Atlassian Corporation (TEAM) Earnings Call Transcript & Summary

All right. We can go ahead and get started since we're starting a little late. Most of you know me. My name is Jason Celino. I'm one of the software analysts here at KeyBanc. Great pleasure welcoming back Martin Lam, Head of IR for Atlassian. First, big congrats on the quarter. Excellent Q4 full year results. There seems to be some significant momentum in the business right now. It's great to see the numbers reflect that. So on the cloud side, specifically, you've seen 5 straight quarters of acceleration. You're consistently pointing to paid seat expansion and cross-sells, the primary drivers of the growth. Maybe can you just unpack those elements and explain, like, what's driving this consistency?

Yes. Thanks for having me, Jason. Yes, we're really pleased with the strong Q4 results to end our fiscal '26. I think what you're seeing is customers really value and understand the value of the overall Atlassian platform with the Teamwork Graph, which is a living contextual layer underlying the platform as well as our overall system of work, basically living system of record and system of work to actually drive those workflows for organizations. To Jason's point, we had a really strong quarter on the cloud side of things. And that outperformance was really driven by 2 things. It's actually consistent with what we saw in Q3. So it's good to see that consistency carry through. But it's driven by strong upgrades and cross-sell to our Teamwork Collection, which is basically customers being able to purchase the entire Atlassian platform and primarily for additional Rovo credits. We get 10x the amount of Rovo credits with the Teamwork Collection. So customers are upgrading for that additional AI capability as well as cross-sell motion into our Service Collection. So that was really great to see. All the while, we're starting to see -- continue to see strong seat expansion across our core products of Jira and Confluence, which I think highlights the importance of collaboration and teamwork and coordination in this AI era, right? Like, we've talked about for quite some time. In the AI era, the need to track, manage, plan all your work across the organization that doesn't change. And I think you're starting to see that continue to play out with strong cross-sell momentum, AI purchasing on Teamwork Collection with that strong seat expansion on our core products with Jira and Confluence. And I think it's notable that the strong seat expansion was across both software development and knowledge workers. So at our investor forum, we recently shared about 2/3 of users on Jira and over 2/3, more like 70% on Confluence are knowledge workers and non-software developers. That trend continued across this quarter where that net new seat expansion happened across both those sectors. So that continues to be really healthy across both segments.

Okay. Excellent. I do want to touch on that, but maybe just following up on some of the numbers stuff first. But margins for the coming year, you're modeling a little bit of contraction. There's some moving pieces here. Maybe can you just talk about some of the headwinds and some of the views on hiring?

Sure. Actually, I would point you to probably GAAP operating margins. I think that's where we're increasingly focused. We talked about one of our strategic priorities, along with enterprise AI and our system of work, is to drive durable profitable growth. And part of that is to accelerate our path to GAAP profitability and expand on GAAP operating margins over time. I think you actually saw that in this past quarter where we delivered GAAP profitability in Q4 and had strong GAAP operating margins. So for next fiscal year, we're guiding to 4.5% GAAP operating margins, which is an expansion relative to how we ended fiscal '26, which was basically flat or 0% operating margin. So that's great to see that progress and, kind of, reflective of the discipline that we're having on that side as we charge and accelerate that path towards GAAP profitability and margin expansion. I think you're quoting non-GAAP operating margins, and there's a couple of different dynamics for you to consider on the non-GAAP side of things. So earlier in fiscal '26, we announced the end of life of our data center product. With that comes pretty significant 606 changes where we're recognizing more upfront revenue on the sales of our data center subscriptions. And that drove approximately 4 points of margin benefit in fiscal '26. So we're recognizing significantly more upfront. That basically fell to the bottom line immediately, but just all timing of revenue recognition. And so that was a tailwind to fiscal '26 non-GAAP operating margin by about 4 points. So I think if you compare that to fiscal '27 and kind of normalize for those effects, plus I spoke earlier about trying to be more disciplined, not only from a headcount perspective, but also how we think and issue equity to our employees. And so we're changing the compensation mix between cash and equity for certain employees and certain roles. And that presents, like, a 3-point headwind on fiscal '27 non-GAAP operating margins. Again, it's just moving compensation mix between cash and equity. So normalizing for those 2 effects, you actually see non-GAAP operating margins increase. Again, it's a lot of moving pieces. So I actually would probably steer you more towards the GAAP operating margin expansion that I pointed to because that's simpler. It just helps you, kind of, cut through the noise. And GAAP is frankly where we're focused now as a company.

Yes. That's a good reminder. As, like, a software analyst, that's new to me. New concept. So maybe if we go back to kind of your explanation on the paid seat expansion. So one thing that investors have been focusing a lot with a lot of the AI worries has been like developer growth and developer headcount and knowledge worker growth and knowledge worker headcount. If we look at different data sources like Indeed job data, we do a CIO survey, and hiring intentions are up. There's a number of other data sources that are also pointing to near-term positive indicators for developers and knowledge workers. What do you think is really catalyzing this? Do you think AI is catalyzing near-term activity? And that this is maybe just a flash in the pan before we eventually see some contraction. We have some members of our KeyBanc IT organization here, and they talked about labor arbitrage with offshore and AI potentially. So help me understand maybe what you think around this.

