noon. We are putting a close to day two of Citi's TMT conference today. I am Fatima Boolani. I jointly head up our software equity research team here, and I am very happy to end the day with James Chuong, CFO of Atlassian. Thank you so much for being here and joining me.
Yeah, thanks for having me, Fatima.
Well, before we get into kind of the meat and potatoes of our discussion, five and a half months in the seat, almost six. Biggest surprises, James, in the last five months as CFO of Atlassian?
Yeah. Five and a half months, so I must have all the answers at this point. In terms of biggest surprises, honestly, I would say that a lot of it is more about reinforcing the thesis as I came into Atlassian. When we really thought about the opportunity that AI presents, I think there was a lot of questions, a lot of noise in the market about whether that was going to be a headwind or a tailwind for Atlassian. I think we have proven over the last couple of quarters, and really even before that as well, with our strategy, that AI is going to be a tailwind for Atlassian. As customers look to really drive more and more value out of AI, understanding that Atlassian delivers an incredible graph, our Teamwork Graph.
That's what's really going to be powering AI across the enterprise, and I'm excited to get into that today. But again, I would say not a lot of major surprises, but pleasantly surprised at the reaction that we're seeing right now over the last couple of quarters.
The other question I did want to ask you is, areas that you feel you've been able to influence the most change. I know you talked a lot about, hey, there is a little bit of an evolution. This was just tweaking the machinery of metrics, guidance philosophy, and things of that nature. But any areas that you feel you've influenced a kind of change that's worthwhile shedding some light on with investors?
Yeah. Like we said, the strategy largely hasn't changed, right? When we talk about the strategic priorities at the company, it starts with AI, system of work, enterprise, and I think one of the areas that Mike and I added was around driving durable, profitable growth. So all of that's continuing to work together and come together over the last couple of quarters here. So I think that's one of the areas I would highlight.
Good segue in the next question I wanted to ask you. A year in review. You just ended fiscal 2026, reported results about six weeks ago. Just two to three most important highlights that are worthwhile for folks to kind of refresh on.
Yeah. Q4 is actually a great place to start with that, right? We posted $1.8 billion in revenue growth, 28% year-over-year. We posted ARR growth of 23% year-over-year, with RPO growing 44% year-on-year. We got a question around what changed overnight, what changed in the quarter. The reality is it was really a lot of consistent innovation and execution across our product teams, our engineering teams, our go-to-market teams, to really deliver and drive those types of results. Two of the areas that really outperformed our expectations was standalone seat growth in our product like Jira, Confluence, and JSM. We saw revenue re-accelerate in our Service Collection in the quarter as well. A lot of opportunities still ahead of us here.
In terms of highlights, I shared a couple of these stats a bit earlier here, but in terms of AI being a driver for us, the growth that we continue to see across our Teamwork Collection. One of the things that has made Atlassian incredibly special over the last two-plus decades is we put the customer first. We want to make sure that we're driving value, number one, and we're seeing that Teamwork Collection is the best way for customers to buy AI with us. Give out 10x Rovo credits, mobile credits, as part of that, and we want to remove that friction of that ability for customers to adopt and use AI with more and more depth. We're beginning to see that 80% of the Fortune 500 use Rovo right now.
As we continue to focus on the depth side of it, we're also seeing Rovo-assisted actions increase 50% quarter-over-quarter. That's really important because it's not just a simple AI search, it's not simple AI chat. Users asking Rovo to take action. That's how you begin to unlock the power of AI across teams and across the board.
James, I want to focus the conversation on the cloud ARR and the cloud revenue. There's been a pretty visible acceleration in your cloud growth as one of the things that we've seen in the last four quarters, cloud revenue growth growing from 26% to 31% in the fourth quarter. Can you unpack for us some of the factors that we should expect to be repeatable and durable in terms of growth drivers for fiscal 2027, and then conversely, factors in inputs that will lap out or aren't necessarily going to be durable inputs to fiscal 2027 cloud revenue growth?
