Atlassian Corporation (TEAM) Earnings Call Transcript & Summary

Welcome, everyone. I hope you're having a great afternoon. My name is Alex Kurtz, senior software analyst here at KeyBanc Capital Markets. And with us is the Atlassian CTO, Sri Viswanath. Sri, great to have you.

Excited to be here, Alex.

And we've got a lot to go through in about 40 minutes. As we've done throughout ETS 2021, really want to capture questions from the audience. So in the interface for the event, please start plugging questions away, and that are relevant to Sri and the cloud architecture that his team has built. And then we'll get to that as we kind of move through the platform through the program. And Sri, I just got approval that we can talk about Key's cloud use of the product. So if you want to reference us as we go through this, Chris said we could talk about 2/3 of our product were in the cloud. That's great timing.

So Sri, let's just start with the high level, the evolution of the cloud platform since launching road map a few years ago at AWS. What was the strategy? What was the intent?

Yes. Great question. I would say, in the last 5 years, we have transformed the company. It's less of an evolution, more of a transformation. And I'll go through different phases on how we got here and what is ahead of us. The first phase -- I joined Atlassian 5 years ago. As you all know, Atlassian's mission is to unleash potential of all teams. We do Jira, Confluence, Trello, Bitbucket and a number of other products. We are in the software and IT and workplace management for all markets. I joined 5 years ago. At that point, we had a cloud offering, but we were running our own data centers. It was quite obvious that we had to be on a public cloud, just so that we can focus on customer value and not just doing all the infrastructure work. So we did a massive project uplift to move our Jira and Confluence and our products onto public cloud, AWS, and made it multi-tenanted, made it much more robust and running on AWS. That project was so complex. Some engineers said, oh, my head is dizzy, and we called it Project Vertigo. And it took us about 2 years to finish. And the thing that Atlassian had was we were running the same code base for both on-prem and cloud at that point when I joined. We forked the code base, we forked the organization and we focused on cloud -- building a cloud-native platform. And that has paid extremely well for us. So what it has allowed us to do is to use native services on AWS and cloud. And over time, we have also created micro services, and there are a number of advantages of micro services for scaling. As you peel off micro services from an old monolithic code base, you can put the team anywhere in the world. And that has allowed us to basically create a number of micro services. We have over 1,500 running in production. And we have teams that manage that all across the globe. And we, in fact, recently went completely remote, which also helps us to be able to invest in architecture. And we created a platform layer so that we don't have to duplicate all the work in every single product, and we started peeling off things that are common across the board but also gives inherent value to our products that are core functionality to Atlassian. As you all know, Atlassian doesn't follow the traditional go-to-market model. It's a land-and-expand model. We try to put as many things into the product as possible in terms of go-to-market and try to expand in customers with landing teams, and we expand enabling teams and it expands in the company. They use more products and more teams use more products. They use Marketplace apps and more teams use more, so we expand in the company. And what we try to do is to step back and say, what kind of product do we need to build that is inherently capable of allowing this land-and-expand model. So the platform that we built essentially started ingraining those kinds of features. When you @mention in one product, it's exactly the same in the other product, and you can tie those 2 together, things like that. So we built those platform services. And over time, what we also did was started improving our scale and reliability because -- and security because that's extremely important for larger customers. And over time, what we have done is we have released a free version. We have a free, standard, premium and enterprise. And it's underpinned by the common platform underneath. And we serve the customers that are small, super small, like in the garage, to all the way to the largest of the large customers, including financial institutions. So that, I think -- the reason it works well is because we have built this robust platform underneath. And I'm happy to dive into each of those in terms of what we have done in security, reliability and how we have done -- in fact, 2 years ago or 3 years ago when we launched Vertigo -- finished Vertigo fully, we could only do 2,000 users on our platform per customer. And today, we have crossed over 20,000, and we keep growing, right? So that's been a trajectory that we keep going towards and continuing. If I look forward, we have done -- in the last couple of years, we have done a lot of enterprise features, which is not just for enterprise companies, but also for the larger base, and we continue to do that. And I'm happy to expand on all of this.

