Atlassian Corporation (NASDAQ:TEAM) Q3 2023 Earnings Conference Call May 4, 2023 5:00 PM ET

Company Participants

Martin Lam - Head, Investor Relations

Scott Farquhar - Co-Founder and Co-CEO

Joe Binz - Chief Financial Officer

Cameron Deatsch - Chief Revenue Officer

Mike Cannon-Brookes - Co-Founder and Co-CEO

Conference Call Participants

Gregg Moskowitz - Mizuho Securities

Brent Thill - Jefferies

Keith Weiss - Morgan Stanley

Fatima Boolani - Citi

Michael Turrin - Wells Fargo

Kash Rangan - Goldman Sachs

Michael Turits - KeyBanc

Keith Bachman - BMO Capital Markets

Fred Havemeyer - Macquarie

Ryan MacWilliams - Barclays

Adam Tindle - Raymond James

Peter Weed - Bernstein

Ari Terjanian - Cleveland Research

Jake Roberge - William Blair

Operator

Good afternoon and thank you for joining Atlassian’s Earnings Conference Call for the Third Quarter of Fiscal Year 2023. As a reminder, this conference call is being recorded and will be available for replay from the Investor Relations section of Atlassian’s website following this call.

I will now hand the call over to Martin Lam, Atlassian’s Head of Investor Relations.

Martin Lam

Welcome to Atlassian’s third quarter of fiscal year 2023 earnings call. Thank you for joining us today.

Joining me on the call today, we have Atlassian’s Co-Founders and Co-CEOs, Scott Farquhar and Mike Cannon-Brookes; our Chief Revenue Officer, Cameron Deatsch; and Chief Financial Officer, Joe Binz.

Earlier today, we published a shareholder letter and press release with our financial results and commentary for our third quarter fiscal year 2023. The shareholder letter is available on Atlassian’s Work Life blog and the Investor Relations section of our website, where you will also find other earnings related materials, including the earnings press release and supplemental investor data sheet.

As always, our shareholder letter contains management’s insights and commentary for the quarter. So during the call today, we will have brief opening remarks and then focus our time on Q&A.

This call will include forward-looking statements. Forward-looking statements involve known and unknown risks, uncertainties and assumptions. If any such risks or uncertainties materialize or if any of the assumptions prove incorrect, our results could differ materially from the results expressed or implied by the forward-looking statements we make. You should not rely upon forward-looking statements as predictions of future events. Forward-looking statements represent our management’s belief, assumptions only as of the date such statements we are made and we undertake no obligation to update or revise such statements should they change or cease to be current.

Further information on these and other factors that could affect our financial results is included in filings we make with the Securities and Exchange Commission from time-to-time, including the section titled Risk Factors in our most recently filed annual and quarterly reports.

During today’s call, we will also discuss non-GAAP financial measures. These non-GAAP financial measures are in addition to and are not a substitute for or superior to measures of financial performance prepared in accordance with GAAP. A reconciliation between GAAP and non-GAAP financial measures is available in our shareholder letter, earnings release and investor data sheet on the IR website.

Please keep in mind that we would like to allow as many of you to participate in Q&A as possible. To facilitate that, we will take one question at a time. Please rejoin the queue if you have another question or a follow-up, and we will do our best to come back to you later in the session.

With that, I will turn the call over to Scott for opening remarks.

Scott Farquhar

Thank you for joining us today. As you've already read in our shareholder letter, we delivered another solid quarter of financial results and continue the steady drum beat of innovation to help our customers transform the way they work. We've just wrapped Team '23, our annual Atlassian conference where we gathered with thousands of customers and partners.

The Atlassian community has never been stronger, and the enthusiasm for our product announcements was unmatched. We introduced Atlassian Intelligence, a virtual teammate, which uses generative AI technology and leverages the Atlassian platform to accelerate customers' work in a way that is tailored to them. With 20 years of knowledge reflecting how hundreds of thousands of software I-team, IP and business teams plan, track, and deliver work, Atlassian Intelligence is like a floodlight lighting up new customer insights. We're excited to see how AI will unleash our customer's potential and strengthen our competitive advantage.

