Everyone, thanks for joining us today. I really appreciate you all making the effort coming out to Team 2024 and joining us for Investor Day today. I'm Martin Lam, Head of Investor Relations at Atlassian. Before we jump into it, we may disclose in our materials that talk about forward-looking statements. So on our screen is our typical legal disclaimer regarding such statements. I'll pause here for a moment to give you a chance to read the legal disclaimer that covers such statements made in today's session. Okay, we've got a great agenda for you today. We've got some fantastic customers who will tell you all about how they run their organizations with Atlassian. We have our executive team here for Q&A with co-founders and co-CEOs Mike Cannon-Brookes, Scott Farquhar, Company President Anu Bharadwaj, and CFO Joe Binz. We'll have three sections that will go over materials.
So we'll start with Section 1, which covers our strategic vision and the significant opportunities. Then Scott will dive into a customer panel with H&M and Rivian. Section 2 will cover our platform and the big bets that they power. Then Scott, Mike Cannon, and Joe will join me on stage for a live Q&A covering sections 1 and 2. Then we'll take a quick break and watch a short video about the Atlassian System of Work, and then dive straight into Section 3, which outlines our path to $10 billion and beyond. The team will once again join me for Q&A covering that section following that. And so we saved a good amount of time for Q&A today. We want to accommodate all your questions as much as possible, as well as anyone streaming virtually. Please feel free to submit questions online.
For those of you in the room that rather submit your questions virtually, there's QR codes on the table. You can submit through that as well. Again, welcome to Investor Day. We just came off earnings last week, so it's good to have a chance to speak with you much more about our longer-term opportunities and dive a bit more about the Atlassian business and help you understand how we'll evolve over the coming years. Again, it's really great to see so many of you joining us in the room today. We'll dive into the materials. For those of you that have never been to an Atlassian Investor Day before, it's a little bit different than what you're typically accustomed to. We don't have the traditional slide deck and presentations.
This is a chance to actually let you have a peek inside of how Atlassian runs our business. We issue 6-pager memos. We think it's a fantastic way for us to focus our thoughts a bit more, be a little more concise with our thoughts and how we articulate them to you. And as readers, it forces you to focus a bit more on the stuff that matters. And you're all getting it all at the same time. So for those of you in the room, you should be seeing it passed out right about now. And then for those of you streaming online, the documents should have dropped on the platform. So this is how we run all our meetings, from product strategy sessions to business updates to our Board of Directors meetings. Okay, so first up is Section 1, our strategic vision and the massive opportunities ahead.
We'll spend about 15 minutes allowing you to read through this document. All right, so now I'd like to welcome Scott Farquhar, co-founder and co-CEO, as well as some incredible customers to tell you about their experience with Atlassian products. Please welcome Mike Hayes, Senior Architect, Digital Workplace at Rivian, and Michael Heldén, Atlassian Platform Owner and Engineering Manager at H&M Group.
I wear it, so I let the gloves come off. In a hurry, the taste of my own blood. It's salty, but I fight for what I love. 'Cause I'm worthy. It's all I need to say. It all feels to break.
Hello, everyone. Nice to see you all. It's the biggest room we've ever had for investors. I appreciate that. Thank you for those who came in last night, and thanks for following us at Atlassian. I hope you enjoyed the keynote. I will enjoy this conversation. But before we do that, I'm going to welcome you both. It might be worthwhile just giving a quick introduction: who you are, what do you do, and what companies do you work at.
Yeah, my name is Mike Hayes. I am a Principal Architect for Solutions for Digital Workplace at Rivian.
I'm Michael Heldén. I'm from H&M. I am the Engineering Manager for the team running the Atlassian Cloud Platform. So that officially makes me the Atlassian Platform Owner as well. Been at H&M since 2010 and been on this journey with Atlassian for a long time.
Great. What was your favorite thing in the keynote this morning? Favorite announcement?
Oh, go ahead.
Can I start?
Please.
Well, one thing, I mean, the Jira announcement was, of course, really cool. But one thing that really struck me is the way you can now work with Confluence and Loom together. Because I'm thinking, when we talk about products, we probably mostly think about software, services, and products. But I mean, H&M is not really a software company. We sell clothes. So what I'm thinking is the opportunities here are really endless. Because our products are shirts and skirts. And we have designers all over the world. And what if they record a Loom, create a Confluence page, maybe with a whiteboard, create some sort of mood board, and submit it to the Chief Designer? And they worked with it in Jira Product Discovery and prioritized, sent it to the team to start working on things. So that was really cool, I think.
Realizing that it's not just a software tool, kind of.
Great.
Yeah, so mine was taking Jira Software and Jira Work Management and calling them Jira. And what struck me about that was it's a small thing. But people who've been doing this a long time, it's not a small thing because we didn't want it. We've been saying we get anxious about pricing. Is it going to be a different model? And no one I know was happy for this. And they did what we asked. And so that's really a testament to our relationship, not just ours, but the community, everyone that came to this event. It's a little thing that most people probably don't care about. But the ones who matters to it, they do. So I liked that a lot.
Great. Tell me a little bit about your journey at Rivian, or maybe the Atlassian journey at Rivian. Kind of how long have you been using our products for? How many people use them? What do you use them for? Use cases, that sort of thing. So I'll start with you, Mike.
OK. So I've been at Rivian for over three years. And I work for the Digital Workplace team, which is kind of a new way to look at servicing employees within a company. So we're technically in IT. And we build solutions for employees. So you want to have a process that tracks tasks or whatever they come to us because they're usually in a spreadsheet. And we build that for them. And we're all on cloud. So we didn't even go to Server at all. And so that's really helped keep the maintenance out of it. I don't know anything about Servers and deploying stuff. I don't need to know all that. I can talk about features and solutions. And what has been the biggest impact is when we were starting up, every team we have 14,000 people. But at the time, we had maybe 5,000.
Every team had their own little tool for doing what they wanted to do their way, which is great, a lot of autonomy, but not a lot of alignment. And so what we did is we went through and standardized on a lot of things, got the processes working across the teams with structured data. And now we have a lot more alignment. They still have autonomy because they can do a lot with these tools. But we have a lot more common thread through each team's workflow.
Talk about the—I think you've replaced some other tools that you were using with Atlassian products. And earlier, you were telling me just the breadth of different use cases you use our products for. And as a manufacturer, I was surprised, actually, kind of the different use cases that weren't familiar. So do you want to talk through, again, other products that maybe have been supplanted by us and the use cases?
Right. So if you're a team and you go to Rivian, and at your last company, you use Asana, and we're in growth mode, we were hiring 200 people a week. So keeping everyone on one tool is really difficult. So their manager would approve, okay, you can have Asana. You can have monday.com. You can have Smartsheet. Because it allowed, again, autonomy. They could get their work done. But when it came time to work between the teams, you couldn't do that. Then you start getting into price. All those tools together are very, very expensive. So we went through and said, okay, we need to bring in our budget, like every company has to do. And as you get more mature, it's something you are responsible for.
So we went and looked at the usage of all these ones that had duplicative features that the Atlassian suite has. And we just—I wish we had pulled the Band-Aid off, but we didn't. We went to them and said, you know, we're not going to renew this. We have a solution for you. We will take your requirements and build it in this other one. And yeah, we got rid of five of these sort of tools that only one or two teams used. And they're all happy now. They want more features. So everyone's happy they did that.
Fantastic. And at H&M, you're, as you said, was it shirts and skirts? But you have a huge technology organization. You want to talk about the size of that technology organization, again, the Atlassian journey, how you use us?
Absolutely. So we started with Jira in 2005 already and Confluence in 2015. But in 2020, we moved to cloud. And then we actually became the first customer, at least as far as I know, to have 5,000 users in the cloud. So we almost even had to convince Atlassian to allow us to do that because it was so early in the days of the Atlassian Cloud. This was mainly within the tech organization. So basically, we are around 5,000 techies. I think we have close to 400 teams. We have areas which are similar to release trains, if you are familiar with that sort of terminology. And we have multiple portfolios or departments, domains. So it's a really huge organization. And what I think makes. I'm not sure if H&M is unique.
But I mean, it's a highly connected enterprise where it's basically just one long value chain from sort of design, production, sales, logistics, and all of that. And everything is so tightly connected. Take a price tag, for example. If you need to change something, that change needs to be reflected in sort of the production and stores and everything. So it's really important for us to sort of have all the teams within H&M connected across departments. So that is something that we have really worked on since we moved into the cloud. We have sort of made a scaled agile transformation. And I believe that the cloud has really helped us in getting there because it's so easy to maintain and change.
There's a bit of a trend in the industry at the moment where we've seen across the industry layoffs, some sort of consolidation. I think you at H&M went up a little, and then you've come back down a little bit. Can you talk about that journey? And then also, again, you got 5,000 people in the technical org. What does that look like? How many non-technical people are there? And are they using the products? How do you interface?
That's a very interesting question. Because in the transformation that you referred to in 2020, we started with 5,000. And then we peaked at 7,600 users. And then we made a transformation in 2023 that brought us back down to 5,000. But now, the latest transformation has also sort of created a stronger connection between business teams and the tech department. So what we see now really happening is that when the business teams work closer with tech and they start seeing that, oh, what are those shiny tools you have? We are sitting here with PowerPoint and Excel. And they need to be able to collaborate with tech.
And so we get a lot of requests from our business teams, not only to be able to collaborate with the tech teams, but also that they themselves, in their own sort of way of working and how they track their work, want to use the same tools, basically, in terms of Jira, Confluence, and so on.
Yeah. So Mike, at Rivian, some of the use cases we were talking about before, even just the sort of non-obvious use cases that people would use obviously, there's the things that are tracking software projects, and those which we will get into later. But tell me some of the non-obvious things. I think you mentioned that just measuring tools and measuring things on the production floor was handled in our products. So give me some examples.
Yeah. So metrology is a science of measuring things. And I say it because I don't know about building cars. I know about process. So we've got the software team set up. And then this other team, the metrology team, says, "We are getting a lot of requests because these teams have to ask them for metrology scan and do stuff with that data." And if they're just getting email or Slack or passing them in the hallway, it's really hard to keep track of it, which ones are the biggest priority, and that kind of thing. So we built them, basically, a way for people to request a metrology scan for whatever part they were working on. And just that example, what that gives you is now traceability between this body and the scan results or this tool or this part.
And now, instead of a spreadsheet saying with a part number, the scan passed, and then the other team having, we did a scan on this part, we know now the history of it. And that's very valuable when you put this stuff at scale. So we do a lot of those with the teams interacting with each other but requesting from each other. So that's been really important. Because you don't want to just let everyone ask you everything all the time. You want them to follow a process. Otherwise, you're just going to get overwhelmed. So it's really helped separate those responsibilities, but at the same time, tie it all together.
And I find that's an interesting one. Because when people think about a service desk in an organization, sometimes they think, my IT team are the people that need to have a service desk. And when you look at companies that are really advanced in how work moves through their organizations, you find that, actually, service desks pop up all over the place. Because no one really wants to handle requests coming back and forth in email. And Rivian, you have quite a large number of service desks. Do you want to talk about? So you've got 14,000, 16,000 employees. How many service desks do you have?
