Prepared remarks
Good afternoon, and thank you for joining Atlassian's earnings conference call for the third quarter of fiscal year 2026. This conference call is being recorded and will be available for replay on the Investor Relations section of the Atlassian website following this call. I will now hand the call over to Martin Lam, Atlassian's Head of Investor Relations.
Welcome to Atlassian's Third Quarter Fiscal Year 2026 Earnings Call. Thank you for joining us today. On the call with me today, we have Atlassian's CEO and Co-Founder, Mike Cannon-Brookes; and Chief Financial Officer, James Chuong. Earlier today, we published a shareholder letter and press release with our financial results and commentary for our third quarter of fiscal year 2026. The shareholder letter is available on the Investor Relations section of our website where you will also find our other earnings-related materials, including the earnings press release and supplemental investor data sheet. As always, our shareholder letter contains management's insight and commentary for the quarter. So during the call today, we'll have brief opening remarks and then focus our time on Q&A. This call will include forward-looking statements. Forward-looking statements involve known and unknown risks, uncertainties and assumptions. If any such risks or uncertainties materialize or if any of the assumptions prove incorrect, our results could differ materially from the results expressed or implied by the forward-looking statements we make. You should not rely upon forward-looking statements as predictions of future events. Forward-looking statements represent our management's beliefs and assumptions only as of the date such statements are made. And we undertake no obligation to update or revise such statements should they change or seek to be current. Further information on these and other factors that could affect our business performance and financial results is included in filings we make with the Securities and Exchange Commission from time to time, including the section titled Risk Factors and our most recently filed annual and quarterly reports. During today's call, we will also discuss non-GAAP financial measures. These non-GAAP financial measures are in addition to and are not a substitute for or superior to measures of financial performance prepared in accordance with GAAP. A reconciliation between GAAP and non-GAAP financial measures is available in our shareholder letter, earnings release and investor data sheet on the Investor Relations section of our website. We'd like to allow as many of you to participate in Q&A as possible. So out of respect for others on the call, we'll take one question at a time. With that, I'll turn the call over to Mike for opening remarks.
Thank you all for joining us today. As you've already read in our shareholder letter, we delivered some incredible Q3 results. Total revenue grew 32% year-over-year to $1.8 billion. Cloud revenue surpassed $1.1 billion and accelerated to 29% growth year-over-year, and RPO grew again 37% year-over-year to $4 billion. These are likely thanks to our team's excellent execution and clear momentum across our key strategic priorities: enterprise, AI and the system of work. This quarter, some of the world's largest enterprises, including Siemens Energy, Rheinmetall and Wayfair deepened and broadened their commitments to Atlassian. In AI, we continue to add millions of monthly active users to Rovo and our AI Rovo credit usage is growing more than 20% month-over-month. Customers using Rovo are also growing their ARR at roughly 2x the rate of customers who are not using Rovo, contributing to our strong cloud outperformance and expansion in the quarter. And more and more enterprises are embracing our platform-wide vision using Atlassian system of work to see the full picture of their organization. This is because the Teamwork Graph connects knowledge, work, people and code, giving our customers one of the richest enterprise context graphs in the world. Context is a clear differentiator for us. And we're saying this is our competitive momentum built. This is our largest ever quarter for competitive displacements from a major ITSM provider. We're taking it from rivals as customers move away from legacy systems and choose Atlassian for a more modern AI-native and much better value service platform. As I've said before, I believe AI is one of the best things that has ever happened to Atlassian. In a world where humans will run teams of agents, context is the only anchor to avoid chaos. And we believe the companies that prioritize context will come truly AI native. With Atlassian, our customers aren't just choosing software, they're choosing the kind of company that they want to become. This is a time of significant change in our industry, and we're moving forward with strong conviction and discipline. We're focused on executing, delivering customer value and driving durable profitable growth. With that, I'll pass the call to the operator for Q&A.
Questions and answers
Operator provided instructions on how to ask questions. Your first question comes from Arjun Bhatia from William Blair. Your first question comes from Keith Weiss from Morgan Stanley.
And congratulations on a really solid Q3 print. It's great to see all these investments in AI and innovation really starting to come to fruition within the numbers. It's important in getting investors more confident in the stock. I think another important part is helping people better understand how the existing software that Atlassian brings to the equation plus AI brings a better result. There's one line in the shareholder letter that I thought was really interesting. When you're talking about the Teamwork Graph and how it makes the AI investments not just smarter, we've been talking a lot about context and how it makes the AI better, but you're also talking about cheaper and more valuable. So one, I was hoping you could dig into that and how the Teamwork Graph and the broader system lowers the cost of these AI investments, particularly as we're hearing more pushback on these credit costs which are really starting to rise and the token costs becoming large? And then maybe a follow-up for James. In the shareholder letter, you mentioned data center outperformance driven by some pull forward of deals from future quarters. Any view you can give us into what that means for FY '27 and what we should be expecting from data center in the year ahead?
