram from Ancestry. He shared a wonderful story with us about their journey with Elastic that is representative of so many other stories of so many users and customers that we've heard over the years. Super proud and humbled to have the team at Ancestry as our customer and humble and proud to have so many more customers and users like them. Hopefully, you found today's presentation to be insightful and useful. We're always here to answer more and more questions if you have them. Thank you for joining us, and we look forward to talking to many of you in the future. Ciao.