Atlassian Corporation (TEAM) Earnings Call Transcript & Summary

Hi, everyone. Thanks so much for joining us today. I'm here with Tiffany To from Atlassian. Gili Naftalovich from Goldman. We're really, really excited to have Tiffany here with us as SVP of Enterprise and Platform for Atlassian.

Tiffany, you have been with us with Atlassian for almost 5 years now working across the company's different solutions. Can you tell us more about the role you have held in the evolution of Atlassian over this period?

Sure. Thanks so much for having me. So I joined Atlassian, like you said, about 5 years ago to head up the product on the Agile and DevOps side. So I headed up Jira and Bitbucket, core cloud products. But as I started meeting with customers and ask them about what they wanted from our product, one of the really big recurring themes was that they wanted us to help them be better tech companies. So it wasn't just hey, we buy our products, they help our dev teams manage the software development process. It was how do we know how the work is flowing. How do we know that we're doing a good job. How do we know we can be -- like how do we be a better tech company. So a lot of that pointed me towards investing in more solutions around the platform. And so after a couple of years, managing the product, I transitioned over to head up the Platform team. And that kind of aligns with how Atlassian has also evolved their enterprise strategy. So I think most folks are aware that we made a very strong strategic decision to move our customers to cloud. We announced the end of life of our server products, which is where most of our customers were, and over the last 5 years, customers have been going through this journey of working with us to figure out what's the right timing for them to migrate to cloud. And so building some of these new use cases that I talk to those customers with 5 years ago, those are now a big part of the business case for them moving to cloud. And then we'll get into it, but AI becomes like a whole another accelerant that is exciting to talk about.

Yes. We will definitely talk about Atlassian Intelligence and everything revolving around that. But even before we get into all that, collaboration has always been a really key component of the Atlassian platform and across a variety of different end markets. How would you characterize as that key differentiation that it allows you to do and approach that a lot like we can continue Atlassian's strong market position?

Yes. I think what's interesting is a lot of times when people think of Atlassian, they think of Jira and they think of us as a dev tool. But if you look at who's using Jira, only half the users are developers. The other half are folks that are product managers, program managers, marketers, all of the other functions around building and developing software. And so when you talk about collaboration, I think one of the really interesting things when you think about technology and development is that, it's not just the developers, right? Like if you want to be an innovative tech company, you actually have to harness the value of the whole company, all of these other functions. And Atlassian has always embraced that. And that's why we -- after Jira, we built Confluence, and you've seen us make a number of both new things we've built, but also things we've purchased, things like Loom, right? And that's because we truly believe that to be the best technology-driven company, you have to enable collaboration between the technology and the business teams. You actually have to connect those groups together. And so to differentiate we want to both have the best tool for that range of different personas, but we also want to invest in a platform that connects the data across those tools. And so one of the big investment areas that we built in platform is something called a teamwork graph. A graph is not a new concept, don't worry. I won't go into like a whole engineering description, although that's my background. But think of it as there's been this huge era of, hey, you have tons of data, try to get value out of that data, right, the whole big data movement. And when you talk to customers and ask them, did you get value out of that big data movement, very few customers have. Why? Primarily because to take that data and actually organize it, clean it, enrich it and make it usable, it's so expensive that very few industries other than, I think, like retail, manufacturing, there's a few industries that have gotten real value out of big data. Now why do I talk about that? Because when it comes to collaboration, one of the hardest things to do is to provide context between people, right, especially if you're on a different team or you're in a different function. And so in Atlassian's platform, the teamwork graph is actually where we manage relationships between the data that comes from our products as well as third-party products. And we do that in a very opinionated way. And we anchor on 4 key concepts that we believe for good teamwork and a tech-driven company, you have to really understand who's doing the work, the teams, right, the goals. So what are the business goals the company is aligned to? How do you map those teams to the goal? The work itself, so all the projects that tie to those goals; and then finally, knowledge, all the different ways that you may store data in your organization and how that maps back to the project. Why is that important now? We couldn't have predicted the GenAI huge ways. But when you think about how to provide context, initially, a lot of that investment was to give context to the analytics for our users to be able to ask questions like what I said at the very beginning, where customers were like, hey, how are my teams doing, which teams are bottlenecked, which teams need help. That teamwork graph helps give the context to answer those questions. But now with AI, the teamwork graph ends up being a really key differentiator for how we evolve those models, those AI models because the AI models can layer on top and tap into that teamwork graph. And no one else in the industry has that. Microsoft is well known for their graph, but it's a much more generalized graph, whereas ours is very focused on those teamwork objects.