In addition to Atlassian Intelligence, we also announced exciting new features and platform enhancements, including Confluence, whiteboards and databases support for up to 50,000 users now Cloud and Beacon, our new threat detection and mitigation product, just to name a few. It's incredibly exciting to think about the opportunities we have in front of us with unique features and capabilities for Atlassian Intelligence and the platform innovations we're delivering. Customers are more enthusiastic than ever to migrate to the cloud and move forward to helping them on that journey.

With that, I'll pass the call to the operator for Q&A.

Question-and-Answer Session

Operator

We will now begin the question-and-answer session. [Operator Instructions]

Your first question comes from Gregg Moskowitz from Mizuho Securities. Gregg, your question please.

Gregg Moskowitz

Okay. Thank you very much for taking the question. Good afternoon, everyone. You mentioned in the shareholder letter that the Q4 guide includes an assumption of increasing headwinds for churn and for premium addition upsells, even though you've yet to see a significant impact in those areas. And I think that's prudent. But wondering if you could contextualize this a bit more for us. For instance, are you basing in a modest incremental headwind or would you say that it's more substantial than that?

Joe Binz

Yeah. Thanks Gregg. Really no change to our guidance approach that we took back in February. When it comes to the cloud guidance, specifically that cloud guidance range assumes the macroeconomic environment gets worse in Q4 and year-to-date trend lines continue into Q4. You’re right to call out. The fact that the low end of that range not only assumes continued weakness in the two drivers that we've seen to date around paid seat expansion at existing customers and free to paid conversions, but it also does include some macro impact to areas that have held up well year-to-date, like churn upsell and migrations. So we do expect and have built in pretty substantial macroeconomic headwinds at the low end of that range.

The other thing I'd call out really quickly is if you take that Q4 guidance, it puts our full year FY23 cloud guidance at 37% year-over-year. That's squarely within the range of the 35% to 40% we provided in February. So really no fundamental change overall in our cloud outlook or our approach to guidance.

Gregg Moskowitz

Thank you, Joe. It's helpful.

Operator

The next question comes from Brent Thill from Jefferies. Brent, your question please.

Brent Thill

Joe, EMEA grew 22% this quarter. I think last quarter was 30%. It seems like you had some weakness in the region. Can you just give us a sense what happened there? And maybe I'm just reading too much into the numbers, but anything to call out there.

Joe Binz

Yeah. Awesome. Thanks Brent and great to hear from you. I'll start and then I'll let Cameron chime in. In EMEA, revenue trends tend to be quite volatile quarter-to-quarter than the U.S. And the reason for that is pretty simple. EMEA has been slower to adopt the cloud, so there is more data center revenue and the EMEA sales mix. And DC, as you know, is inherently more volatile for two reasons. One is revenue recognition. 20% of DC license value is recognized upfront. And then secondarily, customer purchasing behavior around data center pricing changes also creates some volatility. So I wouldn't read too much into that. It's typically just a function of quarter-to-quarter volatility, but I'll pass it to Cameron for more of the customer insight.

Cameron Deatsch

Yeah. Repeating what Joe said there. But I would like to speak just to the cloud adoption in EMEA. Obviously, we do see cloud adoption lagging behind U.S. and the Americas, but in general, over the last year as we've done things like data residency in Germany as well as higher requirements for ban fee for financial services organizations, we've been able to see more and more of our larger customers in Europe choosing to go to cloud. We actually worked with a very large European bank over the last few months getting them signed up for cloud, and we're starting to migrate them today. So I will see that -- although they're lagging, we are seeing that adoption continuing to increase more in line with the U.S. numbers.

Brent Thill

Thank you.

Operator

The next question comes from Keith Weiss from Morgan Stanley. Please go ahead.

Keith Weiss

Excellent. I'm going to sneak in two questions. One, kind of more tactical and short term in nature and one longer term. On the short term, and this is going off of Gregg's question, I think what a lot of people are trying to understand is what the shape of the quarter was and how demand held up through the quarter? And whether that influenced sort of the more conservative Q4 guide. So, maybe if Cameron gives us a view on kind of like how the quarter progressed.