Keith, sure. Thanks for the question. Great, very savvy question. Look, the Teamwork Graph and the Atlassian platform is certainly delivering amazing results to customers. You can see that customers using Rovo are growing their ARR at twice the rate of non-Rovo customers. Their credit usage continues to grow strongly, with results of over 20% month-on-month. And they're upgrading the Teamwork Collection to get more of those credits included in the base offering, but also using many more agents. That agent usage is what you're referring to, whether those agents are Rovo agents or other platforms' agents that are accessing Atlassian's context with the Teamwork Graph; that usage is increasing markedly. That's the compounding effect of intelligence for customers as models continue to get better. But to really accelerate a business, the intelligence compounding is only one aspect. What you also need is the context. That is your knowledge, your work and projects and goals, understanding of your people, so your org chart, their skills and everything else, as well as knowledge of the code. We have a lot of huge announcements coming up next week at Team '26 around this area, but what you're seeing in the Teamwork Graph is the world's best context graph across all aspects of the business. So whether that's a service team, a marketing team, a technology team or a business team, bringing that context to bear in all of those AI surfaces is what's most important. Now when we say it makes things better, faster and cheaper: why is that? Well, we have a lot of statistics and proof points that not only do you get higher quality AI answers because of our search, the Teamwork Graph, everything put together in the knowledge that we have about your business, but you also get cheaper answers. Those are cheaper answers because you use far fewer tokens to get to an answer in the same amount of time, fundamentally using fewer tokens reduces your cost of AI or allows you to do far more AI work, whichever way you look at it. So customers are seeing that. You're seeing that in the usage and then showing up in our financial results. James, do you want to follow on with the second half?
Yes, Keith, thanks for the question. On the data center side, the Q3 revenue beat, as we mentioned, was primarily driven by recognizing greater-than-expected upfront term license revenue within that quarter. Since our announcement of the data center end of life back in September, we've had a couple of quarters now to better understand some of the signals that we're seeing from our customers in terms of their buying behaviors, especially Q3 being the largest expiry base for us. So let me unpack that a little bit more. First, the migration to the cloud is on track and continues as expected. We still expect that to contribute mid- to high single digits on cloud growth. Second, retention rates on our data center business remain incredibly robust and actually outperformed expectations in the quarter. Third, for some of our largest customers with more complex migration, they remain committed to transitioning to the cloud, but it's going to be a multiyear journey for them. They've got deep customizations and change management; it will take time. Often many of these have tens of thousands of users, some with over 100,000 users. So this category of customers, we saw a pull forward of purchasing and expansion activity into Q3 from future periods. We also had a pricing change in March that further catalyzed this dynamic. As a result, that drove greater-than-expected upfront term license revenue recognized in the quarter. In fact, relative to our expectations for Q3, we recognized approximately $50 million more in upfront term license revenue. Those are trends we've been seeing since our announcement, but more pronounced in Q3 given the size of the expiry base and the pricing catalysts I mentioned. Lastly, the cohort of data center customers that are actively planning and transitioning to the cloud are moderating their seat expansion versus historical trends. Again, retention remains incredibly high, but we now expect a more muted level of data center expansion from these customers going forward as they try to move to the cloud. We're still seeing nice uplift when customers move from data center to cloud. Net-net, our largest strategic customers continue to deepen their commitment to Atlassian, whether that's on data center or cloud, and we're working hard to meet them where they are and help them accelerate that transition so they can unlock all the AI and agent capabilities in the cloud. With these dynamics across the year on data center, Q4 is yet to play out where revenue recognition is being pulled into FY '26 from FY '27. We recognize that there's lumpiness in that pull-forward effect in data center and timing impacts reported revenue, RPO and CRPO. Internally, of course, we look at a variety of metrics to manage our business, including a healthy ARR. Next week at our Team '26 Conference, we're going to hold an investor forum to help guide investors through the revenue recognition timing dynamics on the data center side. We'll look to enhance our disclosures and share historical subscription ARR, which will help normalize some of those timing effects and help everyone better understand the underlying strength of the overall business. All up, we feel really good about our execution and the runway that we still have ahead of us. More to share next week at the TEAM event and the investor forum.