What is exactly the teamwork graph enable or unlock in terms of capabilities within Atlassian platform?

Yes, the way that it shows up is when a user logs in, it knows the context of who that person is, what team they're on, what projects they've been working on. So it's going to proactively bring them suggestions on what work they should pick up what's happening in other teams that they should be aware of because there's dependencies between their projects, right? It's going to pick out of the sea of knowledge across Confluence or SharePoint or whatever tools that you're using, relevant knowledge that would make sense for that project. So think of it as like a way to have almost a helping hand that's very customized to each user. And that shows up in different ways across the different products. But especially with the AI features, you really start to see that difference because it picks up on the specific ways that, that company does their particular technology development. So I'll give you maybe a more concrete example. When you apply the AI features, the generalized models that people are using, obviously, is going to look at all data and make recommendations. But with the teamwork graph, we have an awareness of the way that, that particular tech company does their work. So we know the way they write PRDs, right, the product requirement docs, we know their release process. We know what security checks they want to do before they can release software. All of that data then gets fed into the AI queries that we put into Atlassian Intelligence. And so what that means is when you build, for example, an AI agent, a Rovo agent in our portfolio, on top of the teamwork graph, it's going to provide much more customized guidance that fits for that organization.

Yes. And I definitely want to touch on Rovo, I'm just...

Yes, I'm jumping ahead.

Maybe a little bit, but it's a great context. And I think something that you said, I wanted to ask you a little bit about in terms of Atlassian Intelligence and if you're seeing greater usage of that between applications within Atlassian ecosystem or even third-party or within applications and unlocking the functionality of, let's just say, a Jira or Confluence, how does that look like today?

Yes. I think we're seeing everything right now, which isn't surprising. Like all the enterprise customers I meet with, they're like okay, we've got to figure out our AI strategy. They are often designating a counsel that is in charge of determining how to evaluate and adopt all these different AI technologies from different vendors. I think the great thing for Atlassian is we're usually on that top 5 list for them because we're the set of tools that is the system of record for some of their most valuable employees, their development teams and all of those teams supporting those groups. And so the AI features that we've built, they're in kind of 2 arenas. One set of the GenAI features are kind of integrated across all of our products. So think of it as any function that you're already doing in Jira and Confluence, we turbocharge those things, right? So it's like you could read the page or you could read the summary much faster, right? You can write the page from scratch or you could have it pre-populate based on other pages it seem, your personal preferences, et cetera, you could fill out the Jira ticket or you could have the agents fill out the Jira ticket. So a lot of those functions are happening, and we're making those available to customers that are on premium and enterprise additions of our products. And we're seeing really huge growth of that. So about a 3x growth in now quarter-over-quarter we're seeing. We're feeding that growth, obviously, because we're continuously releasing more features. I think we have a couple of dozen features kind of in that arena of deeply integrated into the product. And then there's a second category, which is tied specifically to Rovo. And so there, that Rovo layer is both search built on top of that teamwork graph that I talked about, but it's also a development platform. So part of what we're seeing with customers is that they want our help in building the right agents for their organization. So we have a lot of great data from our 300,000 customers in the research we do, so we are providing a lot of out-of-the-box agent. So think of it as your product manager teammate, your developer teammate, your program manager teammate. They're very persona-oriented. So we're using data from our customers generically to kind of develop those functions. But then since we're building it on top of our forged developer platform, what becomes really exciting is customers or our partners can then take that agent and extend it to their organization, right? It's taking that rich context that I talked about, but then customers can start to make it work the way that they want in their organization. And that's part of, I think, where the real productivity is going to get unlocked over the next year.