Yeah. Really pleased with the momentum that we continue to see on the cloud side of the world. Some of our biggest drivers remains expansion, as well as that cross-sell motion that we talked about. I know there was some concern about cannibalization as it relates to our collection business, but as we saw in the last couple of quarters, the outperformance was driven by standalone seat growth that we saw in our applications as well as continued growth in Teamwork Collection. Teamwork Collection is still relatively early. We only introduced that over the last year, so I believe that's going to continue to compound. If you look at also the Data Center side, right? As we continue to migrate customers over to the cloud, that continues to be a multi-year expansion pipeline.
When we look at cohorts of customers that have since moved from Data Center to cloud, three years on, they've spent 1.75x on average more than before. That makes sense because you've got a whole suite of applications and collections that's available in the cloud. You've got Rovo, you've got AI, you've got the Teamwork Graph. A ton of opportunity to drive more value as customers continue to migrate to the cloud.
You brought this up 2 times now, and I think it's really worth expanding and expounding on this notion, seat growth. I think there's been a pervasive fear, or there had been anyway, pervasive fear in the investor community around, there is a seat-based monetization orientation to the business. The frontier labs and a lot of the innovation out of the labs had been perceived or misperceived in cannibalizing your opportunity by way of, "Hey, you're not going to need to hire as many developers or software engineers." In retrospect, that has not turned out to be true. I want to ask you, what are your current observations in your conversations with customers and customer behavior around just headcount growth, technical headcount growth, just so we can put that matter to bed.
Yeah.
Why or why not should we see continued growth in software developer headcount and engineering headcount?
Yeah. I think as you see the intelligence in the models continue to grow and make progress and compound over time, and also seeing how competitive it is becoming from a pricing perspective, that is going to benefit the market, right? There is more value being driven to the market, to our customers, that is going to benefit Atlassian as well. In terms of the seat-based growth, what that really means is the cost of coding, cost of building is coming down.
We are seeing more innovation in our customers. We are seeing more software being built. We are seeing more seats being added, and that is true across developers as well as non-developers. I think one of the things that is most underappreciated about Atlassian is how diversified the set of business is. I came into Atlassian sort of thinking that it was a tool-based company simply for software engineers, called Jira.
The reality is 65% of Jira users are knowledge workers, non-software engineers, right? Confluence is even higher, closer to 70%, and JSM Service Collection, north of 70%. So incredibly diversified base of business. I think as customers really think about how they are going to get value out of AI, it is one thing to have a point solution product that is going to drive individual productivity. How do you actually get value and productivity across teams? Well, that is exactly where Atlassian sits, right? We have spent 20+ years helping teams collaborate, orchestrate, and that is becoming more important now, not just with people, but through people and agents together. As customers realize that they are adopting more of our Teamwork Collection, they want to be able to go wall to wall with Atlassian, to be able to unlock the value of the Teamwork Graph.
Because the more teams and more individuals that they can get onto the Atlassian platform, the more context that they are creating, right? The workflows that are being built across Jira Service Collection, and all the context that is being captured in Confluence pages that are attached to those workflows, that is really creating the intelligence and the IP within the enterprise that they can begin to unlock.
Historically, our frame of reference in terms of product exposure was tool-based, right? So Jira and Confluence, JSM, and then kind of everything else. So Jira Work.
Yeah.
The artist formerly known as Jira Work. I know there has been a pretty meaningful reconstitution of a lot of those capabilities, especially from the standpoint of looking at your portfolio in the cloud. That has manifested in the effort to build collections, right? So Teamwork and Service. Can you give us a little bit of a refresher on how many collections you have today? What type of KPIs and metrics can you share around levels of penetration with Teamwork? Maybe the type of ACV or ASP uplift you can realize when a singular Jira user goes on to add Confluence, goes on to add other capabilities.
Broadly, we have five collections across the company. Teamwork Collection, Service Collection, Software Collection, Product Collection as well. If you think about the collection space, it is really about job speed across different parts of the organization. Jira is our largest business. It is a $2.5 billion ARR business with roughly 150,000 customers today. You have our Service Collection business, which just crossed $1 billion in ARR, growing north of 30% year- over- year. That is one of our largest and fastest-growing enterprise wedges today across the Atlassian platform. Confluence, we rarely talk about Confluence, but that is a $1.5 billion business with 100,000 customers. When I think about the footprint of 150,000 customers in Jira, and 65,000 customers in Service Collection, there is a massive opportunity to cross-sell. We are still so early stages on the enterprise side of the world.