Your next question comes from Arjun Bhatia from William Blair.
Sorry about that. But congrats on the strong quarter here. I was curious on just Rovo, how you're thinking about positioning that against some of the third-party agents. It seems like you're having a lot of success in your existing customer base. I'm curious, are customers evaluating other agents against Rovo? Or is this sort of an easy add-on given how integrated it is into the rest of the platform with Jira and Confluence and JSM and the rest of the suite?
Thanks, Arjun. There are a lot of places that customers can access Rovo. Think of Rovo as the AI part of the Atlassian platform that shows up in all of our surfaces, whether those surfaces are on the Atlassian platform inside of Jira or Confluence, in our chat app from mobile and desktop, or whether those surfaces are in other agent platforms from Google, Salesforce or any of the foundation model vendors. We want to make sure that Atlassian's workflows, processes and the Teamwork Graph show up wherever is most relevant for the customer. This has required a lot of R&D to ensure that our context graph in the Teamwork Graph is the best out there, the most deeply connected and with the most organizational context. We do the inference upfront to make sure you get those better, cheaper and faster results that we talked about: better quality answers at lower cost and running agents faster. That's an amazing offering to customers, and they're realizing that. Whether that happens on the Atlassian platform or off it, we want the customers to see value in the platform overall. Agents existing natively in our automation framework have the largest access to the platform and appear directly in the sidebar and Jira UI; that's a huge advantage. At the same time, we've shipped features that allow third-party agents from Gamma and Canva to Cursor and Claude to be used in each of our different types of teams. We want third-party agents to surface in Atlassian's context, whether that's on a Confluence whiteboard or a Jira work item; they all use the Teamwork Graph at the core. Customers are seeing that. We certainly get evaluated against other platforms, but customer reaction is strong. We've done a phenomenal amount of R&D to give great quality answers. We see that in increasing usage of our Rovo platform on and off Atlassian, both of which benefit us, and you can see that in customers' expansion rates as they use our AI technology. All AI is not built equal. We build fantastic AI, and we get it into customers' hands. That's what's most important.
Your next question comes from Gregg Moskowitz from Mizuho.
James, welcome to Atlassian. Mike, you may recall my high level of frustration one quarter ago when after I thought it was a pretty good quarter of acceleration, your shares continued to trade materially lower. I don't want to minimize that there's more work to be done. But clearly, this is an impressive result. My question relates to a comment in the shareholder letter that strong seat expansion in Jira was a key driver of the cloud revenue acceleration this quarter. Given that there is so much fear about meaningful seat compression at Atlassian being on the horizon, can you unpack the drivers of the seat growth for us? Secondly, is this a dynamic you think can be durable?
Gregg, thanks for the question. On the cloud side, we saw performance reaccelerating to 29% year-on-year. Migrations from data center to cloud were in line with our expectations and were not the primary contributor to that $50 million beat. They are progressing well and still expected to contribute mid- to high single-digit growth to cloud. The two primary drivers for the outperformance were cross-sell and seat expansion. On the cross-sell side, we saw outperformance in our collections business across Service Collection and, in particular, Teamwork Collection. We have Jira, Confluence, Loom and Rovo. Teamwork Collection is the best vehicle for customers to buy and unlock AI and agent capabilities across the Atlassian platform. Customers are upgrading to Teamwork Collection because of the increased AI credits — we're giving 10x more credits on Rovo versus the stand-alone subscription. Mike can touch on the progress we're seeing there. Importantly, we're also seeing growth in Teamwork Collection while also seeing continued seat expansion in our core Jira stand-alone offering. That speaks to how an AI-driven world keeps Jira and the Atlassian platform core to managing workflows and collaboration to fully unlock AI value. We're launching a lot of value that enables customers to deploy agents to do the work and to capture that agent activity alongside work history, permissions and audit trails and admin governance. So there's a lot of traction here, as shown in our Q3 print.
Yes, Gregg, thanks for calling that out. I hope we've been consistent: we are not seeing any signal of seat compression from customers. If anything, we're seeing the opposite. We are seeing strong expansion numbers, strong cross-sell between collections, strong usage of AI and strong commitment to the Atlassian platform. Many competitive wins, a huge amount of consolidation into the Teamwork Collection. We have a lot of green lights in different places. NRR maintained north of 120% and even ticked up again for the third or fourth quarter in a row. There are a few reasons for that. First, high R&D investment yielding great quality software — that matters. We build amazing applications that deliver great value. Second, the context in the Teamwork Graph and the critical business processes, powered by AI, blur boundaries between teams and roles. That means our platform can expand into nontechnical roles at an increasing rate, which explains seat expansion in different collections and areas. Fundamentally, customers are opting for more workflows on the Atlassian platform.