Yes. And it's actually really interesting. I think Atlassian was one of the first companies that we really look at and at scale in the industry that really talked about agents and through Rovo. What led you into that direction? How fast have you worked to kind of get there to announce that Atlassian '24? What are some initial proof points and applications of that beyond the typical Jira customer or Jira use cases?

Yes. It's in a really fast learning loop. I think like everyone else, it's been all hands on deck, right, to try and like capture this learning as fast as possible and experiment with our customers. I think initially, our hypothesis was customers would want these agents like pretty much out of the box and they would use them. But as we first started experiencing, we realized that the true value was actually that last-mile customization because no one wants a generic developer or generic program manager, that's like hiring a brand new employee that comes in, may have functional experience and background, but doesn't know anything about your company in the way you work, right? And so what we realized was that last mile becomes quite pivotal, which is how do you codify the way your company works, whether those are things that are rules and regulations or softer things like cultural practices. So what we find is back to the beginning when customers were asking me, how do I be a really good -- the best technology company I can be. Some of that is in the tech and the tools, but some of that is in cultural practices. Are you running tech ops practices, right? Do you have the right rituals? Are you doing retros, right? These are actually less about the tools and more about how the way the teams work, and so for some customers, they're hiring coaches like humans to come in and teach these things. But you can imagine now with the Rovo agent, you can actually start to take some of those practices that you want to adopt it and have that coaching happening through the agent as well. So I think this learning has been, to your question, quite recent, but it's been a lot of fun because customers are giving us that fast feedback loop and the marketplace partners as well, as they're figuring out what is their role in this ecosystem. They're starting to do some pretty interesting customization where they're taking the agent and they're customizing them by vertical. So you can imagine like, okay, there's a generalized developer or a product manager and then there's custom to that organization. But there's also really interesting things that are done by, in an automotive organization versus financial services versus a hospital. So some of these partners that work with quite a few customers in these verticals are also providing that additional domain context on top of the agent. So lots to come on this front.

Yes. No, I'm sure we'll still be hearing a lot from you in that sense. But also, and when you think about how it's going to look in the future and you think about the workflows that are existing right now or the way that users engage with Atlassian, how do you see these agents kind of either complementing or automating some of this? And is it -- will you have a product for one versus the other? I think we hear a lot about chatbots versus agents. How does that -- how do you see that evolving within Atlassian portfolio?

Yes. That's a good question. I can't quite predict the future, but I will say that I think my view of it is that our customers are thinking much more holistically about their system of work, right? And when I say the system of work, it's how they think about the operating system that runs a technology-driven company, right? So they're thinking about what are those key workflows, right? What are the key teams in these workflows, but also who are the key strategic vendors that supply the system of work, right? They don't want 100. They want a small set of strategic vendors that take ownership and integrate that system of work together because I don't think they're going to have 1 vendor that does everything. And to give you a sense of who's at that table, like clearly, Microsoft has the horizontal collaboration tool. On the IT side, ServiceNow is there as well. But when you look at the core development teams and the system of record for those dev teams, it's Atlassian, right? And so I think when you ask me like what does that role start to look like? I think it's -- these enterprises having these multiple strategic vendors, it's them having a system of work that organizes several of these core systems of record into that system of work. And then they're going to be writing their own AI agents that help with different parts of those workflows. And I think for an Atlassian, it's going to be very much the connecting of the technology and the business teams together. And so that's both a vertical up and down of, hey, what are the goals of the company, right? Which projects are we betting on that ladder up to those goals, and then what teams are working on those projects so we can track that work? And then Microsoft tools are there horizontally, helping with, obviously, collaboration. We integrate really well with those Microsoft tools and bring that into our teamwork graph. And then as you get into the delivery and then the servicing, then we integrate into a ServiceNow as well. So I think it's going to be a mix of multi-vendor and multi-AI strategy. Is that what your question was?