So we go by we have 1,200 service agents. So we have tons of service desks. And we don't use, sorry, but we don't use Atlassian. I wish we did for IT help. I need a new monitor. That's a different product. But we have more agents in what we're doing than they have for the whole company. So it becomes the interface to that team versus going and just saying, oh, who's in charge of that team? Let me find their email. I'll go on the org chart. Oh, I think it's them, kind of shooting a Slack message off. You have an official thing that someone will get back to you. You can check the status on it. And it's just widely used. Almost all the teams have some sort of way of requesting something from them.
Yeah. I feel proud about Atlassian. We often have a more aggressive pricing strategy than many competitors out there. As a result, we just expand the market in terms of seats. We're not shrinking on an ARPU basis. 1,200 people using a service desk, that's 1 in 10 almost inside an organization. That's just unusual. It wouldn't be if we didn't have the sort of pricing model that we have.
That is true.
At H&M, I think we were chatting earlier that the tax team used our products. Can you talk through that? That's not an obvious use case.
So that is a very good example of how we see that kind of growing organically into the business organization. So it actually started with the tax team just reaching out randomly and asking, does anyone know if we have some sort of workflow management tool? And fortunately, I found that request first. So basically, they had been looking into kind of branded tax management tools. But they themselves had sort of the perception of those tools that it's basically just lipstick on a pig that you are paying for. So we showed them Jira Work Management, or maybe just Jira these days. And they really liked what they saw. So we started to build what they needed in terms of workflows. Because what they do is they want to sort of simplify and automate their administration of recurring tax issues. I'm not a tax expert myself.
So we did that. We also had a couple of meetings with Atlassian's tax team, who are also using Jira Work Management to manage the tax issues. I mean, it was instant love when they realized how that when they close one tax, like for a quarter maybe, they are working on something. Then they need to start exactly the same thing. And how the automation then just created a new task. So they didn't even have to think about creating a new task for the next quarter. It was just there magically as soon as they closed one. So we are now looking into growing into the country tax offices. This is the global tax office. They want their colleagues across the world to be able to collaborate with them. Very good example of this sort of organic growth.
Yeah. It's interesting. At Atlassian, one day, the tax team came to me and said, please, please put us in front of customers. We want to be your best sales team. We want to explain how we use Atlassian products inside our tax team. And so we put them on the road. And now, tax team are some of the best salespeople we have going around telling other tax departments in organizations how to run more efficiently using Atlassian's products. And again, sometimes it's hard to comprehend kind of the breadth of the use cases that we can get used in across organizations. And I figure if tax teams can use us, then it's a pretty good example. I want to chat about Jira Align. You use that across all of H&M to sort of roll up all the work that happens.
Can you talk through that journey, both in terms of what it enables your executives to see and do, and then kind of how it works and the benefits you get?
We are working on the rollout still. But what we really aim to achieve, I would say, is now, I mean, we can get a pretty good overview in Jira together with plans, especially, and analytics. But I mean, the really big advantage that I see with Jira Align is that when we also get the business to work on the platform as well. And we can sort of aggregate all the data. Because right now, we are only being able to visualize and analyze, which is also important, the work that we are doing within tech, which means that, basically, 50% of all the work that we do in the head office is not visualized. And we do not really have the capability to measure time to market and so on. Because we can only measure the time it takes once things hit tech.
But a lot of work has gone into even thinking about what new initiatives should we launch and so on. So when we get this sort of the entire chain into the platform and aggregate it into Jira Align and using the business insights, that will sort of create entirely new insights about how the organization works.
All right. We even practiced these questions. I'm going to take a bit of a stab in the dark. Hopefully, it works. We're really proud about the platform we've built out in cloud. That platform allows things like analytics to run across multiple products. Then our customers get insights that they couldn't get if they were using just an individual standalone product or using products from different vendors. That platform provides things like automation. It's the backbone for all of our AI work. To put you on the spot, because we haven't talked about this, are there features that you use from our platform in that way, analytics, automations, and beyond, that, again, maybe you couldn't get if you used different vendors for that solution?
Yeah. And the examples we both used of building these solutions for these teams you don't know about taxes. I don't know about building cars. But we're making those teams work more efficiently. And it's because we understand the platform. The platform has the ability, gives us the tools to do that. And we understand workflow. So you don't need, if you were to build this kind of stuff on a much larger, more expensive enterprise-type system, it could be done. But you're hiring specialized contractors. It's complicated. And I think there's the expression, the devil's in the details. And people know me. I think God's in the details. And so you can say, oh, let's run a report on what projects are late. But if you don't know the date is the start date versus end date or done date or close date, those things matter.
And so it's not just reporting on it. It's getting everyone on the same page of this is what this date means so that we can actually have meaningful stuff. And that's what you can do on the platform, is you can get as customized as you want with that team, but still make it broad enough that it makes sense to other teams. And that's very hard to do in a Smartsheet or a Monday or that kind of thing that specializes on maybe a project plan or a Gantt chart.
I would say something that is really in our focus now is to improve the collaboration with the vendors that we are working with. Because we have so many vendors in everything from marketing to logistics and warehousing and so on and so on. We sort of want to keep them a little bit apart. Because we don't want one logistics partner to have the full insights into what another one is doing and so on. We have started looking into how can we create a better environment for the collaboration between H&M and vendors to happen. I think there is a huge potential there. Because with our current setup, we basically require our vendors to have H&M accounts and so on. There are new opportunities now, partly because we have become enterprise customers.
We can create multiple sites within sort of the enterprise umbrella and still maintain them within the central admin team. You can now have business-to-business accounts, which makes it much easier to set up the access. I think that is something that is really huge and happening now, how to enable this collaboration with all our partners. Because there are many.
Yeah. I want to add to that just to make sure it's clear here. Atlassian's done such a good job internally making work happen across an organization. Obviously, started in technical teams. Now, as you've heard from examples, it now helps work move across organizations in all sorts of teams. We've invested over the previous few years making it easier for our customers to interface with their customers, both from things like identity management and also making things like Jira Service Management work with their customers. It's not well, I think, known among this group. We have an enterprise product in enterprise cloud. That actually allows multiple sites for our customers under one sort of pricing umbrella. What we find now is our customers then spin up lots more sites in different permutations, sometimes, as you said, internally, but sometimes in collaboration with their customers.
We see that a lot of professional services organizations actually use our products in conjunction with their customers to get work done. That's probably an unappreciated part of Atlassian. I'm going to ask another high-risk question on stage. If Atlassian went away tomorrow, what would you replace it with?
That's a really good question. I'm thinking we have developers who would refuse to touch any bad service management tools. And we have business users that love Trello and the simplicity of that sort of tools. So I mean, for me, it's really, really hard to say, especially now in these days where we I mean, H&M is a big organization. We need to be able to collaborate. I would have loved to say whiteboards and Post-its. But it doesn't work anymore. So from my point of view, from what I know, there is really no other platform that really unites people from business all the way to operations that everyone loves. I mean, some tools have their fan base in one of these sort of sectors. But I don't really see any alternatives.
Thank you.
I don't want to answer. No. The anxiety from that. And it's interesting to contemplate that. Because it's like, is it that important? It is. If you want to do better and be productive and move along, you can do all this stuff in Excel. You can run your taxes in Excel. It's fine. You can collaborate with external people from Excel. You could. But you're not getting any value out of that. And this is to talk earlier, when we collaborate with a supplier, they say, what is your Jira? It's not, can we get on your system or get on your network? It's known as Jira. That's how you collaborate with things. And so it would be very expensive. And it would require a lot of technical people to tie the different things together.
So if you use one tool for project management and another one to do accounting or something with taxes, you'd have to spend money to hire people to tie all those things together and maintain it. So you would be recreating a homegrown sort of Atlassian platform, which now you've got to worry about security, login, licenses or not licenses, but just user management, permissions, all those things come with it. So when we spin up a new site, we just say, this team and those people, and it works. Doing that in any other tool is not fun. Yeah.
If you want to. I'm glad we're recording this. Because later on, instead of transcribing the answer, just look at the three seconds of their faces after I said what would happen if we disappeared. So I think that says everything we need. I'm going to take one or two questions from the audience, if we can. I can repeat them if we don't have roving mics. If you have a feature request or something you'd like Atlassian to do in the future or a market you think we should enter or something adjacent we should do or something we're not solving for you already, what would that be?
You can start.
So this is what I think it needs to carve out, sorry, this thing of, again, workforce being productive. That's not really a thing that companies recognize needs to happen. It's, oh, my boss likes this. So we spend all our time converting it and copying and pasting it over to the format that then sent an email as an attachment. That's coming, the system of work, how people work, how they're productive. And that doesn't exist now. And it's coming. And that's, I think, just more of pushing that and having companies realize, yeah, that's an important thing, the way we collaborate, the team thing, versus just saying we're into teams and collaboration. If you don't take steps or technology or do anything about it, you're not really doing anything.
I think, again, the trend is companies focusing on how do we, what tools are we providing the employees? And are we making sure they're using them to their fullest instead of just them using their knowledge they have from their previous jobs? So that.
Thank you.
I would like to connect to what you were saying. Especially the Teamwork Graph is something that I really like. Because it has sort of created a shared understanding of what it is that we want to achieve in terms of sort of the architecture that supports the DevOps loop or the teamwork, basically. Because that is something that we have sort of been struggling to align on, partly because maybe we have different tools in different parts of so I think the Teamwork Graph and continue working with the Teamwork Graph and really make sure that each entity sort of in the Teamwork Graph really has its home so that we can really talk about what is the master data for a goal or what is the master data for a team.
When we have that, that will create so much clarity and make it really easy to reason about the architecture and how to connect different nodes in the teamwork graph and really have the tools support that. But I mean, from what I saw in the keynote this morning, it feels like you are already working in that direction.
I think we're heading both of those right now in our future. We probably have time for one or maybe two questions in the audience, if someone has one. We've got one down here. If we can get a microphone. Else I can thank you, BK.
Hi. Thank you. It's Keith Bachmann from BMO. First, Mike, I'd like to understand if I could get a discount on Rivian.
I wish I could.
Yeah. We did ask them to both of them bring samples in. It was easier for one of them. We decided we could do that.
Yeah. For both of you, how do you think about your seat growth with Atlassian over the next couple of years? And if you dissected it into, say, Core Jira versus other, how do you think that evolves over the next couple of years? And then as part of that, if you just wanted to give any feedback on the announcements today on Rovo, that'd be great. Thank you.
Thank you. I can start. Yeah. As I mentioned earlier, we have already seen strong signs of interest from the business side. And there are new capabilities on the Atlassian platform that I know the business would love. I mean, H&M is a very creative entrepreneurial company that likes to work visually. So Loom, whiteboards is something that I think young people, creative designers will really, really love. And basically, the business organization is just as big as the tech organization. So I think maybe another 5,000 users within a couple of it. It's not impossible. I cannot promise anything. But I hope that we get there. Because that is the only way we can sort of connect from idea to delivered initiative. And then, I mean, there we have 10,000 users. But H&M as a whole is over 150,000 users.
There is also the potential to grow in the store side with incidents and whatnot, insights and so on.
It's hard to comment on number. I can say, as far as growth, and it's very popular. We have to tell people used to, in the old days, it was like, I don't want to use Jira. And we'd say, no, no, you should do. I mean, the old days, like 20 years ago. But now, everyone's coming. We have a finance team that says, oh, you're doing that thing. Can we have that? So it's growing. I don't know. I'll put it this way. There's not a lot of software in these times that's actually getting a bigger footprint within our organization. And I definitely see Jira or Atlassian stuff doing that. And as far as the keynote, again, been doing this almost 20 years. And that was one of the coolest sets of overall features that's been at one time, in particular, the AI.