Your next question comes from Brent Thill from Jefferies.
Mike, you called out the largest competitive replacements. What are you seeing? What's driving this now where you're seeing an increasing rate of displacement?
Thanks, Brent. We had a great quarter for competitive displacements, especially in the Service Collection. We continue to be incredibly strong in the mid-market, and as a result of many years of investment in our enterprise pillars, we are now going strongly into enterprise and strategic segments across service management in particular. That's not just ITSM, although in ITSM we are growing strongly; it's broader employee service management. We get the statistic that 75% of the Fortune 500 uses Service Collection, and 60% of Service Collection customers use us outside of IT in HR, marketing and other areas. This is a fantastic example of why we're getting those competitive displacements. It was our largest quarter ever for those. Again, it's the quality of the software, the ease with which you can get up and running on Service Collection, the high user experience quality, the comprehensiveness of our data and the Teamwork Graph bringing that to Service Collection. That allows you to operate services cheaper, quicker and with better answers because we have access to a better knowledge graph across the organization. AI continues to win across the board. We're very AI-forward; it's one of the largest areas of usage of agents in automation and automated workflows because it allows service teams to run more quickly. We're only just getting started in customer service and had a great quarter there as well. Asset management is moving into the platform as part of the overall Teamwork Graph. This is an incredibly strong customer story. We're excited about taking more share in that space. $1 billion of ARR for Service Collection is a milestone worth celebrating.
Your next question comes from Allan Verkhovski from BTIG.
Mike, it's great to see the AI momentum here. I'd be curious if you could share what drove the decision to announce the data collection changes you're making? And what are you looking forward to from a product capability perspective on the other side of it?
Allan, Atlassian is driven by its values and long-term philosophy. Our data collection changes clarify exactly how we use customer data and what the data categories are and where they are used. We have a world-class policy and are very clear about the categories of data, where they are used, what benefit they provide to the customer, and what options customers have. Think of it as a large clarification of what happens at different points and the value delivered to the customer. The increasing usage of AI allows us to build more powerful features. For example, the DX business benefits from seeing how an engineering organization compares to others in the industry and size; that requires some of these changes. Usage patterns of different SaaS tools are how we build the Teamwork Graph. It's about being an open company, being clear with customers what we collect, what the advantage is and the trade-offs. We've had a positive customer response because we put trust and openness at the core of that relationship and explain to them what they get from it. Those usage patterns are important to building fantastic software, which is our highest priority.
Your next question comes from Raimo Lenschow from Barclays.
In your letter, you talked a lot about momentum in ITSM. Can you talk a little bit about what you're seeing there? What's driving it? And how meaningful is this for you?
Raimo, we are seeing great momentum in the Service Collection, as I mentioned earlier, and we're celebrating passing $1 billion in ARR there. We try to put a different focus in the shareholder letter each quarter. Last quarter we focused on Teamwork Collection when it passed 1 million seats and 1,000 customers less than six months into that offering. This quarter we focused on Service Collection because of the milestone. Strength is across regions; we had a great quarter in EMEA and a lot of large wins in Europe, Middle East and Asia with increasing strategic and enterprise customers. We're seeing strength in non-IT use cases as well, which blurs roles — one tool for IT, one tool for employees and HR is less necessary when you can connect teams across the organization. That is powerful as organizations become increasingly service driven. We also have many AI features delivering value from AI ops in IT to using Rovo broadly. We're seeing increased adoption of our MCP servers and CLI tools. We'll talk more next week, but especially in Service Collection, enabling customers to get access to contact craft built into their service offerings produces fantastic results, a better value proposition and allows service teams to execute more quickly and at lower cost.
Your next question comes from Alex Zukin from Wolfe Research.
This is Arsenije on for Alex. So Mike, what is working best with customers when adopting Service and Teamwork Collections that's driving stronger cross-sell growth contribution in cloud? And then a brief follow-up for James: you mentioned the data center revenue growth deceleration comment — can you give more color on how data center ARR is trending? Or clarify whether we'll get any DC ARR figure exiting the year to better understand growth when we lap tougher comparisons next year?