The longer term question comes from sort of, I think the debate that all of us are having around like large language models and generative AI on whether they're going to be disruptive, or complimentary to a lot of existing businesses. You guys have already rolled out functionality when it comes to your service -- the ITSM side of the equation. So how do you guys think about that question of the -- by sort of the risk of being displaced by people utilizing these new technologies versus your ability and what's going to make you sticky and being able to benefit from them.

And does that necessitate at all kind of changing the pricing model, moving away from the pricing model, maybe to better sort of account for the value that you're adding that doesn't take place with the agent itself, but takes place on the platform actually doing a lot more of the work.

Joe Binz

Thanks for the question, Keith. This is Joe. I'll take the first part with Cameron and then Mike will chime in on the second. In terms of the linear nature of the quarter, we did not see anything unusual about the linearity of our billings in the quarter, nor did we see anything unusual in the month of April where trends were consistent with Q3. Cam?

Keith Weiss

Perfect.

Cameron Deatsch

Yeah. I'd say the only thing is, as all of you know, we did a price change on our server and data center licensing in mid-February. That, of course, causes some behavior throughout the quarter, but no -- nothing significant throughout the quarter beyond the -- reaction to the price changes. Mike?

Mike Cannon-Brookes

Yeah. Thanks Keith. Great question. Look, quite clearly what I can say from Atlassian's point of view, we think that AI large language models, as you mentioned, is a huge opportunity for us as a business. If you look at our fundamentals, we've always tried to solve human problems, not technology problems, right? Which means we have a lot of customer knowledge and data and trusted to us over the last two decades. We have a fantastic platform to build on top of that connects that, that knowledge together. And we have a guarantee team to build features on top of these things. So AI allows our customers to be significantly more efficient to get to our mission of unleashing the potential of every team. That's what we believe it will do.

So from our point of view, we think our best opportunity to grow our business through the Atlassian Intelligence feature set is to drive customer growth, right? We have a goal of -- in the long-term winning the Fortune 500,000, not the Fortune 500. And we think that these technologies will help us get there. That's why we're including the features in all additions rather than as a separate skew. It's not a good customer experience and we think we'll slow adoption to put it that way around.

We also, obviously, have a major trend as you roll -- well, we're all in -- the migrations to the cloud in our business. Our largest customers in data center have the most amount of knowledge and data built up over time, which hence they can get the most amount of efficiency. And with this feature set being only available in the cloud, it's a further reason for them to migrate or bring forward their migrations. And as we know that the migrated customers have really great economics for Atlassian, faster expansion rates, et cetera. So, that's the reason we are looking at it that way. Fundamentally, yes, a huge advantage for our customers, which is why we're in this business in the first place. And secondly, that's how we're looking at growing Atlassian's business through the Atlassian Intelligence feature set.

Keith Weiss

Awesome. Appreciate the color.

Operator

Your next question comes from Fatima Boolani from Citi. Please go ahead.

Fatima Boolani

Good afternoon. Thank you for taking my questions. Mike, maybe for you, or Scott, on that the Jira Service Management product pillar, a lot of activity and action in really boosting the use cases that JSM can be deployed in specifically around external customer support use cases. And we're familiar with certain vendors in that market. So I'm curious what else needs to happen to really drive more monetization scale in JSM, and particularly in the context of some of your very large customers who are working through a headcount retrenchment process where JSM can be a very labor inflationary investment. And then a quick follow up for Joe, if I may.

Mike Cannon-Brookes

Sure. I can certainly take that for you, Fatima. Look, we remain incredibly bullish on our position in the ITSM and ASM spaces. And obviously Jira Service Management is a huge part of that. I think as customers naturally are looking to be more efficient with their spend in a more turbulent and economic time, that would lead to advantages in Jira Service Management and continue to push those. We've invested heavily with Atlassian Intelligence feature sets and virtual agents and all sorts of other technologies that make customers more efficient, make agents more efficient, fundamentally allow them to process more requests per day, per hour, with hopefully greater custom fidelity as well. So, that's one of the reasons we're incredibly bullish at the moment.