When you think about our customer base right now, our $3 million ARR cohort is growing 50% year- over- year, about 164 customers, right? Our $5 million ARR cohort, that is growing 70% year- over- year. That is roughly 69 customers. Think about that against a base $6.6 billion in ARR. There is a ton of headroom to continue to grow on that enterprise side of the world and the cross-sell and expansion opportunity there.
One of the other meaningful efforts beyond the reconstitution and the packaging has been around the go-to-market motion. Historically, we have all been familiar with- You were kind of selling itself.
It was a product-led motion, right? There has been a graduation into this more classic enterprise-
Yeah
focused sales approach. You've had a new CRO, I think, who's now been in the seat for the past two years.
Yeah, almost two.
What have been some of the most dramatic outputs from the changes that have been instituted in the go-to-market process and institution, and where should we continue to see that envelope move, and where should that show the most yield?
Yeah. Brian, who joined about a year and a half ago, Brian Duffy, our CRO, he brings 18 years of experience from SAP. Atlassian being the heritage was product-led growth, the product selling itself. There was a ton of opportunity to continue to evolve and mature that go-to-market enterprise motion. We're already beginning to. It wasn't that long ago, really two, three years, that we only had about 115 reps selling. We're continuing to grow that towards 400. Even then, I think that's well behind the demand that we're seeing in the market right now. You're seeing that in terms of the capacity that we're building out to be able to deliver against that demand across our customer base.
You're also seeing the evolution across our partner programs as well to make sure that we're moving from what historically had been more volume- discount tiers into a much more strategically aligned approach with the Atlassian priorities, which is to drive cloud adoption, and over time, more and more, drive AI adoption.
Part and parcel to the relationship that you've had with partners, historically, that's involved the Data Center business, and the migration thereof. As we enter 2028, as we enter 2029, those are the stated milestone years you've put on the board where there's going to be end of support, end of life. Just by extension of that, how does that change some of your relationships in the partner community for whom the migration business is inevitably going to dry up? How should investors generally think about you and partners' ability to recapture and actually expand the value on the back of those migrations to build more enduring cloud practices?
Yeah. I think this really speaks to the evolution of that partner program. They're an incredibly important part of our ecosystem here. As customers, especially the ones that remain on Data Center, they're some of our largest, most committed customers, but they also have some of the largest customizations across the platform as well. It's going to take deep partnership with our solutions partners, with GSIs globally to make sure that we help our customers change management as they migrate over to the cloud. That's really one aspect of it, and they remain incredibly important part of our ecosystem. The other piece is the marketplace. Many of these customers on the Data Center side rely on the marketplace vendors, the third party parts of our ecosystem that drive tremendous value for both Atlassian as well as our customers.
And so making sure that we move and incentivize these marketplace providers, not just to build on the Legacy Connect system, but on our next generation Forge platform. And so that our Data Center customers are able to take advantage of those capabilities as they move to the cloud. Whether it's the marketplace, whether it's the solution providers, they're an incredible part, an important part of our ecosystem. That's an area that we're
James, as you migrate this last cohort of Data Center customers over by your commentary, they're more complex by nature and by design. That's why they live in Data Center right now, and they didn't move in the prior migration iteration going from server. How should we think about the handholding process from you? And then relatedly, what their wallet potential looks like in the cloud and how could that be expressed from a net retention rate perspective potential?
Yeah, it's a great question. For our Data Center customers, they really do know that cloud is ultimately the destination, right? Today, they have a limited set of products and capabilities when it's on Data Center. They've built robust, incredible capabilities around that, but they know that cloud is ultimately the destination. That's where you can unlock all of our different applications. You can unlock our collections. You can unlock AI, Rovo, and Teamwork Graph. None of that's available on our Data Center right now. So, as customers continue to move towards the cloud, we want to make sure that we partner deeply with them over these next couple of years. And once they get to the cloud, I shared a little bit earlier, three years on, they're increasing their spend by 1.75x. That's on average.