Someone had to say it first, almost at the end. But just thinking about how all these things that we talked about, if they're connected, but then the person getting a summary or seeing that information the way they want to see it. So if I'm an executive, I just want the five highlights. If I'm down here in the weeds, I want all the details, where it came from, who's the last one. And so I really am excited. Just, AI is going to take our jobs. And it's thinking that we have all these tickets that come in, vehicle feedback, thousands of them. And we've not gotten around how does you not have somebody put them where they go? Is it a software? And this is it.
So what we do is we take someone who's a project manager and say, your job is to now triage those, but also be a project manager. So we didn't know that person's not going to lose their job. Because they're just going to be able to do their job, only their job. And so that's been we were just talking about it before. It's going to be big, that part.
All right. Well, we can keep going all afternoon. I want to say a huge thank you to our guests for coming in. Hopefully, you learned something about Atlassian. Thank you. Appreciate it.
Thank you.
Thank you.
Hopefully, you enjoyed that as much as I did. So next up, we'll be distributing Section 2, our platform-powered big bets. So we'll take about 20-25 minutes to read this one. And then we'll bring Mike, Scott, Anu, and Joe. And then we'll take a quick break with the customer video after that. All right. So now that you've had a chance to read through sections one and two, let's dive into some Q&A. So as a reminder, we'll focus on questions from the audience, but of course, we will look to take questions virtually as well. So I'd like to welcome Mike Cannon-Brookes, Scott Farquhar, Anu Bharadwaj, and Joe Binz.
Let you how could I let you go? I've been dancing on the floor all day, and now the beat just gone. And all the lights are on, you see. I tried to find you, but why can't I describe you? Maybe only one I'm saying enough.
I think before we get started, Scott actually wanted to double-click on something that you read in section one.
Hey, everyone. Am I on?
Yes, no.
Test, test, 1, 2. There we go. That's better. I'd like to just you read this in, I think, document number 1, if you want to flick it open again. But we did this work on total available market, total addressable market, serviceable addressable market. We basically have done this work internally about what we think the opportunity is for Atlassian. Now, I have a healthy skepticism for any total available market numbers. I find that in many cases, you can kind of juice the numbers to sort of work out whatever story you want. But we did this work internally, really, so that we could work out where our investments should go, and where should we be putting our energies, and what is the opportunity inside our market. And so it really was an internal process.
And again, we love sharing things we do internally with you. So we look at this opportunity we have at Atlassian, and we did some work to look at if we had our current products at our current prices to our current customers, what does that look like? So if we wrote not a single line of code again, and we didn't acquire a single another customer again, what does the opportunity look like for us? And that's exciting, because I can go to my go-to-market teams, my sales team, my marketing team, and say, hey, this is your opportunity for you to deliver on. And that number ends up being, you can see here, $18 billion. And a large part of that opportunity is in our enterprise segment. So that's, you know, hear from us about our priorities, where we're putting our effort and time.
Enterprise, you can see the huge upside we have. So we're a $4+ billion run rate business, and we can get to $18 billion without, like, acquiring another customer. We've done a relatively conservative way of doing this. So, for example, if you're a small customer of ours, we didn't say that we would sell the enterprise version of our products to you. We said, well, actually, which products would fit your particular size and scale of company, also based on what products you currently have, and then what other incumbents and other opportunities we have to displace other products inside that base. If you look at our free base, which we don't really talk about at Atlassian, we don't call them customers. They don't actually fit in our customer definition, because they don't pay us a certain amount a month.
But we've, you know, increasingly grown that free base of users. And you can see here, 200,000 instances that we really don't talk about very often. And we could double the business of Atlassian by just taking those free customers that use our products today. And again, looking at those customer company sizes and the seats opportunity that we have inside that, we could double the business there. And of course, we continue to grow Atlassian, acquire new companies, new use cases. And there's a huge opportunity for us, again, 4 million companies. That's not every company in the world. That's the companies that are knowledge workers that we think would fit with our current products that we have. And that, again, is a huge opportunity there. And so again, these are our current products at our current prices.
So this doesn't include things like FedRAMP, which is coming down the line. This doesn't include pricing and packaging changes that we might do. This doesn't include some of the enterprise scale work that we're doing to scale up. And so this is a conservative view of how we do things. But when we look at this, you know, it becomes obvious. I think if you look at the same chart, you'd say, hey, we've got a huge opportunity in our enterprise customer base today. So I'm sure you read through that in the page, but I wanted to just give a bit of a voiceover before we answer any questions. Back to you, Lam.
Let's start with Itai up front.
Thank you. Congrats on a great event. I've been to this event for many years, and it's quite amazing to see how big it is right now. Congrats on all your success. I have a couple of questions. First of all, I want to dig in into the logic of consolidating work management into Jira. Maybe you can give us some color on the thought process behind that. Why should we not take that as a negative? It wasn't working, just folding it into the product. Then second question around AI. It sounds like last year, when Gen AI popped up, it was a surprise for everybody. Everybody's kind of scrambling. You weren't sure at the time what model you're going to use. It sounds like from my conversations with your people on the floor that OpenAI is something you're leaning into quite heavily.
Is this a done deal? How do I think about your model in the future? Would you develop something on your own, keep it diversified? We'll have to get more color on the logic there.
Sure, I mean, I can take both of those, and anyone can chip in. Interesting. Should you see combining Jira Work Management and Jira Software together as a negative? I think was the first question. Or why should you not see it as a negative?
That work management has a work.
Sure. I mean, we gave the stat in the thing that Work Management, Jira Work Management, is growing extremely strongly, right? North of 50%, huge customer penetration, etc. I think there's two main reasons for that, and we tend to be incredibly customer-led. Firstly, Jira Work Management was a great opportunity for us to experiment with a different audience, right? We don't need to take all of the things that software teams need to be able to build into that, and we get to experiment with new things. We've done that, and it's been hugely popular with customers, as we've seen from the uptake. The second thing is a lot of software teams like, hey, all the new stuff you put in there, we want that as well, right? Because, as Anu said in her keynote, like, they like colors too. They want their stuff to be colorful.
And so why didn't we get the colors? So you ended up getting an ability to experiment. And then, you know, a lot of the features are getting folded back and cross-blended between the two. Secondly, from a strategic point of view, the number one piece of customer feedback that we've had is we want our technical and non-technical teams to work together. We want them all to work together on one platform. And so it seemed like a logical move as we've done the R&D work to start folding the features together as well, to look at the packaging over time, and to put them back together again. And at least it's only been a couple of hours, but the 2, 3 customer interactions I've had have all been very positive on this. And they believe they'll have more seats as a result.
And so that's a good trade for us, right? So it's nothing to do with Jira Work Management going badly or not working. In fact, it's quite the opposite. It's actually been a big spurt of innovation and increased the pace at which that's going. On the AI side of things, I'm not sure I quite understand the question. We do use OpenAI. Was it about OpenAI?
Yeah, it sounds like Atlassian Intelligence is built on OpenAI.
No. I mean, yes and no. So we have 6 different vendors running internally at the moment. I think 15 different models from 6 vendors. Not all 6 vendors have models running in production, I should say. We have 15 different models running in production, I believe. It's around that number, some of which definitely come from OpenAI. But we also use Meta and other people in production at the moment. The way we see it is that's our job. The customers don't have to care. As long as the legals and the compliance and all the, you know, that's our job. We use different models for different features, depending on speed, cost, and quality, so the performance of the results we get. And we've given a few examples of different things. So we're definitely great partners with OpenAI.
They've been really helpful, but we are also using a lot of other models simultaneously. You talked about training ourselves. Again, that's up to us to do, I think.
Michael.
Thanks very much. Thanks for doing this, and thanks for the notes and disclosures. I guess one of them mentioned Jira still growing faster than the overall business. That's probably not a surprise to you, but it might be more of a surprise to some of us that haven't seen the product update in some time. So I'm hoping you can first just speak to what's enabling the continued growth of core Jira. And then when you think about the product mix and how it informs your investment roadmap and R&D priorities, given you're going after three pretty significant categories, like, what are the trade-offs you're contemplating, and what informs that? Thank you.
You want me to get you? Can direct traffic if you want.
Look, for sure, I'm not sure if it's a surprise to you or not that core Jira is growing faster. Look, Jira Software has been a strong performer for us for 20 years and continues to be so, right? I think our main point there is that we're far from any sort of saturation point in its penetration. Even in our existing customer base, as Scott just showed in some of our TAM slides, it is a good thing to be spoiled by a series of large products that are all growing fast and competing with each other on that growth curve. It's not a bad thing at all.
I do think that combining it with Jira Work Management and having customers be able to get that, as well as, I should say, the increasingly consistent and coherent platform experience that customers really resonate with, like the fact that automation, we ship features now that work in multiple products simultaneously. And as we keep going through that, I think it just will hopefully continue the growth path of Jira Software. Like, I think that's, I don't know, hopefully it's not a surprise now to everybody. But yeah, sure, it's a fantastic product and continues to go strongly, as are all of our large products.
And I would just say mechanically, Michael, it's all the things we've talked about on the call, right? It's paid seat expansion. It's pricing. It's cross-sell. It's upsell. All those things are contributing to the Jira Software core growth.
Also just from a product perspective, it is incorrect that we haven't made updates to the core product. In fact, we've made quarterly updates to the core product. So our Cloud customers experience those experience improvements, updates, and want to actually use the Cloud products over Data Center. And we see that clearly in the customer feedback that comes through. But to link it back to your second question of how does that inform how we make R&D investments? So we make R&D investments in the primary pillars that we talked about in the investor letter. And those three things are very carefully chosen.
So the enterprise investment, AI investment, and harmonizing the portfolio of work across our products, all three of them form a virtuous loop to really help increase some of the business levers that Joe talked about, as well as give compounding gains in terms of experience benefits that Jira users experience, Confluence users experience by virtue of having that common platform.
Ryan, up front.
Hey, Ryan MacWilliams, Barclays. Really cool to see the Rovo launch today. Love to hear about the backstory on the creation of that project or product, and along with the opportunity you see there. Separately, we'd love to hear about the decision on why to charge discretely for that product and why to offer it for Cloud Enterprise and Premium SKUs. Thanks.
I mean, the backstory is how would I answer that question? I think if I take the highest level in the popular consciousness, LLMs, what is commonly pseudonymed as AI, but not necessarily technically correct, is seen as generating content, right? We're all used to ChatGPT, write me a Seinfeld script, it spits one out. This is really cool. This seems magical. One of the things that we have already is a lot of customer knowledge and data. One of the things LLMs can do for the first time ever is read that data and understand it. And then we can turn that into amazing customer features, which is almost like using it in reverse. That is the genesis of what became Rovo. And you see that in a lot of the knowledge and definitions and some of the what seem like small, simple features on stage.
Customers like the acronym thing. I want that. And I'm like, whoa, that's actually like quite a small feature in the scale of what Rovo can do. So I think we've got a long roadmap to continue to show customers how we're doing that. But the genesis of the idea and the origin story is running an LLM in reverse. When we have amazingly valuable to our customers data and knowledge, they have more and more data and knowledge every single day, and it's not going to shrink. How can we help them use that knowledge better to learn from it in real time, to understand it, to find it, those sorts of things? And then the action piece, we have a series of pillars that have become really valuable as our platform has gotten better and better and better, as our Forge development system has gotten better and better.