Second is Jira Service Management, just to remind people, it lands predominantly in IT teams, but it does have a great expansion story to other parts of the organization in terms of like HR, finance, marketing, legal areas which are fundamentally service-driven parts of an organization, that have their customers being other parts of the organization. That has long been our thesis. Again, our fundamental philosophy is that there are far more agents we believe inside a company than other companies, think other vendors think, because we're not just targeting the IT team. That said, obviously we have a wonderful landing spot through the development and IT organization and expansion, and we continue to see that throughout the quarter as customers increasingly use Jira Service Management outside of IT teams. And that's a long term growth factor for us.

Fatima Boolani

I appreciate that detail. Joe, the question for you is just on the cloud trajectory. I know a lot of us have been grappling with how that segment is digesting and absorbing the pretty meaningful moderation you've seen and see expansion. And I know you alluded to in your shareholder letter that churn levels necessarily have not increased. But can you help us sort of understand, or at least, maybe put a -- some sort of a downside cap on how you're thinking about that dynamic where you're really actively trying to manage gross seat churn in order to support growth in that segment?

Joe Binz

Yeah. Thanks for the question, Fatima. When we look at the churn rates, especially customer churn, it's been very consistent and stable throughout the year. That suggests there's nothing that's changed from our structural competitive position. Customers continue to come to our site, try our products, our monthly active users, our free instances. All of that seems very healthy and strong. So churn has not been a big issue for us.

The primary driver of the pressure that we're seeing in that cloud segment is really around paid seat expansion at existing customers and the free-to-paid conversion rates that we've talked about. The remainder of the business, whether it's migrations or upsell or cross-sell, all those drivers pricing -- all those drivers continue to be very healthy from our perspective. So, that's really been the core issue that we've been working through.

Fatima Boolani

Thank you.

Operator

The next question comes from Michael Turrin from Wells Fargo. Please go ahead.

Michael Turrin

Hey, great. Thanks. Good afternoon. Appreciate you taking the question. Maybe one from me on margin. We entered the year expecting mid-teens operating margin. You just delivered a second consecutive quarter of 20%-plus. Maybe Joe, if you can speak to what's driving that uplift.

And then as a second part, at the investor session, there was a comment around fiscal 2024 as a year of continued investment. So any color in helping us square that with just anything else you're doing to help optimize on the cost side as we're thinking through the puts and takes of margin trajectory and some of the offsets you have within your control. Thank you.

Joe Binz

Yeah. Great question. Thanks for asking. If you look at our operating margin performance in the quarter, we did outperform our expectations by six to seven percentage points. Five points of that was driven by lower operating expenses and better operating leverage. Those operating expenses were -- favorability was driven in two primary areas.

First, we had less than expected payroll related costs from lower bonus expenses and headcount, including about $10 million in savings related to our restructuring. And then secondarily, we had less than expected discretionary costs in areas like professional services and T&E as part of our plan that we discussed earlier in the year to be responsive to the macro impact on our business and align our cost structure with revenue growth.

I'll let Scott talk to the comments at Team '23. The last point I would make just is the focus for the team has been around maximizing the return on every dollar spent, making discipline prioritization and resource allocation calls, and driving operational efficiencies. And I really believe the team has done a really good job this year of executing across those three things, and that's what you see showing up in that operating expense favorability. Scott?

Scott Farquhar

Yeah. Just to reiterate what you'd mentioned, Michael, for those that weren't at our investor call, our team conference, but yeah, we did say that. We see FY24 being a continued investment year. We still see huge opportunities out there in those areas. We talked about migrations, enterprise ITSM, that we're being aggressive in investing behind. And we expect those investments to continue through FY24. Of course, they're taking into account everything posted.

Michael Turrin

Thank you.

Operator

Your next question comes from Kash Rangan from Goldman Sachs. Please go ahead.