But if you look at even our NRR on the cloud revenue side, I think that's indicative, right? It's north of 120% year-over-year growth right now on cloud NRR. And that's indicative, I think, of the opportunity that we have in front of us. I mentioned marketplace a little bit earlier. We're continuing to incentivize marketplace providers moving to the Forge, and building on that platform, by making sure that we've got some very attractive rates across that. Again, I think that'll be over time, a tailwind for the business.
On the AI monetization side, we talked about this a little bit earlier, but I think it's worthwhile getting into a little bit more nuanced conversation about. I generally feel that there is some misunderstanding or some noise around what your opportunity set looks like and your relevance looks like in an AI first world. We talked a little bit about Rovo, so maybe you'll take me there. How do you, as Atlassian, intend to monetize AI over time? What products will be the financial manifestation of that success that you're actually seeing in AI, such that you are an AI winner?
The number one way that our customers want to buy AI today from Atlassian and deploy AI is through our collection, Teamwork Collection. That's very intentional. As we bring together Teamwork Collection, which is Confluence, Loom, and Jira, we also provide 10x more Rovo credits. If you look in the market today, whether it's our standalone apps or in our collection, the amount of credits that we allocate out is fairly generous. That is to reduce the friction as it relates to adoption and usage across our customer . This really gets back to, number one, keeping the customer first, number two, driving value, and ultimately making sure that we're being really thoughtful about the pricing advantage that we occupy today to be able to take share in this. If you look at Jira today, it's roughly $7 per user.
That'd be some of the best B2B enterprise value in the market today. So that gives us a lot of opportunity from a headroom perspective. If you put yourself in the customer's shoes, they're having to think about the cost of seats, and they're having to think about the cost of tokens and agent deploy. But over time, more and more, they're going to have to think about that holistically. You think about the pricing advantage that Atlassian has, bringing together the seats, the AI agentic value that we're delivering today, we're already starting to see that, and our customers are reacting, and you're seeing that in our results.
Just to be clear, the $7 per seat for Jira, is that after the compounding price increases that you've put into place in the last couple of years?
Yeah. We always think about driving value for our. To the extent that we are continuing to drive more incremental value for our customers, we want to make sure our pricing also reflects that. I think when you look at the cloud NRR growth of north of 120%, I think that is indicative of the value that we are driving and how sticky the value is across our entire.
There has also been some consideration around your ability to take price in the last several years. That is part and parcel to it being an idiosyncratic dynamic to you in that you had, five years ago, customers on server, and you were using the carrot and stick approach to get, mostly carrot approach, to get folks onto the cloud, right? So there were pricing increases pursuant to that. Same thing with the Data Center migration journey, right? Again, the pricing increases have been somewhat idiosyncratic to you versus your ability to just generally take price. I want to ask you, what do you feel is your scope and ability to take more price in the install base?
Because I think, as a financial analyst, I see, on average, it has been 15%-20% price increases for the Data Center customer in the last couple of years, and those numbers can get big after compounding for five years. Right? So, does that still leave you with optionality on price?
Yeah. Listen, I think when we go back to the core values of driving value for our customers, some of the ARPU, actually a large part of the ARPU lift that we see across our business is actually not necessarily driven by those list prices that we are talking about here. It is customers opting to expand opportunity sets across Atlassian. It is customers increasing their seat count. It is upgrading their editions from premium to enterprise. It is actually going from single SKUs to collections and then multiple collections over time. I think that, again, gives us a lot of permission and room from a pricing perspective. We also think this is a moment to take share.
As more and more customers think about how to get value out of AI, it is to make sure that their employee base are coming onto these platforms and systems of work and systems of record. If you look across large enterprises today, it is typically not 20, 22, 25 massive platforms of scale and ecosystems that they operate in, nor is it one. But it is typically three or five, these ecosystems. Atlassian is one of those ecosystems. As more customers put their employee base, whether it is software developers, whether it is HR teams, finance teams, sales teams, onto Jira, Confluence, Loom, and more and more collections like Service Collection, we are seeing more AI value. It is not just me sitting up here saying that. If you look at the stats that we have shared, Teamwork Collection customers are driving 2x more agent development. They are deploying 2x more agents. They are using 2x more Rovo credits.