We're adding some small things on top of that that become Rovo agents effectively, but built on, you know, the shoulders of giants. And secondly, compared to anybody else, we have a huge product surface area. And given our extensibility DNA, we've been able to build those agents into user experiences. The genesis of the thinking there is that people don't want to go somewhere else to use the new things, and they don't want to operate in different ways if they don't have to. There may be occasions where they need to operate differently. But by and large, you know, if you can just go to a Jira issue, mention an agent, you know what it's going to do, write what you want it to do, and it comes back with an answer or a file or something, that is literally how you would engage with a human teammate
So a lot of the design genesis was to do that, use all our platform capabilities, whether it's notifications or mentions. Even in chat, it's just like Confluence page. You have the full editor there. It can give you back tables and checklists and Jira issues and all sorts of cool stuff. We could have done literally 90 minutes on Rovo. We really had to compress it down to 20 minutes. That is the core genesis, is that we have a lot of really interesting customer content, and they're only going to get more and more, and we can understand it. And secondly, man, with our extensibility, we can do some awesome stuff and just make it literally be like they have AI-driven teammates all over the place. And we weren't lying.
We have 300-odd agents that have been built internally, tens, I think it's north of 50 now, used on a weekly basis in real-world scenarios. A whole bunch of those agents written by product managers and marketers in totally no-code scenarios on top of all the Rovo frameworks. Marketplace coming in as well. So we're pretty excited about its possibilities and potentials. But again, like a lot of things we do as innovation and R&D, it's super early days, right? We do the same cycle that we will do every single time that we did last year with JPD and Compass. We put it into an Early Access Program with a very select group of customers today. We slowly open up the Early Access Program to more customers over time and learn from them and tweak it and improve it.
Then it will go GA at some stage, and then, you know, away we'll go. What we stress to customers is our AI stuff. We're determined to ship it. These aren't fluffy demos of things that maybe one day will be good. No, no, no. Every single thing we showed you last year, we shipped it in under 9 months, I think under 8. So judge us on our ability to get it into your hands to use it in the real-world scenario, not make marketing-level demos and just kind of disappear quietly on the sides. It's different in our monetization policy. When we talk about AI monetization, we're very clear. Again, our number one driver is migrations. We think the AI features in the cloud can continue to drive the migration story for a long time yet. And it's very clear why we can't deliver AI features on-prem.
It's very, very hard to. So that's a number one monetization driver. We think it can highly differentiate, given the platform and the Teamwork Graph, all our existing Atlassian Intelligence features. Again, we're shipping Atlassian Intelligence features in every single product that we have. Every single product team has delivered amazing features. You saw how many of you guys did, like 5, 6 on stage, right? There's 30 major features. Last year, we had 8-10 somewhere in that realm. So like significantly more features this year in the way that we measure a feature. Driving new customer adoption, awesome. Competitive, better than someone else. It's just a straight-up heads-up battle. We've got to do that too. And obviously, driving adoption expansions in Jira Software, Confluence, JSM. No part of that is changing. The difference with Rovo, it is it is an entirely new product.
It is a product that could not exist without AI. Doesn't necessarily mean it's an AI product, if that makes sense. It could not exist. I use the example of Uber as a comparator, not for the product itself, but Uber is designed to get you from point A to point B. It can't exist without a smartphone as an app. Like you couldn't build it pre-smartphone. It didn't make any sense. It doesn't mean it's a Samsung phone or it's an Android operating system. It's just neither. It's not a mobile thing. It's just an app that couldn't exist without the mobile era. I think Rovo is a good example of a product that could not exist without LLMs, but it's not an LLM. It has a huge amount of surface area underneath to deliver it. And we're pretty excited. So hopefully that makes some sense.
We can go to Alex right there in the front.
Hey, guys. Alex Zukin, Wolfe Research. I really liked the disclosure of the paid seats by deployment. I thought that was really interesting. I wanted to just ask if I think about. I love the future column. But if I think about just how that trend of Data Center to cloud, seat to seat kind of migration has trended, and how do we think about, you know, the time, not the timeline, because I know you're not going to answer that, but just the pace of conversion from seats, from Data Center to cloud. And in the end, is it possible that that future state is actually a lot more Data Center and cloud, and it's just larger on aggregate? Or do you really feel that the vast, vast majority will ultimately be cloud?
I mean, I can give a high level, and then maybe others can chime on. For sure, we're very clear with our customers that their ultimate future is in the Cloud. As we like to say, the best experience of Atlassian products is in the Cloud. We cannot build Rovo on-prem. We cannot build Compass on-prem. We can't build the Atlassian platform and the Teamwork Graph. There's so much stuff in there that cannot exist on-premise. You have a more scalable, more secure offering. You know, a lot of our Data Center customers will update 5, 6, 7 times a year, maybe if they're super on the ball, some twice a year. Like we literally update our Cloud like multiple times today since the keynote. And so the experience is going to be better. We're extremely understanding of Data Center customers that they have a lot of needs and demands.
I think we've shown over the last three years that we keep constantly like improving our Cloud towards their needs and demands. And so we believe, yes, that is the ultimate destination. For a lot of those customers, though, it can be a little hard to look at pure seat count because of the hybrid situation. So for a lot of customers that we talk to, they have, let's just say they have 20 Jira Software Servers running inside their organization. They don't move all 20 on one day. They move some. There are some that maybe they need to move three years from now because it's there. Some they may never move because they're like, actually, those two are running a project that's going to finish their version of Server or something, right? They're going to like, actually, let's not move those.
So I think you're going to see way more in the enterprise side when you get to the Data Center and the bigger customers, way more of that hybrid scenario than you did with Server migration transition, where it was more of a flip for most customers on a singular day, which will mean those two numbers get a bit more blended over time, if that makes sense. But certainly, we're, you know, maintain extreme bullishness over the long-term prospects of those customers. And I think we're doing a good job keeping them really happy in Data Center at the same time and continuing to deliver value to them on a regular basis. But, you know, being clear, and that's a huge part of why we're here this week as well.
I would say two quick things to add on to that. One, we've actually accelerated the DC seat conversion to cloud. As we put out in the letter over the last year, we've had more DC customers come in. And for them, there isn't really a forcing function, unlike server, right? So these people are choosing to come to cloud simply because of the benefits they see of what Mike was talking about. So we don't really worry about, oh, is there enough value proposition in cloud for DC customers to come? There's already evidence that there is plenty of value proposition. Second, I would think of this as also, if you look at some of the stats we've put out around cloud expansion, paid seat expansion, as well as cross-sell, we just have more things in cloud to sell to our customers, right?
So we don't worry about, oh, when DC customers come to cloud, there is naturally greater surface area for us to be able to serve them with more products, more additions, and therefore be able to attract more use cases like the ones that Scott was talking about with the Rivian and H&M rips.
If you think about the hybrid specifically, if that's a shape, like if you were to, I don't know, put some sort of a graphic for the hybrid portion where it is today and where it'll be, and just maybe clarify for us, how does that RevRec work again?
Joe?
RevRec?
From a RevRec perspective, we would place that in the native area that it's deployed. So if they buy a Cloud contract, we would deploy it there. If it's a Data Center contract, we would recognize the revenue there. We do have hybrid ELAs that we sell today that enables a customer to run Data Center and move to the Cloud on their own terms in the term of the contract. We recognize that as a percentage of revenue of the absolute price, so relative pricing within that contract. So it's roughly 50/50. So you'd find 50% of that showing up in the Cloud revenue growth number. You'd find 50% showing up in the Data Center number.
Fatima up front.
Fatima Boolani from Citi, thank you so much for doing this. I had two questions on Jira Service Management. By all accounts, on my napkin math, that business is slated to get to $1 billion much, much faster than your predecessor, very large products, Jira and Confluence. So what I want to ask you is that what are the limiting reagents within the Jira Service Management opportunity? And if you can frame that from a sales and marketing in an R&D perspective. And then my second question is just around Rovo, because it's been so topical. You've got a tremendous marketplace business, and it seems like a lot of what you're doing with Rovo, in terms of reverse engineering, you know, customer-driven features and functionality, ordinarily that would have been done by your partners in the marketplace, and you've built an incredible ISV ecosystem.
How do you maintain or balance sort of the terms of engagement as it relates to monetizing Rovo, but keeping your marketplace partners happy? That's all I have.
Scott, do you want to take the chance?
I can answer about JSM. Super excited about what we've got there. It is a fast-growing business. There's some large incumbents in that particular space. There's also actually a relatively large number of competitors in the mid-market space that I think we have shown over time that we have a larger breadth of product offering out there. Only we are integrated with dev and IT, like in ways that no other real vendor can do that. So I think we have some fundamental advantages that are very long-lasting. It would be hard for someone else to catch up, given our base in developers. So the question is, well, what's the constraint on revenue growth? Why could we not grow that faster?
I think historically, you know, feature parity was something we've done, and we've done quite a few acquisitions in that space, and both acquiring that and then building it into the core experience. And you saw some announcements this morning in the keynote about how we've made more and more of those experiences sort of native in the product. I think we are sort of cresting that hill in terms of feature parity. I think we have the features that go up and against the best and largest competitors in this space, for all but the largest customers at the moment. So I think the feature R&D kind of parity is getting close or tailing off. On the go-to-market side of things, that's an area where we just continue to take market share. And every quarter, I look at the competitive switch-outs that we get.
We see a lot more of those happening for cost reasons. The ROI is pretty compelling in those areas. They can often switch out and get a partner to do the data migration and realize all the benefits in under 12 months, just from the license savings they save from moving off a competitor. So that's there. Obviously, we can continue to put more effort into our R&D and go-to-market and have a larger base there. But we do so well selling to our existing customer base. Like that is a very efficient sales motion. And so we could chase faster growth in the short term, but it would be sort of less efficient growth there. So that's the sort of lever we need to trade off on the go-to-market side there.
But I don't really see a ceiling in terms of what we can do in terms of market share on that product.
Yeah, Fatima, I can chime in. Always very perceptive. You get credit for the question. Let me explain why. So yes, so in Rovo, you get 20 agents out of the box at the moment. We're committing to delivering at least 20, right, as we tried to explain on the slides. And it's an incredibly easily our most complex product to try to explain to customers in a singular launch. So we'll be doing that. You can see over time to continue to do the education with the customer base. You get 20 agents out of the box. Those are agents that we've built that we use and that you can customize for your organization in very simple ways. Usually, that's about giving it different types of knowledge. So the Comms Crafter that I explained in terms of branding guidelines, you can give it investment-compliant guidelines.
You can give it any basically knowledge, and it'll help you tailor creating of content. That's one agent that you can choose. What sort of process guide do you want to give it? Branding guidelines is just one of many examples that teams could use that agent for. And they can make 20 of those. Bad number choice. They can make 12 of those internally if they want, right? And that's one agent that they can make copies of effectively with no code. They can build their own. So the Marketplace agent, the middle column that we tried to explain, we have built a number of agents internally that we're not shipping to customers that we built for our own business processes and everything else just to make ourselves faster, literally.