Yes, definitely it does. And it kind of leads a little bit into the next question in terms of thinking about when you talk about ServiceNow or when you talk about Microsoft, you often think enterprise, right? And I think that there has been a lot of changes into how Atlassian is going to market with the enterprise, but particularly a lot of the functionality that we're just talking about is only available in cloud, in particular SKUs that you mentioned earlier. How are you balancing these investments and the needs of these companies that are really, really eager to adopt AI with also the desire to move more of them to the cloud?

Yes, it's been interesting to see those conversations evolve over the last 5 years. When I joined Atlassian 5 years ago, we had just announced that, hey, everyone needs to move to cloud. We're building the cloud platform for you. And I think customers at that time were like, okay, "yes, we get it, and we'll get to it when it's a priority for us." In Atlassian, you've got to unblock all these other features, right? There are a lot of things they needed us to address from a scale performance, data residency, bring your own key, there's like a long list of reasons why customers often are more comfortable keeping their deployments behind the firewall. But as they saw us continue to execute on that road map and unblock and also increasingly add new value in the cloud on the platform, that conversation has shifted to, okay, we'd like to get to cloud. We want to start, help us build those business cases to get there, right? Because for a lot of our largest customers who have hundreds of thousands of seats, like Bosch have 600,000 they need to move, it becomes a multiyear journey. And so I think the evolution for Atlassian is that it's not a onetime type of thing. It's actually having a much deeper long-term partnership with these customers, but that is opening up a much richer set of conversations, like what I was just explaining earlier about the system of work. 5 years ago, we wouldn't be having that kind of conversation with these customers. It was much more -- yes, you guys are a vendor, you guys give us Jira and Confluence, that's great. The dev teams love it. Now it's, okay, we've got to figure out how to make our tech company, we're a tech-driven company, more effective. We've got to figure out our AI strategy. We have like 200 SaaS tools. We've got to rationalize those tools. They have a lot of these much more strategic initiatives that they want to take on. So that migration conversation has really transformed into, hey, how do we help you be a better tech company? How do we help build that multiyear business justification to move to cloud? And it ends up being a combination of things. It ends up being both near term, consolidate some of your apps as they inventory, they look at wait, I have 5 project management tools. I don't need that, right? That ends up helping make the business case for the initial migration. And then it becomes how do I deliver new levels of productivity with the analytics and AI and the other use cases that you guys are building in cloud. So I think of it as us learning to be a better partner to our enterprises. I think that's been the big shift for the last few years.

Yes. And when we think about that, right, I think you talked about multiyears and you talked about a 600,000 deployment, right? So that's I'm sure also across a variety of different applications, maybe mostly Jira, but maybe not only Jira, right? And so which applications are first moving to cloud or which applications or maybe department, I don't know how you might break it out or classify it. Should we be thinking about maybe moving to cloud like over time, basically and said otherwise, right? Like in other words, how do we balance the complexity of some of the Jira and Confluence with the usage of those and likely like the need for those to happen sooner rather than later?

Yes. There isn't -- there isn't a one-size-fits-all. Like I think in the first wave of customers that migrated to cloud maybe 2 years ago, it was very much a lift and shift, so they would take -- they would inventory their full footprint behind the firewall, lots of Jira and Confluence, oftentimes Jira service management, and it was lifting all of that, moving it to cloud, but a lot of those customers were smaller, like probably in the 10,000 seat and below. As we've shifted to the 50, 100, 500 (sic) [ 50,000, 100,000, 500,000 ] like 600 -- the Bosch example, 600,000, then they have to take a much more strategic approach to it. And so there -- it goes back to what I was saying earlier, where they kind of have to layer those business cases. Oftentimes, they're looking at, okay, what are the most critical things that we would bring across first and then build a business justification to move the next set and the next set. And so oftentimes, what we're seeing is they'll bring the core apps, Jira and Confluence, but they're also looking at building business cases to expand into other use cases beyond that. So growing the ITSM side of their use case, bringing more of those business teams in through that consolidation play that I mentioned earlier, right? When they're de-duping the Monday and Asana and bringing that into Jira, that ends up being an expansion of the Jira and Confluence footprint for us.