Customers that use Rovo, we are seeing their ARR growth grow 2x higher than like-for-like customers that don't use Rovo. We are seeing that broad-based strategy of driving breadth of adoption, depth of use, and over time being able to monetize that.
We talked a lot about seat count, but we actually talked about human seat count. What I want to ask you is, as more agentic systems take on the roles and personas of a human user to undertake tasks and touch systems, they would necessarily need to knock on the door of Jira for tracking and workflow information and Confluence from a Kanban standpoint. It might be early days, but how are you thinking about agents touching your system and the Teamwork Graph, and how are you monetizing that? Because as far as the eye can see, agents are going to be multiplicative, if not orders of magnitude more in volume than human users and seats. What is kind of the philosophy and thought process there, and what are you seeing customers do there?
That's a great question. When we think about Rovo and the Teamwork Graph and customers using agents and agentic use cases across the Atlassian platform, we see a couple things. I shared some of the stats a little bit earlier, but Rovo-assisted actions are up 50% quarter-over-quarter. What does that mean? That means folks are actually using tasks, using Rovo and the Atlassian platform to complete tasks and drive more productivity, one. If you think about also the Service Collection side of the world, we've seen agentic automations grow 3x over the last six months in our Service Collection business. If you also think about what has made Atlassian incredibly valuable to enterprises, we're an open system, right? It's not just the Atlassian platform, it's all the APIs and connectors into the other platforms and ecosystems that I mentioned a little bit earlier.
When we look at MCP servers and CLI, the monthly active users has crossed 1 million right now. We're seeing MCP calls grow 400% quarter-over-quarter, right? MCP-driven Jira growth, these are items in Jira, new pages being created in Confluence. That's going 4x right now. That's important because what we're seeing is as we open up the platform, customers are taking advantage of the Teamwork Graph. We're seeing write back into the system. It's not just an extraction, right? Last week we also announced that we will begin to enforce consumptions on those meters starting in December. In large part, customers continue to buy AI through the collection because of the generous credits that we provide. It also puts into a mechanism in place for us to make sure that we're driving value for our customers, but also sharing in those.
I wanted to shift gears to Service Collection. You talked about JSM Service Collections being a billion-dollar product franchise for you. How should we think about some of the technological debate brewing in ITSM, where agents are on track and on pace to take end-to-end action, right? How does that impact maybe the pricing model? Are customers clamoring more for usage and outcomes-based modality there? How does that maybe create opportunities but also challenges for the growth trajectory of the Service Collections franchise?
Yeah, we're seeing a mix. I think right now, and I mentioned earlier, the 3x agentic automations that we're seeing within Service Collection right now. I think there's going to be different workloads that lend itself really well to outcome-based results, and we already have some of them in our CSM product. I think the advantages that Atlassian has as a horizontal platform is the ability to stand up different types of agents, the ability to tackle different types of workloads and work that happens across organizations. Right now we see that through Teamwork Collection in terms of the ability to monetize and folks choosing Atlassian to be able to deploy their AI. The other way, as I mentioned, is through some of the CLI and MCP use cases that we're seeing across the system as well.
Now, for simple queries into the Teamwork Graph, we do not charge for that. If you are writing a ticket, if you are trying to resolve a quick ticket, but if you are tapping into the Teamwork Graph, it is much more long horizon or complex use cases that draw on that Teamwork Graph. We do burn down those credits for customers. Again, it is building habits, like we said, across our customer base first, driving that adoption, that depth of vision, and over time being able to get that from them.
Which, as a reminder, you are metering as of December.
Yeah, that is right.
I want to move down the P&L and then talk to you about the gross margin and gross profit trajectory. You talked about being extremely generous with Rovo credits, but you are eating the cost on that, right? Yet we continue to see you keep a pretty strong toehold in kind of the low 80s gross margin ZIP code. Can you talk to me about some of the puts and takes on your ability to kind of maintain that very high watermark while you are still providing the type of value you are providing to customers?
Yeah. Martin's back there staring at me. He's saying, "Mid 80s.
Mid 80s.
On the margin side of the world.
Name of the day.