Kash Rangan

Hi. Thank you very much. Just a quick couple of introspective questions here. The cloud growth trade is certainly looking to be set for deceleration in Q4. As you look at next year, is there any reason to believe that cloud growth rate could reaccelerate at some point in fiscal 2024 based on your assessment of timeline of migrations and the product readiness? And also as it relates to generative AI, I'm wondering if you can -- and it's been hard to get clarity out of company management because it's such a nebulous topic. To the best of your abilities, can you prognosticate if this will mean that you will need fewer developers and fewer service people? Therefore, there is the idea of a shrinking TAM, but then you gain share relative to the TAM because your technology will allow to displace workers that do not have the generative AI capability. Or is it that -- it actually grows the pie because the number of developers will actually grow because the barrier to software development comes down. So people like me can start develop, hopefully that should not be the case. But how should we think about the TAM implication from a headcount perspective as a result of generative AI and what it does for developers and service people? Thank you so much.

Joe Binz

Hey, Kash. This is Joe. Thanks for the question. Obviously, we aren't providing specific quantitative guidance on FY24, but I would offer up the following how to think about it directionally, specifically as it relates to the cloud space. The big caveat I always offer here is on revenue. There's lots of moving parts, but the biggest being the macroeconomic outlook, which is highly uncertain. And as you've seen in FY23, that can have an impact on our business. We're not immune to those factors.

Beyond macro, I'd have you think about the following. Mathematically, we will have easier comps as we move through the next year. We will have a significant event in the second half of the year with the server end of life. And our focus there is really on migrating those customers to either our cloud preferably or data center offerings. We'll also have some benefit from pricing as prior price increases and less loyalty discounts layer into the model. And then lastly, the decisions that we've made this year around rebalancing and reprioritization to reinvigorate revenue growth will begin to land and have positive impact as we move through FY24.

So in the end, macro's going to play a big role and we'll see what that ultimately brings. But hopefully that helps you think through some of the other dynamics that will relate to cloud revenue and FY24.

Mike?

Mike Cannon-Brookes

Yeah. Kash, I can chip in if you're looking for prognostication, that's probably my column. I would answer two ways personally. Firstly, the software world, if you want to think about it, you ask if there's going to be less developers, et cetera. My view and our house view, I suppose is no, right? The simple answer to that is software is not demand constrained. It is supply constrained. We are constrained by the supply of developers in the world, not the demand, the number of ideas for pieces of software we could create. So, when you create tooling that makes it possible for efficiency of any form, you will soak up more of that demand with the existing supply. The supply won't go down. The number of developers, I don't think it goes down that's not the way, sort of human creativity, which is ultimately to software works.

Secondly, when you make software developers more efficient, if you're talking about artificial intelligence or large language models with code assistance and writing code, et cetera, or creating parts of software that makes far more software. The more software there is in the world, that's good for Atlassian in a generalized sense. We are not necessarily working on individual developers. Again, in Jira software, they're about a quarter of the audience are software developers. The more software you have -- the more need you have for still designers and program managers and making sure that we have the right software that we're building, you now have a lot of software to operate and run and manage. So the more software there is in the world, that's generally I think a good thing for Atlassian. There will be puts and takes across this world as it flows through. But those have been my two fundamental points.

We maintain incredible optimism for artificial intelligence and what it can do for us as a business to help unleash a potential of our customers, to help them create and manage more software in that market of our three markets.

Kash Rangan

Thanks Mike. I will learn the code [ph]. Thank you.

Operator

Your next question comes from Michael Turits from KeyBanc. Please go ahead.

Michael Turits

Hey, guys. I wanted to just talk about data center versus cloud. You're seeing that some of those impacts around expansion free-to-paid impacting cloud, data center beat in the quarter, and it sounds like you're looking for only a minor decel in data center going into the next quarter. So could you tell us what's supporting that growth on a relative basis versus cloud?

Joe Binz

Yeah. Thanks for the question, Michael. This is Joe. You're right. Data center had another strong quarter of revenue growth that accelerated to 47% and was better than we expected. And that was primarily driven by two things, stronger than expected renewals and better than expected seat expansions at existing customers. We were helped a little by deal pull forward resulting from the price change in February, but it was largely consistent with prior year and only slightly higher than our expectations and certainly not nearly as significant or pronounced as it was two years ago.

Having said that, migrations from data center to cloud remain very healthy. Year-to-date, over 50% of cloud migrations are coming from data center and that's up from about a third a year ago. So we've got good progress there as well.