Sure. Yes. You gave us a lot to chew on there. So let's just start with -- and thanks for that question. Please start punching questions in, and we can start working through those. So I think one of the big questions is as Atlassian has moved into the cloud, one of the core values of that is net expansion of other products, right? So easier to add on in the cloud than on-prem. So can you just take us a little bit into more of the weeds about why being in the cloud, and how you design these products, they kind of like kind of co-sell together, and how they get integrated. And then it allows a customer to more rapidly adopt them.

Absolutely. Yes, that's a great question. So our goal is to make sure that going from one product to the other product is literally clicking a link and using the other product.

Right.

And the underpinning of this is we have built identity service, which uses the same authentication and authorization and permissions across different products, which is extreme -- that's step one, right? So you need to have that. Step two is the user experience for our customers. We want the products to look similar. So when you go from one product to the other, it doesn't feel like you came to a completely new world. So what we have done is we have created what we call a Teamwork Platform underneath, which has a common editor and capabilities like sharing capabilities and @mentions and decisions and tasks and to-dos. And there's a number of these elements that -- the collaborative elements that we have created that are common across board that helps us to go one from the other. I'll take a very specific example. Let's say you are in Jira. You're typing in a long-form text. It's actually done well in Confluence. So we can provide you a link right there that when you click on it, it moves that content to Confluence and start using Confluence. It absolutely helps out. The other thing that people don't realize in cloud that's like an inherent value is we are innovating rapidly and consistently on cloud, right? So on a day, we do like 1,200 deployments, right? So from the time we started until now in 10 minutes, we have probably done like 50 deployments into production. So there is always this constant improvement that happens that we see an issue, we fix it. We improve products. We add features. So it keeps improving. And as you know, customers don't have to manage this and maintain this. So the total cost of ownership is much lower. And the other thing that's inherent in this -- in cloud that's an advantage for Atlassian is the Marketplace and the extensibility that we have. The Marketplace, it's a thriving marketplace. We have lots of apps that people use. And it's a lot easier on cloud because, let's say, you're using some of the Marketplace apps but you also use Slack. You also use Google Docs. There are integrations that you can quickly one-click install and continue on your process. You don't have to download something and figure out where to install and involve lots of people. It makes it a lot easier in terms of how we bootstrap the overall setup.

So the other -- and we've gotten a couple of questions here, and we'll get to it in a second. But from an M&A perspective, I imagine a lot of the companies you're acquiring are all container-based in the cloud. So is it just kind of speeds up your M&A strategy a bit because you don't have to go kind of rewire these products to work on-prem?

Yes and no.

Okay.

It's not that clear cut. So let me break it down.

Yes, sure.

It does help us in terms of the platform that we have built is built on a clean architecture in terms of micro services.

Right.

So what it helps us to do is to pick a strategy that works really well for that case, right? So -- and as you know, M&As fail or pass or win because of the integration, like it's the most critical part of the M&A process. And the integration part, we do extremely well as a company, and I can talk about the culture that we have and the values that we have and the open company, no bulls***. And that all works really well. So we do well in the people integration part. The technical integration is not like you turn on a switch and it just works. That's why I said the no part. But the thing that it allows us is we have the services that you can pick and choose and strategically move the products over. It's not like a pause everything, wait for 2 years to move everything over and then we launch something. Like we don't have to do that. What we can do is we can pick a service and say, let's move identity over. And even inside identity, we have identified like 3 different layers of integration. We can do a simple integration where it works with the same authentication log-in, but there are some things that are different. Or you can go all the way to having exactly the same identity across both and it gives you a lot more features. So there is the gradation of integration that we can do, and we plan that specifically for each of those integrations, and that has worked well in our case.

Okay. A question here. Some users have noticed the cloud product is a little bit more -- a little slower than the on-premise product. That hasn't been our experience here at KeyBanc, but that is what this question is proposing here. Have you seen performance metrics come up from the cloud customers where you have a sense of how it's performing relative to on-premise?