Those are built either with Forge, if you want to write code and do really advanced computations and other things, or you can build those entirely with no code. So if you go to the booth, you'll see a few examples, the Decision Director and others, that are quite complicated agents that are built entirely with no code by product managers and marketers and other business people inside an organization. Our goal with those is to be able to achieve real functional improvement of your business processes with someone who's smart enough to write Excel functions. That's kind of our guide of like the hacker, maker. They're not a software engineer, but they understand how kind of Excel works. It's like quite a complex piece of software, actually.
The reason I bring up that column of making your own, we know there are a lot of customers who will make a lot. Again, we've made 200 internally already in like the last 3 or 4 months. Our partners are very excited. And I suspect we'll be more so in the early accesses. We've given a few of them about their ability to go into customers and not just write add-ons for the customers. These are the technology partners, if you like. Not just to write add-ons for customers, but to build agents for customers, right? Not just to write automation rules for customers, but to build agents for those customers. They'll be for specific business processes. And I would suspect as a consulting opportunity, that's a really big thing. I have to understand your business as a consultant.
I have to understand the Atlassian technical frameworks, Forge, various other things, and then build you a unique application that runs in the Atlassian service area that solves specific business problems for your organization, whether you're making cars or fashion or whatever it is, you have a particular thing that you want to do. That is a big opportunity, I believe, for our marketplace over time as Rovo rolls out, depending on its customer base, etc. The third category of marketplace add-ons, you need to be a Rovo customer. So you can think of Rovo as coming with a whole series of things. One of them is the agent framework and 20 agents. The marketplace add-ons that you can download come from our marketplace partners, our vendors, the people on the show floor with booths. Those can be paid or unpaid.
So you have to buy that marketplace add-on or app on top of Rovo if, you know, that's a paid app from that partner. A lot of them are free. Some of them are paid. And that will have its own economic model, obviously, out there. I would expect, and this is total prognostication, we've got about 10-15 partners working with us. They're all super excited. Most of them right now have ideas of what they would like to build. Realistically, what they're building is taking their existing apps and adding agents to them, right? So a marketplace partner that already has an app doing, you know, task X can now just do task X and add agents to it. So you can chat to task X. It can interact in various different ways. You can suddenly have it appear as a person in various places.
That's fine, right? That's an additive kind of experience. It will probably take a little while until they're building entirely new agents that are purely just a paid agent that does something specific. But all of that is a potential possibility of surface area with Rovo. But, you know, we've still obviously got a ton of work to do to get there. But yes, you've sort of nailed at various economic possibilities of agents for us.
We'll go to Luv on the line.
Thank you. Luv Sodha, Jefferies. Thank you again for taking my questions. Wanted to ask the first one. I think a key concern has been around the sustainability of the growth. And so outside of these three big products that you have today, which one out of the broader portfolio set do you think could become the next, you know, product, if you will, that grows at that scale? And then secondarily, a question for Anu on, you know, R&D, just walk us through the priorities for the next few years. Obviously, the past three and a half years, cloud has been a big focus, right? But, you know, what's coming in the future, I guess, in terms of investment there? Thank you.
Sure. I can take a shot at both questions and maybe Mike can follow up.
Yeah, go ahead.
So the way we think about newer sources of revenue, like we broke out, is that, one, the three core products themselves have different levers of revenue. So it's not just selling the three products to different customers, but it's around expansion, what new use cases are we able to serve, which is why we talked about Jira Work Management and Jira Software getting together today. Just yesterday, I spoke with an analyst that was really thrilled with the unification message. It was quite the opposite of the first question we received, which was they hear clearly in market that Jira Work Management is seeing the kind of breadth of adoption that they're not seeing with any other vendor.
So they're really excited about taking the core product of Jira and being able to sell it wall to wall to a lot more different kinds of teams than they've traditionally sold to. So the opportunity within the three products, I would delineate that as it's not just one linear path, but really the ability to address different kinds of audiences with different kinds of additions as well in Cloud. So that's one thing that we have continually invested in, which helps us drive more sustainable growth in the core products. In addition to that, the way we think about other sources of revenue really is the continual product innovation engine that we have with what we talked to you about over the last few years as Point A. We've put out a lot of stats around how Compass has grown, how Jira Product Discovery has grown.
And we continue to run that from an R&D perspective with a very high level of discipline around. We will seed multiple ideas. The ideas that look like they're getting good traction, that are seeing product-market fit, we then fuel those further so that they can take off into different places. When we talked about the system of work, that really gives you a solid idea of what are the different radial directions in which we can grow revenue. The teamwork foundations, which are our three core products. But now we have Loom joining the teamwork foundations, which is another universal product that really serves different kinds of teams and has a tremendous potential going wall to wall across different kinds of teams. Also a universal use case, just like we have with Jira and Confluence.
And on top of that specialized collection, so Compass has seen phenomenal customer interest and adoption. So has Jira Product Discovery, also a rapid grower. And because of our strong investment in platform, which is a consistent thing that we have made year-over-year, it allows us to create even more such new sources of revenue. And lastly, in the platform itself, we are able to create things like what Mike talked about in Rovo, not just as a first-party source of revenue, but potentially in outer concentric circles of second-party and third-party sources of revenue such that we build an entire system. So it's mostly revenue sources that we think about as a matrix, not just as one linear thing saying, "Oh, here's a list of revenue creators for us." Your second question around how do we think about that on R&D?
Typically, in R&D, we invest a little ahead of time to create future sources of growth. We've been talking to you consistently about how we've invested in the cloud platform so that we can bring server customers over to cloud. That has worked very well. As we see more of the migration work crest, effectively, what we're doing is we've already redirected some of that work into enterprise readiness, which is how we are able to make all of the deliveries around data management, enterprise performance, scale, which lets us bring more of the data center customers, larger enterprise customers over on cloud. A big priority for us is enterprise. We invest in enterprise scale readiness at the platform level. We see that benefit across all of our products. A second big investment area is around AI. We talked a lot about that.
So I'm not going to go into a lot of detail, except that the one thing that I would like to repeatedly underline there is we have a very unique set of data, a unique set of connected data within the Atlassian ecosystem. So we are going to be able to do something a lot more unique and differentiated than any other vendor can do. And we are seeing more strength in that area with breadth of adoption of different teams. So that graph underlying our product set is getting stronger and stronger. So that's a great virtuous loop for us. And therefore, that's an area that we will continue to invest in and experiment to see what are different kinds of sources we can create.
Lastly, the investment in platform, I often get this from investors around, "Hey, when will it stop?" The way we think about it is not that it's a finite start-stop, but more around how can we keep redirecting it to different areas of the platform such that we can lift all boards, such that we can lift all of the different products built on top of the platform? You continually see proof points of how that is helping us create new sources of revenue.
If I can just add on one small thing, I just want to connect sort of two of the new thoughts. You've seen us have a, well, at least almost a decade as a public company, 2 decades as a private company, history of balancing durable growth with balanced profitability, right? We've always been incredibly capital efficient. We continue to be so, I think. And durable growth, one of the reasons that comes is making great products, but taking the time to make them properly, right? I had someone ask me, "How come it took 2 years for Jira Product Discovery to kind of get out of Point A?" And the answer is because we wanted to make it great. We wanted to make it really work with customers.
Because part of the flywheel business model and everything else that we have is the product has to start by being really good, right? I would argue with both Jira Product Discovery and Compass, they've reached that point of customers saying, "I want this," right? One of the pieces of customer feedback I got, or internal feedback about customers, was people showing up to the Compass booth yesterday and saying, "My developers have told me that I need to come and ask you about Compass. What is it?" I'm like, "Awesome question." That's not a negative question. That's a super positive question. That is literally the question that we spent 2 years plus, I don't know how long Compass spent, 18 months, because that's the durable growth. That's the customer that will come and they will buy Compass over the next 12 months and start in a small way.
And then they will grow over the next 4, 5, 6 years in a durable way alongside Jira and other things. And that is by design how we want to be. We want to be durable in that growth path for that customer. I love that question. But we do take time to make sure the products are great when we ship them.
We have time for just one quick one. Jason.
Hi. Thanks for fitting me in. I think, Scott, you said something interesting about the mid-market for ITSM and how it's still quite fragmented. I hope you don't take this the wrong way. But what might investors not fully appreciate about maybe being the number two player in ITSM? Or what do you think is underappreciated about that market?
That market for us, like the ITSM market for us, is just super exciting because, you know, we're winning market share in every front there. And I do believe the reasons we're winning are very, very durable reasons to win. We, in some cases, have a price advantage, like, you know, in some cases, significant price advantage over incumbents to the point where, you know, it pays for itself, you know, to migrate to our products, like in six months, three months, nine months, like, you know, the ROI is there. We're much more usable than the competitor products out there. If you go talk to users of either our products, they'll tell you hands down, like, that our products are way easier to use. Three is that we're tied to development.
And when you talk to our customers, they say, "Well, why don't you choose Atlassian?" Well, only Atlassian can actually tie our customer requests to solving them. And no one wants to, you know, just collect customer tickets. They want to actually solve them and solve things on behalf of their customers. And increasingly, that's a software problem. It's a development problem. It's a technology problem. Atlassian can connect those ones. And so we see things like, "Hey, we used to use this third-party vendor, but all we did was then copy the request into Jira. And wow, we can do that much better if we're using the same product." And lastly, the platform that we've built out, like I showed some stuff this morning around automation. Now, you know, other vendors can build all that stuff out.
But because we are amortizing those investments in AI, analytics, automation, Forge as a technology platform, all these things come for free. Well, for free, we had to spend a huge amount of investment. But effectively, we can amortize that investment across all of our products. And so you see in the ITSM market, our feature and functionality is significantly elevated compared to competitors. And again, we can deliver that much more efficiently. And like, it's much more durable. It's very hard for competitors to catch up. So super bullish about that. And again, I don't hear of customers moving off that product. I hear a lot of customers moving from other products to us. So I feel very good about that.
Maybe a quick one to add to that. The way you asked the question, often we get asked the question around service management as purely IT service management. But the way that Jira Service Management works really is it creates new opportunity. You heard back from the Rivian folks around how the previous tools that they displaced wasn't a large legacy vendor necessarily, but it was Excel sheets and email. So we also serve a long tail of use cases that are adjacent to IT. And it's more of a patient monetization model where it surrounds all the breadth of different teams and then also enters into IT. So we have that bidirectional advantage that other players don't.
All right. I think we're going to jump into a video. Then we'll move you on to section three. Thanks.
So we're switching seats? Whatever works or something.
All right. So happy to jump into questions. Actually, I knew it was mentioned to me that Jackie had a question for herself. I thought it's relevant to bring up with the broader audience. I don't know where Jackie's sitting.
Can you just hold on?
Oh, there. Jackie, do you want to ask a question?
Sure. I'm super sorry to see Scott moving on, though. I know at some point this was going to happen. But I would really love for you all to walk us through kind of how the management team will be taking over his responsibilities and kind of how you expect things to work going from here.
If you don't know that I'm leaving Atlassian, then this is going to be very awkward. So yeah, after 23 years, I'm going to hang with my family. I got a young family after my son's 11th birthday is today. I just got off the phone with him. And I'm super excited to be able to be in a position to do this because Atlassian, I feel, is in such a solid position and on so many different fronts. We've finished server-end of support, which is a huge project that we've done. And we executed that as close to flawlessly as anything we've ever done at Atlassian. We've spent a decade building out a cloud platform.