You're seeing some of those displacements?

We are. We are, yes. And they're happening not just -- I mentioned the business case aspect. And so it may sound like it's only for cost savings, but as I meet with a lot of our enterprise customers, they're also sharing that it's not always for budgetary reasons. So for customers like Amazon, LinkedIn, Adobe, it's not because they're trying to save money necessarily on the tools. What they've realized is in providing like full freedom for our end users to pick all of the tools that, that actually hasn't delivered higher productivity, it actually delivers more complexity and confusion because people are having to copy and paste and move a lot of data back and forth between all of these tools in the same function. And so many of them now have these app rationalization processes where they're trying to dedup those things in enterprises. So we're helping them come up with not just a more cost-effective, but a much more kind of end user, better usable experience, I think, in terms of interfacing with all these SaaS tools to collaborate on digital work.

Interesting. So it's not -- like how long would it take to move like the Jira or the Confluence to your point? I'm just trying to think about if net new -- sometimes you talk about net new workloads, the net new seats kind of are joined on cloud and there's like integration with the data center solution? Or should we be thinking about it as really like are they lifting and shifting? Are they re-architecting? How are they -- how long does that take? And what is the involvement of that migration of this complex and integrated applications.

I think for the ones that are in that 100,000 seat or more, it's a 3-year process. So they take anywhere from 3 to 6 months to kind of inventory what they have and build a plan with us and a partner and then they're going to take another 3 to 6 months to migrate probably a good chunk of the initial seats, right? From there, they're often figuring out how to get marketplace apps and then plan the next set of use cases. So it ends up being kind of multiple phases over those 3 years. What I will share, though, is also for some of these customers that are in the more regulated industry, there's always going to be this remaining maybe 10%, 15% of seats that they want to keep on-premise. And so you may have heard we've started talking about our hybrid approach to this. And I think that's important because it is reflective of what you see in the market from other cloud vendors like Microsoft, where we end up having a private cloud option for that remaining 15%. So when I look at someone like a Bosch with 600,000 seats, at the end of 3 years, they probably won't have 600,000 seats in cloud. They'll probably have the majority and then they'll have some set that they may keep in private cloud for some of these higher security use cases. So biggest customers a couple of years to get across. Folks that are in that sub-10,000, they can usually do end-to-end that whole loop in about 6 months. Is that good context? It's hard because we've got 300,000 customers. So it's a pretty wide range of deployment models.

Yes. And if you have the mix, you love that. But if not, we can also just kind of move to talk about the pricing of cloud. I think that oftentimes you really hear about the cloud, enterprise and premium SKUs being relatively expensive even compared to the data center. How are you thinking about -- how are you tackling this first of all? And then second of all, also like elevating the value proposition that they can really get from the cloud solutions for data center?

Yes. It's definitely a big focus for us on the enterprise business. It really comes down to building that business case and educating customers on the fact that cloud is not the same products, right? Like we've invested for the last 5, 6 years on building a world-class cloud platform and so what that means is it's not just offsetting the work of managing the instance, which is the infrastructure layer, but we've invested and made some acquisitions on the data side to build a really robust intelligent data layer. That's where we've built things like the teamwork graph as well as out-of-the-box analytics, and so when we work with customers and we say, look, what tools are you using behind the firewall to track all of the work and do the reporting on the work. Oftentimes, they're paying for expensive Tableau instances, or if they have a developer productivity initiative, they might be buying third-party products like a Jellyfish, all of those things actually add up. And then when they realize that when they buy our premium and enterprise edition, that things like automation and data analytics, the teamwork graph and AI, as we talked about, are built into those additions that helps them make that business case to move to cloud.

Understood. Understood. And what else from an investment standpoint, do you think are still needed within Atlassian platform? And maybe how are you prioritizing them internally?