Yes. Absolutely. This is one of our key metrics for our platform. And even in the last year, in my team, we have had significant investment in performance, and we have seen massive improvements in performance across the board. There may be some cases where specific areas -- when you are using lots of add-ons, a specific version, maybe it's lower. But the good part in cloud is we actually know that. We measure performance of every customer across the board and at all levels. So when somebody is -- and we use these Rails metrics, that's Google's data standards on how long does it take for you to render and to actually the product usable, not just some random metrics. And the usability metrics, we have a specific goal. And the way we do this is we've been improving and we'll keep improving. We have -- I have like a large portion of my team is working on the performance improvements. The thing that we do really well is we have what we call SLOs, service level objectives for each of the teams, and they can't breach that, right? So we maintain for every single place, and we do these reviews every quarter where we can dig in and say, wait, you had this customer below this bar for a while, and we need to go fix those. And in the last 2 years, I would say, 2 years ago, sh*t hit the fan, like it happens when you run a cloud service. And what we did was we stopped all development for a month or so and improved our reliability and performance. And over the last 2 years, we have made significant progress. So if there's a specific customer that's not -- well, I can take it off-line. But in general, I think we have made significant progress, and it's going really well.

Fair enough. I want to get to a lot of other stuff, but there is just some good questions, so I thought we'd just hit them right now. Just generally, we had GitLab, the CEO of GitLab on before you. And we were talking a little bit about Jira and how they leave that to you. That part of the market in the enterprise, you obviously have a huge franchise there. I guess, the essence of the question is Jira's been around for a while now. How do you keep innovating on Jira? Because you could argue that over time, that product could get stale and there could just be a new opportunity for an entrant to come in. So when you think about the core product, Jira and Jira Core, what are you doing to make sure that as a stand-alone product, not all these integrations with Confluence and other products you have, but that core Jira remains one of the best products in the market?

Yes. Yes. Number one is the investment and focus on Jira. It's extremely important. And we are also bolstering the features and capabilities that Jira has over time. In fact, last year, we launched the Jira automation, which allows you the automation on top of Jira. We are adding features, but not just adding features. I think that's -- many companies do that. What we also have done in the last few years is to step back and say, how do we make it modern? How do you make it -- both from the user experience perspective, and I can talk about the design improvements that we have done and progressive disclosure [ and on ]. Don't show everything all at once, but show the minimum thing that you start with. And then as you learn more, you can add more things and make it more complex. So we have done that really well in terms of user experience. In the back end, we've also done extremely -- it's not just like a incremental improvement, surgery in the back end to make sure that we get to a micro services architecture that makes it snappy, that makes it faster. And we can also have like additional functionality that we couldn't do before. For example, we can index and search much, much faster, even in a pretty large setup, and we have gone through much, much larger like millions of issues for a customer. And that works really smoothly and fast in terms of -- it's a -- if you look at it, we invest heavily in security, reliability, performance and also the functionality for Jira and Jira sister products.

Right. Okay. Another couple of good questions. We'll put those on hold for a second. So let's transition in machine learning and no-code, low-code and how you think about the future Atlassian through that prism, right? It seems like there's a lot of opportunity around machine learning. There's also a lot of opportunity around automation. A lot to chew on there, but it must be obviously on your road map. So what does that look like, the next couple of years from Atlassian?

Absolutely. Yes. Jira automation, we are investing in it. We have launched it. As you know, you've seen it.

Yes.

It's our Automation for Jira. Trello also has an automation in place. And that's now moved into a platform-level capability that's -- that may improve all the products in terms of automation. The way Atlassian -- we look at it as we are trying to solve a human interaction problem, like human collaboration problem. And the hardest part in automation is actually the interaction when the workload involves a human in between, right? And that's something that we are trying to solve as part of the automation problem, which is unique in terms of -- because that's the real problem in the world, right? So when you have some people undergoing an approval and they have to make workflow changes, like that's really hard. So what we are trying to do is to bring the technology of automation and the people practices into it so that we have both the practices and automation work really well in terms of it. And the other thing that we also do is we invest heavily in what we call smarts. Like we don't call it machine learning or AI and create a separate versions or editions for that. What we like to do is to embed that as part of our products, right? So if you, for example, in the cloud product that you mentioned, you go in and start typing somebody's name to @mention. It figures out who you work with and knows the network of people you work with and gives you the right relevant people. So the first few are likely the ones that you would add. So those kinds of things make the product inherently better and much more usable and are the delight features for our customers. And that's something that we keep adding and we'll keep adding in our road map.