And that cloud platform is now seeing that sort of momentum in feature delivery, whether that's new products like Atlassian products hitting real kind of growth on revenue and usage, whether that's things like delivering data residency now pretty quickly, or the pace of the AI features that we could deliver. We've got our cloud platform to a stage where we're really getting ROI on that. And I can tell you that that is a huge investment under the curve before you start seeing that turnaround. The Teamwork that I shared earlier in terms of the opportunity we have ahead of us, particularly in enterprise, I could go on and on. But so for me, it feels like a perfect time to be able to pass the baton.
As much as I'd love to say, well, Atlassian is going to be totally different with me gone, like that would be great for my ego. At a practical level, Mike and I have made decisions together for the 23 years we've been working together. Those decisions are everything from which markets to enter. Back when we decided to enter the ITSM market, we made that decision jointly all the way through to what is our growth profile going to look like? What do we want our margin profile to be going forward? All those decisions are made jointly over a long period of time. I don't expect there to be significant changes with me stepping down. Of course, Mike, you can talk a bit more.
Sure. So many ways to address this question, I suppose. Firstly, I told someone yesterday that we're heading into this new phase with sadness and happiness, but no fear, right? And they said, what do you mean by that? I said, well, we start by being all very sad that Scott won't be around, right? Just on a human personal level, I think that's been the biggest overriding feeling from the company is just, man, not that anything's going to change or be different, but just a sense of sadness. It's a natural human emotion, right? After 23 years for me, 30-odd years almost in terms of interactions, although it's not like he's dying. So that 30 years continues, I guess. I'll know him for 40 years in 10 years' time. That part will continue.
But the 23 years as an operating executive with Atlassian, if you want to call it that way, there's a sadness, right? And that's around the company. And that's a totally normal thing, right? We have many, many employees who've been around for more than 10 years. I think the happiness is the happiness for him, right? He's making a great choice, leaving Atlassian in a fantastic position, bullish about where we are. And that happiness feeling is mixed with the sadness. But it's a balance. The no fear is like, I don't think we have any fear that Atlassian is going to be a different place. We're going to have suddenly different culture or values or a mission or strategies. We're really well set up for the long term. We make very long-term, pragmatic, thoughtful decisions. We will continue to do that as a philosophy.
But we are well set up in the current period to go and do that. We know exactly what our priorities are. And I think that's sort of bounds of emotions. And then I think from all of our point of view, just a thanks, a gratitude, right? I said to 11,000 Atlassians that their overriding feeling should be gratitude and thanks for Scott being a major contributor, the arguably major contributor to getting Atlassian to this point over the last 23 years. And all the things that we take for granted inside the company, small things all the way through to big things, he's had his hands and fingerprints on creating. And his expectation is that those type of things continue, right? Whether that's the values or whether that's other things. So I don't know if that helps from a human point of view.
From a practical, operational, pragmatic point of view, I know a lot of you very well. Like the human stuff, yeah, I get it. Now let's get to the operational details. As I've said to the company, the only person whose job is changing is mine, right? It's actually quite simple. I've never not felt 100% accountability for the company. I suspect Scott has never not felt 100% accountability for the company, right? It's not like we sort of felt responsible for 50% of things. The biggest change in my job is taking on go-to-market, right? Well, I built and ran that for 13 years. So I feel very comfortable to be able to take on sales and marketing and customer success and all the functions that we have there.
So we've both been intimately involved in finance and intimately involved in talent, whether that's recruiting, Team Anywhere, whether that's our culture and workplace experience. These are areas that are probably some of the most shit areas in terms of the culture of the company over time. So from a comfort level, I feel extremely confident. I know you're all going to wait and say, see the results. That's totally fair. That's what I would expect. From an executive team point of view, we have the most experienced exec team we've ever had, right? We have incredible average Atlassian tenure. We have great experience in the enterprise market as a broad whole over careers. So I think the exec team feel very confident. I mean, these two can speak to themselves. Only two of them are on stage at the moment.
But we feel very confident we can continue to take Atlassian to better and better places over time. I don't know if the exec ops members want to chime in.
That's fair. That's true.
We'll start with Arjun on my left. BK is actually quite an accomplished triathlete. So it's good to see you jogging down the aisle there, BK. Well done.
Thank you for that. Two questions, maybe if I can. One, to start off with, when you think about the opportunity to expand within your customer base and the 20% medium-term CAGR that you've laid out, how much of that is predicated on seat expansion, maybe coming back versus the maybe bigger cross-sell opportunity that you have with the product breadth of the platform now? And then two, I'm just curious how you think from a product perspective broadly about verticalization and whether there is an opportunity to do that, whether it's JSM or Jira or other products that you have in your portfolio. Thank you.
Yeah, great question. Let me start with the first one. In terms of the 20% CAGR or greater than 20% CAGR that we expect, that doesn't assume any improvement in macro. So from that perspective, I just want to be clear that it kind of assumes stable macro conditions to what we see today. We do expect that we can inflect paid seat expansion for all the reasons Anu talked about earlier. It's the ability to position our products to apply and relate to broader teams in the organization. It's the penetration in enterprise. And it's really positioning us to address more wall-to-wall opportunity than we have in the past. We believe that's going to be the fundamental driver of improvement on paid seat expansion and both on the cloud and in data center.
Yeah. And in terms of the verticalization question on product, the way we think about our portfolio of products is through the System of Work that we talked about. So the Teamwork Foundations has a universal set of products that we believe have the potential to go wall-to-wall across every organization. And we're seeing several proof points of that with customers today taking these products wall-to-wall. So I would think of that as we have a set of universal products that every single team in your organization can consume. And we build on top of that foundation. So the verticals that we think about on top of that foundation are for specialized teams, the ones that Scott talked to in his keynote today, specifically for software development teams, for IT teams, for service management teams.
I think we will continue to look at what are those adjacent verticals that we can keep building on because we have the luxury of having a common foundation that we continually invest in. A second lens to look at it is in terms of verticalization by industry, which we do today through our solution partners, through our marketplace partners, and the ability to go do templatization and configuration across our horizontal products. That's always been a strength for Atlassian. That's why we see the kind of stickiness that we do, especially across our existing customer base.
Two tiny add-ons. We do verticalization here. For example, we have an automotive customer group that get together where we connect our automotive customers together. We do a little bit of that alongside some of those partners in the industry-specific things. The second one is when Anu talked about templatization, just because no one's mentioned AI in about 11 seconds, I wanted to add in you saw a great example from Scott on stage about how we're going to use AI to continue to enhance those templates, right? Our templatization used to be human-driven, right? Create an HR service desk, et cetera. To be able to use AI with all the templating capabilities we've ever had to make more detailed templates for what you need is a real strength. We're only just at the cusp of sort of exploring that.
But you saw some of it in Scott's keynote, how we can verticalize our teamwork foundations products and some of the more broader products like JSM using AI. I think it's a really verdant area for us to keep exploring.
Then I know Arjun mentioned cross-sell as well. Do you want to touch on cross-sell a bit more, Joe and team?
Yeah, absolutely. As you've seen the strong product portfolio, the strong roadmap of innovation we're going to be rolling out. We're going to have increasingly relevant products that we're going to be able to cross-sell into that installed base of customers and seats. And so that will be definitely a lever that we'll have going forward. It'll probably be more impactful in the future than it is today, just given the level of innovation that we're going to be bringing to the market over the next three years.
Why don't we go with Greg right there, Wendy?
Hi, it's Gregg Moskowitz with Mizuho. Thanks very much for hosting this event. So the guidance for a revenue CAGR of 20%+ over the next three years, along with surpassing, as you said earlier, $10 billion within five years, which I think is a CAGR of 18%+, both of these are better than investors, generally speaking, I would say, were expecting. Two questions around this. First, how much go-to-market focus would you say was being diverted, if you will, towards migrating Server customers ahead of end-of-support? And then secondly, when I look at your business today, it's growing very nicely, very significantly. But you still have less than 500 customers that are spending $1 million or more. Feels like it can go a lot higher.
How much might the go-to-market need to change to drive a lot more wallet share with enterprises that will help propel you to these financial targets?
Let me jump in that one. It's my final go-to-market question I get to answer. I think the answer is yes to both of those. Let me get into more detail. We have prioritized migrations for our customers over the last few years because that's the gateway to all the R&D goodness that we're building in cloud. And so there's no point trying to cross-sell them something in cloud if they're not in cloud in the first place. And so that does mean there is a sort of disproportionate investment in the go-to-market motion to help support that. And that's everything from migration managers. And that's getting people to have all the pitch decks and so forth to convince customers to make that choice. Now, we still have a large number of data center customers that are moving. And that's not going to go away.
The proportional investment I expect to slow down over time in terms of the amount of effort we're putting into migrations. We've built the pitch decks to convince people to come across. We've got the migration management, even at a technical level, which is not a go-to-market, but the R&D investments we've made on all the tooling to move people across, where again, those investments can moderate over time. So what we're now seeing is that those go-to-market investments and those people have more time to have conversations with customers. And so that's from two aspects. One is that disproportionately, they're spending less time on migrations. And two, more of our customers are already on cloud, which means that then they've kind of passed through the gateway. And we can sell them additional products.
I do expect for us to get better at that sales motion over time because we have been investing in migrations. I think that's got a long way to go before I say we're world-class at all those things through a sales motion. So that's the first part of that. On the second side, in terms of large customers, you said we're 500 over $1 million. Enterprise is just such a big opportunity for us out there. And I talk about internally that to double our revenue doesn't mean we need to double the number of lines of code we have written for our customers. And in the enterprise, there's a huge opportunity for us to have high R&D leverage in what we deliver to those enterprise customers and commensurate high ARPU for those things that we deliver.
And so I do expect to see sort of both R&D and go-to-market leverage over time as we get better at serving those enterprises.
Why don't we go up front here? Marshall?
First to comment, Joe and Martin, this has been a great presentation with a lot of incremental disclosure that's really granular and helpful. And thank you for that. It's fantastic.
You're welcome.
Then a question for Anu. If you do the rough math, it looks like your ARPU for Fortune 500 customers is about $900,000. G2K x Fortune 500 looks a little under $750,000. You came from a company where those numbers would have been much, much larger. When you think about the headroom in ARPU, what needs to be done to get there? What is product? What is go-to-market? What is just sort of behavioral, longer-term pricing, et cetera? And when we're at $10 billion, what do you think the mix looks like? Is it heavier to G2K, or is it similar?
Yeah. Thanks for asking the question. It's a great one. It's one that we talk about a lot because that's where we see the opportunity in enterprises, as we've told you. The way I think about the customer shift, just answering it from the reverse direction, is that like Scott talked about, we already have a lot of these enterprise customers using our products. It's a very different sales motion from the place I came from for the rest of the room. I came from Microsoft. It's not like we have salespeople having to go to a brand new enterprise and start telling them, "Hey, here's a company called Atlassian. And here's something called Jira." It's not like the enterprise sales motion typically that other vendors do.