Yes. I'm biased because I had the platform team, so I'm always like, there are so many things we could add more of. I think we're just getting started on the data layer. Like I think the uptake on analytics has been immense for customers, and they're asking for more and more features there. The great thing is that we've built it in a very platformized way. And so at this point, we're enabling marketplace partners to build on top of that. But I think there's a lot of exciting things when it comes to -- when you think about AI in the platform and then orchestrating these workflows, I think build or buy, there's some new opportunities around low code that will be really interesting for us to add to the platform. Because if you think about it, it's like -- we've got a data layer that now gives you all the insights, right? We've got an AI layer that helps you create agents to take action. But then the final piece that is kind of missing there is, yes, the agent could write code, but really what you want are prebuilt pieces that you know work in your organization the way you want your app. And so I think having more of those available to teams outside the core development team means that with the addition of AI, you could have the entire company like driving that innovation loop, right? Because I think the traditional model is like let's get the insight quickly to the dev team so they can build the thing and get it out the door. And then we had this influx of, okay, there's low code things that non-devs could do, but there is quite a lot of limitation in terms of like how powerful those apps could be, but now married with AI, I feel like we could really like complete that loop and get much faster innovation for these companies. So that would be on my wish list.

Yes. No, that's -- I think that's one way -- like that is probably driving a vision that you and the broader Atlassian team have on your sort of like your strategic and competitive positioning, right? So when you think about your TAM and specifically across the 3 different markets that Atlassian plays in, how do you see that evolving? And maybe if any specific areas you see it particularly driving Atlassian in?

Yes. I think the stat that stuns me all the time is the fact that we have $14 billion in opportunity in our existing enterprise customers. And that's just with our current products, right? And so that's not even like as the portfolio grows. And that just shows how broadly we spread the seeds of Jira and Confluence and Jira Service Management. And so when I think about the potential, I see that in our core market of development, there's still a huge opportunity to grow, like we're a standard, but there's lots of teams that have other tools as we talked about, but I think consolidation will help with that growth. And then I think the other 2 really big opportunities are the ITSM space, right, whereas companies use Jira and then they ship a lot of these great products and services and then have to manage them, there is a very natural workflow that goes into the service management side. So JSM has been growing so rapidly for us, and I think that continues to be a huge growth vector. But the one that I'm the most excited about, I think, is the broader bringing the business teams and the technology teams together. So I think it's making people aware that Jira is not just for developers, but it's a place marketing can be in and HR can be in. I think as we ride this wave of consolidation and then architect this larger system of work for customers, that, that's a huge untapped opportunity for us.

And how does that affect the way you're thinking about go-to-market then, right? Atlassian has always been really oriented around individual products and elevating those specific capabilities. How do you expect that to evolve over time as you think more about the system of work that you've outlined with us today?

Yes. I think it means go-to-market has to evolve with us, right? It means that we're selling. I think you hear it in the narrative that I'm sharing. It's much more of a solution story, right? It's not a transactional like how many more seats of Jira do you want? It's the hey, what are the business problems that you need solved, which bottlenecks in your development process can we help you with? And so there's investment in bringing the right talent in to do that, which we've been doing the last few years. And we've been doing it thoughtfully. I think the biggest investment has to be the R&D side because I think being enterprise-ready takes these big resourcing and R&D commitment. And Atlassian has been in a good position to do that because we are -- we focus very much on R&D. So the last 5 years, it's getting the platform ready. And then now as we were saying, with the GenAI trigger, it actually becomes this massive accelerant. So now if we invest in those go-to-market motions that can really bring fruit fast, I think. So we are bringing in the right folks to do the solution sales model. And from my experience, that training can happen very, very quickly. And I think with GenAI is like the door opener, we're seeing a lot of great traction in the system of work conversations.

It's a great place to end it. And it's all the time we have today, but thank you so much for coming and joining us. Hopefully, we'll see you here again.

Yes, that was fun. Thank you, too. Thank you.