And then in terms of the guidance, we do expect growth rates there to moderate a little bit in Q4 and beyond, and that's really driven by increasing migrations from data center and cloud and server to cloud, as we remove migration blockers and add compelling features and value to our cloud offering. And then secondly, in the data center, we are starting to lap strong prior year comparables in Q4. And so that's another factor in the growth guidance that we've given.

Michael Turits

Okay.

Cameron Deatsch

Yeah. This is Cameron. I'll -- let's add a little bit on the data center side. Just largely basically talking to customers, we were at Team '23, roughly a few weeks ago, our many of our largest customers are there and many of those customers are still on data center, have been on data center for many years. Now just say just the tone with those customers over the last few years. I'd say last team, or a few years ago, three years ago, it was always about like, why should I move to cloud? What are the capabilities? Are you going to get to my data privacy requirements? You name it. I'd say that tone this year, just talking with all those customers is, okay, how do I get there? Right? You delivered the new AI, we have these new functionality. We're bought in now help us, guide us through this plan. And I just think that overall tonal shift has been super positive and helping get more data center customers moving to cloud going forward.

Operator

Your next question comes from Keith Bachman from BMO Capital Markets. Please go ahead.

Keith Bachman

Hi. Thank you. My question relates to what Cameron was just talking about. As you think about just philosophically in FY24, would you expect data center to continue to outgrow cloud or, do the compares and some of the other feature set things you were just talking about?

And then underneath that, you've talked pretty consistently about cloud is getting the benefit of 10 points of growth associated with conversions. As you think about what happens to that 10 points as you anniversary or get beyond the server date, which is mid-Feb a year away from now, does that cloud conversion go to zero? Or is there continued benefit as you migrate longer term from data center to cloud when you in fact perhaps move to enterprise version? So in other words, is there still a longer term benefit of cloud conversion or does that 10 points go to zero? Thank you.

Cameron Deatsch

Thank you. So, this is Cameron. I'll talk to just largely the customer choice ahead of us. You're absolutely right. So server end of life goes into effect February of next year, and we still have a variety of server customers still sitting out there on a variety of sizes that are effectively need to make that choice between now and February. And many are doing it month by month. Obviously, many are going to wait till the last minute to make that choice. And they do have two choices.

Now, obviously we're going to lead and try and get as many of those customers to choose cloud, but obviously we have data center as a very strong capable version for them going forward. So, I see that continued option out. Post 2024, we'll still have a sizable data center customer base on which to continue migrate customers to the cloud. So, this migration journey that we are on does not stop February of next year. It is a multi-year journey as we continue to get all of our on-premises customers eventually to the cloud.

Joe Binz

Yeah. Thanks Cam. And I would say, nothing more to add to that. Beyond we continue to invest to grow and accelerate that migration to the cloud. We're removing blockers. We're making migration easier through tooling and support investments. We're also closing the gap on pricing between advantage pricing on DC and cloud with our overall pricing strategy. And just overall the cloud platform provides the best experience, whether it's analytics or automation or AI or collaboration and better TCO, those factors will only increase over time. And then, just to reiterate what Cameron said, there's a lot of runway left on that cloud migration story. And we expect to continue to see that even after the server end of life.

Keith Bachman

Okay. Perfect. Many thanks.

Operator

Your next question comes from Fred Havemeyer from Macquarie. Fred, please go ahead.

Fred Havemeyer

Hey, thank you very much. I wanted to ask about Beacon. Actually, it caught my attention in your shareholder letter. Offering -- it's an -- of course an early access product, but shifting and focusing a bit on cybersecurity as it relates to just Atlassian's cloud ecosystem seems an interesting adjacency here. And I'm curious, with this early access program, are you seeing signs that there's demand across your customer base for just a broader set of cybersecurity solutions? Is this something that you think is specifically monetizable? And is there any sort of scale of customer that is particularly applicable to?