But as we think about enterprise customers, what do we have to do at Atlassian for our very unique business model? It's very much, how do we sell them the enterprise edition of our products and help them go wall-to-wall such that they see the value of deploying our products wall-to-wall and deploying specialized solutions of our products for specific teams? For IT teams, our JSM and related products. For developer teams, our Jira, Bitbucket, the rest of the developer products, which is a slightly different GTM motion. And like Scott mentioned, our first step there was getting them to cloud because that's where we have all these additions and different collections of products, so to speak. So I'm very pleased with our progress in taking step one, which is getting enterprises onto cloud. And that required significant product investment, to your question.
So from an R&D perspective, we've made multi-year investments on performance, on scale, data management, compliance. All of these are pretty much checkboxes we needed to first check off in order to even get to the 500 number you were talking about. Now, that R&D investment is continuing. And like I described earlier, from migration, we are able to take some of that and redirect it to enterprises. And I'm confident that things like FedRAMP, HIPAA, a lot of these are ongoing investments that we will continue to deliver on and therefore be able, from a product perspective, to have a really strong offering to get more and more enterprises to adopt us in the way that we would like them to.
On the GDM side, I think this is an advantage we have that we are not going to have to do these cold calls and suddenly start selling these at different places. But like Scott described, there is a way to go there. There are some ways that we need to cover in terms of what do we do on GDM, not just on sales, but pre-sales, post-sales, customer success? How do we actually scale this with the efficient model we have established? As a company, we are very disciplined about making sure that we take incremental evolutions there. And like we've described, we think taking incremental evolutions on the GDM model still puts us very much within line of sight of what needs to be done to get those enterprises on our platform.
So together with the central platform and the continued investments in R&D and the evolutions in GDM, I think we will be in a strong place to increase that 500 number.
Let's go to the guy on the left here, sir.
All right. I tried. Thanks. Joe, a couple of questions for you. First, on the marketplace, if I remember correctly, you've eliminated your takeout of the cloud apps in order to drive your marketplace to shift to the cloud. When do you turn back on? Is this a perpetual zero, or once everybody's there, then it's time to take your cut out of that as well? And then second question, in your commentary about the impact of data center to cloud transition, you talked about mid-high single-digit impact on cloud growth points coming from this transition. But can you talk qualitatively, what % of your data center seats do you expect will convert in this three-year time frame, if not an exact number, which will be great?
Perhaps qualitatively, the majority, anything that could give us an extent of how much of the Data Center do you think truly moves over? I appreciate it.
Yeah, I'm not going to be able to give you a specific number on the seat count. I think the mid to high single-digit estimate of cloud revenue growth coming from data center migrations is a good rule of thumb that you can use. What gives us confidence on that is just the incredible size of that data center install base and all the things we're doing that Anu mentioned earlier to enable our customers to move to the cloud. On the marketplace cloud take rate, we'll continue to keep the take rate as it is for the foreseeable future because it is driving the right incentives around creating cloud apps and building momentum around that. So for the foreseeable future, we don't see a change. But we'll thoughtfully evaluate that and be pragmatic as we see that play out over the next few years.
Maybe just a follow-up on the data center to cloud. In three years from now, what would be a reason for a customer not to move? Do you believe that all the scale and enterprise features that you are still trying to put in will be largely in place? Would it be a reason for anyone not to move into cloud in three-year time frame?
I'll take it from the backwards. Given the size of our customer base, will we ever reach a singular point where every single run of our customers has no reason to move? Maybe you could argue no. Right now, the majority of our Data Center customers, I would argue, have a reason to move today. We're continuing to do things like delivering on our FedRAMP commitments, as an example of many. We have a public Cloud roadmap, which you can all see. And we're delivering against that roadmap. Those are going to continue to take away any impediments there may be for customers or some portions of those customers not to move. Again, to my exhortation, Data Center customers are not singular entities, usually.
Even as FedRAMP, as an example to continue that thread, other than big government or obvious sort of customers, there are a lot of customers who say, "Well, part of my business needs FedRAMP, and part doesn't." Great. So part of it's already moved. And some's going to move in the future, et cetera. So look, we're continuing to work really hard on that. In three years' time, I would say that number will be even higher than a majority of customers today have a reason to move. We also want to make sure those customers move at a pace that makes sense for them as well as us. Will we have more compelling offerings in the cloud in three years? I guarantee it. I can guarantee our cloud offering will be more compelling in three years than it is today.
So that will continue to give them even more reason to move over time. We've only had two—what have we got here? On Sunday, Monday, Tuesday, so a couple of days with customers. Again, I have not run into a single customer who said, "I'm not going to cloud." It's just, "Here's my time frame. Here's my schedule. Here's my thing. Sometimes, here's a list of things I need." But more often than not, I get it. Wow, you guys have done a great job with those sorts of things. The scale improvements really helped me. That's a good example. Or data shares is a good example. The external data sources and analytics, some of these things that may not be broadly understood.
If I can explain the analytics example, a lot of customers used to go straight to the Jira or Confluence database with SQL to do effectively analytics. And we don't let them in their cloud go to our databases. And sometimes, we don't even have a database nowadays. So we need to do the data modeling to put in the data lake and give them a SQL-like interface and analytics. It's a good example that's an unblocker for those large enterprise customers that is not often talked about. But our continued investment in analytics, I can tell you, is really resonant. Two or three customers have already mentioned to me, "Wow, you guys have come a long way in 12 months in analytics or 24 months since announcing, I guess, 12 months since GA." So a lot of those capabilities will make a big difference.
We have a financial-oriented question for Joe. I think it's a good one to address. In the medium term, how do we think about free cash flow margins relative to operating margins?
Yeah, thanks for the question. Typically, we will see free cash flow margins correlate with non-GAAP operating margins. We typically see about a five percentage point gap between those two. It may vary period to period, just given timing. But for the medium term, we expect to maintain that type of spread between the two. So you can think of it as non-GAAP operating margins plus five to six points to get your free cash flow margin estimate.
Sandra in the middle. That might be challenging for you, Wendy.
Sanjit Singh, Morgan Stanley. I want to echo Marshall's compliments on all the great disclosures today. That's been super helpful. Then Scott, it's been awesome working with you for the past 10 years and an outstanding career. All the best. I know we're going to have plenty of time to say your goodbyes. All the best in your future endeavors. I wanted to talk about two topics, one on growth when we sort of frame out the TAM, the serviceable, addressable market, as well as the 20% growth. To what extent are you guys anticipating any sort of headwinds on seat-based pricing? It was a question I asked on the earnings call.
But if you're driving a lot of productivity, let's say, in agents and customers don't have to hire as much as they did before, don't need to expand, how are you guys thinking through that as part of the overall growth equation? I know you can expand to more teams to address that. But just on the seat-based pricing debate, what's sort of the team's latest thoughts on the revenue side of the equation? And on the margin side of the equation, I wanted to get a better understanding of now that you don't have a server base to support, how much capital, along with productivity improvements with your engineers, how much capital does that freeing up to invest in the go-to-market enterprise? And how will that ultimately drive margins to that 25%+ target over the next couple of years?
Let me take the first question, which is, I think, on everyone's minds: AI drives productivity. If you're a SaaS vendor, you charge per seat. So therefore, if I need less seats, what does that look like for the growth out there? And I think everyone asks other vendors out there know that there's some sort of usage-based pricing in our future out there. Now, we've looked at this internally. I don't think we have anything out there publicly. But this is something we're looking at in conjunction with our customers as to how do we deliver value based on usage-based pricing? Two other things that I think are interesting that I think we have advantages relative to other SaaS vendors. One is that we seem to expand the market for what we do compared to many other vendors.
And you saw we listened to Rivian up here, where they have 1,200 seats of a service desk in a company with 15,000 people, roughly. So one in 10 people have a service desk. And every other vendor, it's sort of out by an order of magnitude or more what most other vendors would have. And so we have a history at Atlassian of effectively expanding the market by being more aggressive on pricing for what we do. And so that helps us a lot. Maybe Rivian saved 10% of their seats by AI. But we're 10 times larger than our competitors. The second one is our per-seat pricing on a monthly basis is just significantly lower than many of our competitors out there.
So the thrash and difficulty to go through that transition from a seat-based pricing model to a usage-based pricing model is going to be a lot easier for us. If we're charging $20 a month for a service desk agent, someone else is charging $200 a month for a service desk agent. When you switch to usage-based pricing, if someone's like, "Well, to do that, it's going to cost you $4 every email you send," it's going to be really difficult for customers to swallow those types of usage-based pricing. But when we're a lot cheaper, "Oh, it's going to cost you $0.04 to send an email," hypothetical numbers. But the idea is, it's going to be a lot easier for us to make that switch to consumption-based pricing. I don't think anyone can stand to say exactly what that will look like over the period of time.
But when I look at the value we deliver to customers, AI so significantly increases the value we can deliver for customers that the optionality there, for however we price it, puts us in a way better spot than we are today. So I don't know. Without knowing all the details of what that looks like in five years' time, I just feel really confident about that being a tailwind for us rather than a headwind.
And then on your margin question, you're absolutely right. With Server end of support, we do free up capital. We do free up resources that we're focused on that, both on the R&D side and the sales and marketing side that can be redeployed. By the way, we're going to have the same phenomena around the Data Center migrations, a lot of the work that's happening in the company around enabling people to move to the Cloud, enterprise-grade capabilities, certifications, reliability, scalability, all the things we've talked about, performance. Once we meet the bar and we deliver on those things, that does give us the ability to move resources within the organization. And that is built into that greater than 25% operating margin or the ability to return to historical operating margin. So you see that in some of the line items like R&D.
We expect that to be a smaller percentage of revenue going forward. That's a big part of it. You don't see it as much in sales and marketing because we are making the big investments in the enterprise-grade capability there. It is absolutely part of our resource reallocation process that we're going to play out over the next three years.
If I could speak in just one follow-up on the 20% CAGR, I just want to make sure that doesn't assume at least 20% over the next three years. That's a three-year CAGR. Is that the right way to?
It's a three-year CAGR. Correct.
It's a three-year CAGR. Yep. And greater than 20%.
Greater than 20%.
Yeah, 20%+ .
I'll go to Fred.
Hey, thank you very much. Fred Havemeyer from Macquarie. I was thinking of doing an AI question. But I think I'll give you all a break from that for a moment. So I want to go macro instead. I think when taking a look at some of the job postings in the marketplace for IT professionals, software developers, et cetera, it looked kind of like a Bitcoin chart in ways over the pandemic, went up a lot. And of course, it's contracted quite a bit. So I wanted to consider that 20% CAGR over the next 3 years and actually the $10 billion over five year sorry, $10 billion in revenue over the next five-year target against this backdrop and just ask, is there anything in your forecast or estimates that really requires a turnaround in the overall hiring landscape for software engineers, IT, help desk?
Or are we really looking at this as a function of Atlassian able to drive growth in a really diversified manner, and especially with the under-penetration enterprise? Thank you.
I would say it's definitely the latter. Firstly, I just want to be clear. It's surpassing $10 billion in under five years, which is a little different to $10 billion in five years, just if you do any math. Nothing in our forecasts is currently based on the market rebounding back in a massive way. I think we said that before. The numbers that we're giving out broadly here are based on a constant macro environment as a way to say it. I'll leave the Bitcoin analogy to others to determine whether that's accurate or not. But I just wanted to make clear that that's not based on a massive rebound in job postings or other things at all. We do drive diversified growth across our business. To the question on durable growth beforehand, again, we've always, for the longest time now, wanted to grow longer.