Yes. I think it's important to note, as I mentioned earlier, that, that strong seat expansion or strong seat growth that we saw in Jira and Confluence is actually across both software development and importantly, non-software development or knowledge workers. So teams like HR, marketing, legal, finance. So really good traction there, and that continues to be an area of focus as we focus on penetrating more of the enterprise customers that we have and reaching and serving more of those users. But on the software development side, I think it's an indication of ability to create software now with AI is becoming greater than ever. And you're seeing that ability for companies to become software companies, whether you're a traditional nontech company, you now -- all of these companies have to drive more digital transformation, create more software, create more digital services as part of their overall strategy. And AI is lowering that cost, you're able to drive a lot more software development. Now the next part of that challenge is how do you actually make this -- all this increased software move in the direction for value for your customers or for the enterprise customers. And that requires a different level of coordination, right? Like a lot of the AI capabilities we've seen to date have been focused on personal productivity or individual productivity. Our ticker symbol is TEAM. We've always been focused on teamwork. How do you, again, I mentioned earlier, coordinate across your organization. How do you manage, track and plan work to make sure that we're all moving in unison towards the organization's strategic goals and delivering value to your customers because that's actually what ultimately matters as opposed to, again, all these disparate individual tracks happening. And I think that highlights what you're seeing in play.

Okay. Interesting. Yes, it's like the narrative changes every quarter. It's like every conversation I have, I feel good or feel bad. But at the end of the day, the numbers have been pretty good. So one thing that you introduced this year is subscription ARR to, kind of, hopefully smooth out how to view growth at the business given the model changes with the data center. So maybe just how often do you plan on providing the metric? Do you plan on guiding to it, just philosophy around some of the forward-leading indicators.

Yes. So we introduced subscription ARR back in May at our Investor Forum for the first time to help investors understand, kind of, the underlying health and strength of the business and our subscription base. We are, as I mentioned earlier, going through a cloud transition as we sunset our data center offering and migrate customers to the cloud in the coming years. And so to help normalize for some of the 606 noise that I mentioned earlier because we now have greater upfront term license revenue recognition on the sale of data center subscriptions and cut through all that accounting noise and timing noise, we've introduced subscription ARR, which, again, smooth things out and helps give you a better read on the underlying strength and momentum in the business. That continues to track incredibly well. We are guiding to that for the first time. We came off a quarter where we grew subscription ARR 23% year-over-year. And we're guiding initially for fiscal '27 to end fiscal '27 at 18% growth year-over-year. And so that's our initial guide. Obviously, we want to take a prudent approach with that guide. It is a new muscle for us. But again, I think it helps investors understand the entire subscription base, so both data center and cloud, cut through the ASC 606 noise and help you identify or I guess, cut through the noise of the migrations of people going from data center to cloud because I think historically, investors have had outsized focus on, say, how much of the cloud revenue growth is coming from migrations. Subscription ARR helps normalize for that because if someone moves from data center to cloud, you kind of just see that all within the overall ARR results.

Okay. I will take some questions at some point. I do want to keep this interactive. But I'm not that great of a mathematician, but some other people have noticed that your ARR growth, again, new motion, new muscle memory, but it decelerated a little bit versus the prior quarter, but then your cloud revenue accelerated. Anything specific on why that might have happened?

Yes. I would say 2 things. There is quarter-to-quarter variability in ARR. We are going through an enterprise go-to-market evolution as we continue to scale up our enterprise sales motion. Today, we only have about 400 quota-carrying enterprise sales reps, which is an incredibly low number, and I think it highlights the opportunity ahead of us as we continue to grow that team. Alongside, we have the cloud migration. So with these 2 dynamics, we're always intentionally or thoughtfully introducing changes into how do you align partners or how do you incentivize those salespeople. When you introduce sales compensation changes and/or motions associated with our cloud transition that can influence customer purchasing quarter-to-quarter. So there's quarter-to-quarter variability. So I encourage you to look at some of the full year dynamics. A perfect example of this was Q3. We spoke about this on our Q3 earnings call. We saw some pull-forward activity from customer purchasing behavior from Q4 into Q3 as a result of us changing pricing on data center, right? So we do things like data center pricing changes or stop selling data center subscriptions to new customers. All these things are in the vein of moving customers to the cloud and accelerating the cloud migration, but those motions also can shift customer purchasing behavior from quarter-to-quarter. So again, I think what you're actually pointing to is some of that pull-forward behavior that we pointed out on our Q3 call coming from Q4 into Q3 and again, creating some variability in the Q3 growth rate versus Q4.

Okay. Yes. No, that makes sense. Any questions from the audience? Yes. Sure.

[indiscernible].

Yes. Can you paraphrase the question?

Yes. So the question, I think, is around really the context and data that lives within our systems. I think earlier, I talked about a lot of the AI tools out there today are geared towards personal productivity. What our advantage, I think, is in the Teamwork Graph and the context that lives within all the different workflows and our system of work. So we have the data -- you're talking specifically about engineering. And in that, people document the requirements of what you're trying to build or what you're trying to accomplish. Did the team actually build and accomplish that? What is the customer feedback? How do you iterate on that development? All that context lives within our products like Jira, like Confluence. And I think that's an incredibly important asset for us, especially as an organization like KeyBanc tries to deliver for your customers or your employees internally. And so I think that's a very valuable asset. That's a very important value proposition of the overall Atlassian platform. So we talked more recently about the Teamwork Graph. It's a living contextual layer that understands relationships and what tools and what people are working on and again, what they're trying to accomplish as an organization. But more importantly, with that context, it delivers better results for people as they utilize AI tools. So we've talked about the Teamwork Graph delivers 48% more efficient token usage because then you're not having to have AI kind of blindly search across the organization. You have to understand the contextual relationships across KeyBanc. And then secondly, it delivers better results. So 44% better results because you have that understanding of what are your different teams working on, what is the context within Jira, context within Confluence, the decisions made that were previously made, what steps we're trying to drive as an organization. Understanding all that delivers the better, more efficient token results.