Yeah, look, Atlassian's gone through a pretty heavy investment cycle over the last couple of years on the R&D side of the world, right? If you think about the investments required to stand up an enterprise-grade cloud, to be able to build an AI architecture and platform that can deliver Teamwork Graph today, we're seeing some of that investment right now, and we're able to reallocate some of those resources. We also have an incredible engineering team that's great at optimizing against a lot of the platform optimizations that we're seeing. Also, one of the advantages is to be able to look at the seven model makers that we work with, the over 70 models that we use across the platform on behalf of our customers, and be able to optimize against that as well. So driving leverage across gross margin.
As customers continue to use more and more and find value in our Rovo credits, we're going to continue to see some of that pressure show up on the gross margin side. But that's also why we focus more and more on GAAP operating margin profitability. Saw some of that flow through in Q4. You've seen us set a positive 4.5 GAAP operating margin for FY 2027, and as we drive more value to our customers, having that accrue to the top line, being incredibly disciplined on the investment side of the world, we're going to continue to see that margin.
Speaking of which, managing that operating expense envelope with a very assiduous mindset on GAAP profitability and durable growth with GAAP profitability.
Where are the toggles inside the operating expense envelope, between R&D and sales and marketing and anything you can express in terms of headcount growth? Generally, I think you've been very judicious there as well.
Yeah. I think that's largely reflected in the FY 2027 guide at the moment. We talked about some of the levers on the COGS and gross margin side of the world. From a headcount perspective, something that Mike and I talked a lot about was around reallocating investments across the board. We continue to see a lot of opportunity on the enterprise side, so you'll continue to see us fund the AE side of the world and continue to grow there, and grow the capacity to kind of follow the demand that we're seeing right now. AI will be another area that we continue to invest in, certainly. If you think about also the leverage that we're getting, 99% of Atlassians internally use Rovo today. We've got 15,000 agents stood up, completing roughly 40 tasks every single day.
Driving a lot more with the resources that we continue.
Rounding it back up to the top line, one of the things that you introduced as CFO at the Atlassian user event in May was Subscription ARR. It's to abstract away a lot of the noise that is prevalent in the model because there have been shifts in the idiosyncratic changes around migration that we talked about. Subscription ARR has been accelerating.
Yeah.
You're guiding towards a deceleration in fiscal 2027. What's on the either side of the ledger that you're accounting for in terms of assumptions that you can help us? We'll have you back here at the conference a year from now. What are the two things that you feel are going to be most potent in your ability to upside surprise on your Subscription ARR performance after you put a caption on fiscal 2027?
Yeah. Subscription ARR was introduced back at our investor forum in May, really to help, to your point, reduce some of the noise. It is the best indicator in terms of the momentum that we are seeing in the business right now across Cloud and DC, 23% year-over-year in Q4. As we think about what also contributed to that 23% in Q4, there was a point of growth driven by DX acquisition that we did in our fiscal Q2 as part of that. As we look out into FY 2027, number one, ARR is a new metric, as you mentioned, that we are introducing for the first time, and making sure that we are being prudent around that, as well as some of the macro uncertainty, the fiscal policy that I think we wake up to every single day right now, and it is still early in the year.
In terms of the growth drivers going into the rest of 2027 is about executing. As I think about the capacity that we are continuing to build on the sales side of the world, the adoption that we are seeing across our customers on Teamwork Collections, Service Collections, I think those will all continue to be upside drivers for us.
James, I have been asking all my presenting companies at the conference this year this question. I gave you a magic wand and had you wave away one major salient investor misconception or an element of your investment story that is underappreciated by investors. What would that one wave away point be?
Yeah. I think, as I reflect on the last five months and the last couple of quarters for Atlassian, it really is that AI is a tailwind for the company. I think enterprises, organizations, they are realizing how incredibly challenging and hard it is to be able to drive AI value throughout the enterprise. The days of talking about vibe coding your own CRM, your own Jira, et cetera, I am not hearing that. We are not seeing that in our results. We are not seeing that in our seat expansion, the adoption across our collections as well. But really, as companies think about where to drive value, it is in that graph. It is in that Teamwork Graph that we have. That is what is actually opening up a lot of the discussions with the C-suite right now and driving a lot of the results.
Thank you so much, James. It's been a very eventful five months. We're looking forward to see what the next year brings.
Yeah.
Thank you.
Awesome. Thank you.