We still want to be growing strongly 10 years from now and 15 years from now. That's the way that we've set up the company to date and the way that we're intending to run into the future and continuing to do that in a capital and margin-efficient way as we've got a reputation for doing over a long period of time. As a part of the durable growth, it is about setting up multiple products and multiple streams and multiple markets, which are all growing, to some extent, connected, as we've seen from the dev and IT spaces, but also independently in their own ways. That's one of the ways that we get durability and resilience, I suppose, in our growth through market conditions up and down.
Let's go to Michael here in the front.
Thanks for taking the question. I would echo there's a lot of useful information in here that we've been asking for. Thank you for all of that. I want to spend some time on the cloud expansion rate that's in here, so 120%. It's really good given 70% still SMB and what's happening within that segment of the market. Can you help just contextualize some of the trajectory you've seen there, Joe? It sounded like from the most recent earnings call, seats are still a headwind. Maybe just help us with how much cross-sell and some of the new products and visibility in cloud is helping reinforce the growth, even with some of the seat dynamics at play there.
Yeah, thanks for the question, Michael. I'd say if you look at the cloud revenue growth drivers, the number one growth driver we've had historically is paid seat expansion. If you go back two years to our expectations for cloud growth and compare it to what's happening today, the delta is entirely explained by paid seat expansion change. And so the rest of the model continues to perform really well, whether it's migrations, whether it's cross-sell, whether it's upsell, pricing, et cetera. So the overall business remains healthy. It's just really that one piece of the business. And as you point out, 70% of that business is SMB. And SMB has been disproportionately impacted by the macroeconomic environment. And we really haven't seen it come back. Enterprise, on the other hand, is relatively strong. And so that part of the business continues to perform well.
The trajectory on paid seat expansion, as we talked about, it's lower quarter-over-quarter. But that trend is moderating. And so we hope to get to a point in the near future where we can say it's quarter-to-quarter stable. The thing that's holding that back is SMB right now. Longer term, we've talked about the drivers of what gives us confidence on how do we improve that. And it's really around moving into enterprise and this broader team sell and wall to wall sell. Cross-sell is doing really well. We talked earlier about the number of great new products that we're bringing out that really helps drive that cross-sell number. So as we bring new products and new innovation to market, as we broaden out the user base, that's going to become a bigger and bigger part of our growth profile going forward.
Couldn't be more happy with the premium and enterprise uptake in the Cloud. We'll continue to add value to that. And as we go into enterprise, as you know, those customers predominantly purchase those premium and enterprise SKUs. So just by virtue of the mix shifting over to enterprise, you should see more upside from an upsell perspective. And then pricing, we've just been kind of consistently raising prices as we've added value to the offer. You can assume that's going to continue going forward as we do continue to add innovation. So for us, that's how we think about the growth drivers. And then the last one, obviously, is migrations. We took a question earlier about that, where we expect mid to high single-digit contribution to Cloud revenue growth from migrations from Data Center to Cloud. So that's sort of the trajectory we see.
We do see some improvement on paid seat expansion. We do see cross-sell becoming a bigger part of the mix. And then we just see pricing and upsell continuing to kind of stay on trend relative to where it's been. Hope that helps.
Let's go to Keith. He hasn't had his question yet.
Hi. Thank you very much. Keith Bachmann from BMO. Joe, I had a couple for you to start with, just to follow on there. If you think about the growth algorithm, you've been heavily weighted to existing customers. And do you see that changing at all in the next three years? Well, let's just start there.
Yeah, I don't see that changing. I think with our land and expand model, we'll continue to land, as we always have, with that really efficient and differentiated flywheel model that we have. And so we'll generate a presence in an organization. And then we'll expand from there. So I really don't see new customer logos in a one-year period being a material driver. And so I do expect that ratio to continue to be dominated by our existing customer base.
OK. And then the following was on DC. And as you think about it, you've talked about mid-single-digit price increases. But should we assume that the price increases on DC will be significantly above that as you're trying to candidly shape behavior on maybe inducing that migration path? And just to add on to that, does DC grow over the next three years?
It does grow over the next three years. I think if you look at if you strip away the migration impact, if you look at the underlying drivers of Data Center revenue growth, they're actually very healthy. It's new customers. It's seat expansion within those existing customers. It's cross-sell within those existing customers. That's been very healthy. If you strip away all the migrations noise and some of the one-time event-driven purchasing, you'll notice that it's pretty healthy. Then pricing, obviously, has been a big part of that as well. I'll let Mike and Scott chime in on the pricing for DC. In general, our philosophy is we really price to the value of the offer. We have to be very thoughtful from a Data Center perspective because we do want to provide the right incentives for customers to evaluate Cloud versus Data Center.
But at the same time, we want to make sure that the Cloud platform and the Cloud offer is ready to capture those customers. Typically, those Data Center customers are our largest and most complex customers. As we add value to Data Center and as we think about the offer in Data Center, we'll continue to raise price in conjunction with that. But from a broader strategic perspective on incentives to the Cloud, we're going to be really thoughtful about the actions we take for that.
I mean, I would just say we continue to philosophically optimize price. I think it's been a strength of Atlassian for a decade plus, two decades almost. It's, I talked about the cloud take rate as an example. We don't always raise prices. We try to optimize price to where we believe the customers best have the demand. You've seen us doing that over the last few years in lots of different ways. Sometimes it's Jira Software up and JSM down. Sometimes it's a lot of different things. So I don't think philosophically, we'll be changing that in the visible future.
I'll just one last one. I promise I'll pass the mic. But Joe, why is now the right time to give a longer-term guide? Atlassian's never had kind of a longer-term revenue margin guide. But why is now the right time?
I think we've always been a company that's managed for the long term. And I think the point we want to make is there is a durable long-term revenue growth and a profitability profile in the company. We've committed to returning to historical operating margins. A very legitimate question we get from the folks in this room is, can you put a time-bound aspect to that? You've left that very open-ended. So we want to demonstrate and illustrate the commitment we have, the path that we see to being able to achieve that, and the timing of that. And we want investors focused on total revenue growth. We've got a great Data Center business. We've got a great Cloud business. Both of those are very healthy. And we'll grow over the next three years.
But we really want the focus to be not on each individual one, but in total from a subscription revenue perspective because really, we feel that reflects the true health of our overall business going forward.
I think it's also probably fair to say, and Joe, correct me if I'm wrong here. But I'm pretty sure we all feel this way. The visibility into that future is a little different than it was a couple of years ago getting through the end-of-Server support transition. And as you saw, Data Center migrations are up 90-odd% year-over-year. So the visibility of the major movements we have inside our world and getting through the February end-of-Server date gives us a different profile over the next few years than we've had over the last few.
Colin here in the front.
Hi. Colin Ducharme with Sterling Capital. Thanks for taking my question. Keith was a great springboard. I'll riff off the disclosure ask. Mike, Scott, Anu, you all projected very confident views on the durability of growth, great disclosures with the new products, et cetera. So thank you for all of that. My question is, there were several of us in the room today that have been positively surprised by some of the product-level metrics disclosed in the documents here. And I guess my question for you, Joe, is, is there an opportunity from a disclosure standpoint as the shape of the growth begins to change and move more towards cross-sell to provide a revenue attribution by product, by customer type, a more regular NRR update? So that way, you can, I guess, true up the market expectation to your own confidence of the durability of the growth.
Yeah, it's a fair question and a fair comment. And that's feedback we'll take, Colin. The reality is we will change and evolve our reporting as our business evolves and changes. The metrics, the communication that we had in the past when we were behind the firewall company should change as we get more towards the cloud. So as you all know, we love feedback like that. We take that into account. And we try and evolve and adjust our communication and metrics and KPIs to be more relevant to the business that we have. And so we'll take that into consideration as we think going forward.
Thank you. Then a real quick follow-up, more specific on the SBC line item, you guide towards 25% with a slight glide down over the medium term. Can you give us, I guess, a qualitative view on how you think about that gliding down over time? Because there are companies that are at your scale or perhaps larger that grow at rates or higher that have a lower SBC percentage. And that's a great leverage item for your GAAP profitability over time.
Totally. You're thinking like I am. Absolutely. So great question on this one. So stock-based compensation, I would say we think about this much like we think of other elements of compensation expense, which is attracting and retaining the best talent possible. Nothing has really changed or will change fundamentally about our compensation structure and framework, which is all market-driven. If you look at that stock-based compensation line today, that reflects decisions that were made two or three years ago on grants in a totally different market environment. So we actually think we have good levers in optimizing the plan to reflect the current market reality that we're operating in. Then, as we've pointed out in the docs, we think we have the opportunity to moderate headcount growth relative to revenue. Headcount growth is really a primary driver around stock-based compensation. So those are the levers.
We're highly confident we can get there. You're right. When we think about setting goals, we look at our peer comparables and say, hey, we should be able to get on a glide path down to that level of stock-based compensation as a percent of revenue. There's no reason we shouldn't be able to do that plus or minus. That is the plan going forward. I wouldn't read too much into the slope of the arrow in terms of the North Star and the goal that we have. It really is focused on looking around at peer comparables and benchmarking ourselves against that and putting those plans together. You're right. You're absolutely right. That is a key lever in our goal to achieve sustainable GAAP profitability.
Let's take one final question, I guess, Alex.
Hey, guys. I guess maybe without giving I appreciate the focus on total revenue and really appreciate all the disclosures and the focus on durability of growth. But as we've seen the last or at least the guidance suggests an acceleration in cloud for Q4. And next year, you have a tough data center comp, particularly in the second half. So as we think about the slope of cloud versus data center, maybe over the next two years, not so much the years beyond that, is there is cloud accelerating and data center going to go like this? Or what's the right way for us to think about that dimensionally?
Yeah, we'll give more definitive detail on the FY 2025 guidance, Alex, on the earnings call in July, August when we report Q4 results. In the docs, we've tried to give you the drivers, the core drivers for each one of those. I would touch and just go back to what we talked about on the call, which is we do expect healthy revenue growth in FY 2025 from both Cloud and Data Center.
I talked about the shape of that Data Center revenue growth, that we are going to run into tough comps in the second half of the year, both from the comparison to the end-of-support event-driven purchasing as well as the fact that we're going to see the roll-off of Server migrations to Data Center. And so you're only going to have a headwind at that point, which is the Data Center to Cloud migrations. So that's that.
Then from a cloud perspective, we shared our drivers in the document. We'll talk more about those drivers going forward. Overall, we expect to see healthy revenue growth from both for the full year.
Scott, do you want to close us out?
Oh, yeah. We're finishing up. I think we're having drinks outside after this, I think. So I'll get to catch up with you one-on-one there. But this is my last, I guess, investor day on stage. And that's a bittersweet moment for me. And I want to say just a huge thank you to all of you. There's a few new faces in the room, but many, many, many faces that have been around on the Atlassian journey for close to a decade. And there are two things I think we really want from investors. We want investors that think about Atlassian long-term because that's how we operate as a business. And we want investors that spend the time and effort to understand what we do as a business and be informed about that. And I could not be more blessed.
And we could not be more blessed to have investors like yourself who think long-term and invest the time and effort to really understand the Atlassian story. And I know many of you have traveled. I think some have traveled all the way from Australia today to come be here. And I just want to say from the bottom of my heart, thank you. It's been a great decade for me. And again, thank you for all your time today. I hope to see you at drinks afterwards. Thank you, everyone.