Arsenije, everything is working really well. The Teamwork Graph and our data are central, as is the speed of adoption and the user experience. We have customers with 500-plus different service desks in an organization, and the ability to get new service desks up and running with organizational data creates efficient offerings for finance, operations, workplace and HR teams. Our investment in user experience is paying off. We continue to do strongly in HRSM and service management around HR and other business functions. This continues to be a source of strength for Service Collection. Our traditional connectivity between dev and IT teams, between technology teams and business teams, has always been a source of strength for Jira Service Management and has deepened with Service Collection because both teams are becoming more AI driven. A single context graph and AI offering across products lets customers take that offering into other tools they use, making service-driven parts of the business quicker and faster to operate. We're seeing strength across the board. Internally, our customer service management adoption has hit more than 70% AI resolution rates across hundreds of thousands of conversations, allowing us to run more efficiently. Customers are seeing that too as the customer service offering rolls out with a strong set of features.
Arsenije, as it relates to FY '27, it's too early to discuss guidance at this point; we'll provide that in August with our Q4 earnings. As it relates to ARR, we're seeing lumpiness in revenue recognition on the data center side. Next week we'll share historical subscription ARR across the overall business to help smooth those timing effects. That should give a better understanding of underlying strength. As a reminder, for the data center end of life announcement, we began to recognize greater upfront term license revenue which results in greater upfront revenue recognition in the period, but there's a corresponding drawdown in RPO and CRPO. When you normalize for the impact of ASC 606, our RPO would have been north of 40% in the quarter and our CRPO would have been north of 30% year-over-year in Q3, much more in line with recent trends and underscoring strength in the backlog we're building.
Your next question comes from Fatima Boolani from Citi.
I wanted to ask about diversifying some of your pricing strategy. Collections has been a huge step consolidating adjacent capabilities into a more intuitive selling motion. Many peers in enterprise software are investigating or testing usage-based pricing. I'm curious what you think about that approach and particularly how pertinent it could be for the Service Collection. To the extent you're A/B testing any of this with certain products, I'd love that perspective. Quick one for James: there's been a lot of focus on driving efficiencies and leveraging AI internally to make Atlassian more efficient. What qualitative learnings and quantifiable yields are you seeing as you deploy AI internally?
Fatima, on pricing: our philosophy has always been to meet customers where they're at. Customers generally like our pricing and we will continue to be customer-led. Today the largest amount of value is delivered through seat-based pricing. Collections have transformed how we do that, giving broader value. When people move to Teamwork Collection, which has great momentum, they get software value across Jira, Confluence, Loom and platform assets and increased Rovo credit allowance. Teamwork Collection customers use more than twice as many Rovo credits per user and have more than twice as many active agents. We want to build features that use those credits so customers see value. We also have a series of consumption or usage-based pricing meters — over 10 or 12 meters, from assets to customer service index subjects, extra Rovo credits, Forge extensibility and Bitbucket Pipelines. We continue to be customer-led in pricing as customers consume our AI offerings and grow. Token usage growth of 20% month-on-month is an achievement and shows the value delivered. Fundamentally, it's about selling outcomes to customers and understanding value. You see customers increasing their length of commitment and dollar-based commitment, reflected in strong RPO growth. Normalized for ASC 606, RPO north of 40% shows customers voting for the long term for the Atlassian platform and our pricing models continuing to adapt to their needs.
Fatima, on margin expansion and efficiencies, we're in a unique opportunity where we're seeing strong demand signals and will continue to reinvest in AI and enterprise sales where we see opportunity while balancing a disciplined fiscal approach. We elevated driving durable, profitable growth as a strategic priority alongside AI, enterprise and system of work. Margin expansion will come through efficiencies and continued top-line growth as we drive value for customers.
Fatima, to add: we've seen our engineering investments pay off in terms of platform efficiency. Our cost of operating the platform continues to improve even as we scale with larger customers expanding. We're running the platform more cheaply without reliability hiccups, which is a credit to our engineering team. There's an R&D story and a finance story here — we feel strong about our ability to deliver durable profitable growth in the future.
Thank you. That's all the questions we have time for today. I will now turn the call over to Mike for some closing remarks.
Thank you all for joining us on the call today. Thanks to the Atlassian team for a fantastic quarter. As always, we appreciate the thoughtful questions. I believe one of our long-time friends, Keith Weiss, is retiring after this call. So Keith, thank you very much for all the questions over time in person and virtually. We appreciate your thoughtful questions, especially today. To everyone else, hopefully, Keith, you will join us next week in Anaheim for Team '26. We have a series of incredibly exciting announcements as well as an investor forum. Whether we see you online or in-person in Anaheim, we'll see you next week. Otherwise, hope you have a kickass weekend.