Scott Farquhar

Thanks for the question. Scott, here. Just a reminder, so firstly, there's a big demand for this sort of product from Atlassian from our customers, particularly at our use of conference. And just customers are excited by this. And if you think about moving to cloud, customers want to make sure that their data is secured and there's things that we can do, and all that so we can do, and we can see the shape of usage of our products that allow us to work with our customers on flagging things that might be of interest to them.

As a reminder that this is not a generic security product. This is across our data and our cloud. Now we have advantages because we have one platform in the cloud and we can offer Beacon across most of our products. So, there's an advantage there from a sort of consolidation perspective and from an IT admin play in a similar way you've seen us to access on the user authentication side. But this is really a product to help customers be very comfortable and with how their data gets used in the cloud. And in most cases, what we can offer here is, again, ahead of what the customers could do themselves in a behind the firewall setting. And so this is just yet another example of how the cloud is overall better for our customers than what they can do themselves and another reason for them to migrate.

Operator

The next question comes from Alex Zukin from Wolfe Research. Alex, please go ahead.

Unidentified Analyst

Hey, guys. This is Alan on for Alex Zukin. I just wanted to ask a financial question. If I think about the shape of NRR through the quarter, I know this isn't a metric you guys report to. But just to better understand the growth that you're seeing in the business, can you just at a high level talk about what that shape look like through the quarter and through April?

Joe Binz

Yeah. Thanks for the question. Unfortunately, no specifics to share with you today on the NRR. Given the macro pressure and headwinds, we do see on paid seat expansion, it is trending lower. There was nothing unusual about that trend in Q3 relative to Q1 and Q2. Those trends just continued into Q3. And I'd say lastly, the underlying fundamentals in our business remain very strong. We see no change to our structural competitive position, so we do expect those retention rates to recover once the macro picture stabilizes and improves.

Unidentified Analyst

Got it. Okay. And just as a quick follow up. I appreciate the color on the guide for Q4 of cloud growth benefiting 10 points from migrations. Just so we kind of have the numbers correctly here, do you mind just kind of telling us like what that exact benefit was in this quarter and the last two quarters or so? Thank you.

Joe Binz

Yeah. We've consistently said we're getting -- overall approximately 10 points of revenue growth in the cloud business from migrations that hasn't changed dramatically throughout the year.

Operator

Your next question comes from Ryan MacWilliams at Barclays. Ryan, please go ahead.

Ryan MacWilliams

Thanks for taking the questions. Just on a vertical standpoint, any verticals worth calling out that may have impacted the quarter or the guide? And at this point, do you have a sense maybe what percentage of server customers might choose to remain on server even past the end of life for a couple quarters after? Thanks.

Cameron Deatsch

Yeah. This is Cameron. I'll take that. Yeah. Just from a customer perspective, as we mentioned, with over 250,000 customers and plus all industries geographies, the size of all type, we see that as kind of a massive competitive differentiation for Atlassian, giving us plenty of growth opportunities across the customer base for quite some time. The -- of course, the trends we've seen in seat expansion slowing down is broad based across the entire customer base that we see today. One advantage we have is that as every company starts bringing in more and more technology to deliver value to their customers, we obviously get to take advantage of that as we are helping companies with technology teams and business teams work better together going forward.

And then the second part on the server end of life thing, largely most of those customers we see -- there will be some that is a perpetual license. There will be some customers, we believe that most will choose cloud, some will go to data center and obviously some will continue to use the server licenses unsupported. Just about every customer size that I speak to largely is under a compliance requirement or just general it guidance is internally that they do not run unsupported software and that we'll be choosing cloud or data center post the end of life. So I really see that as a very small portion of the customer base.

Operator

Your next question comes from Adam Tindle from Raymond James. Adam, please go ahead.

Adam Tindle

Okay. Thank you. Maybe one for Mike or Scott. I was going back to my notes from this time last year where you announced free additions of cloud products. And you talked about it echoing an approach from 2008, 2009 where you brought in your customer base by offering a $10 starter license. So if we fast forward one year later, your customer account is still growing, so you're certainly broadening the base like you did in 2008, 2009. The question would be maybe take us back to the upsell motion from the starter license, and compare and contrast that to upselling from the free cloud, what you learned then and what you can apply now to improve cloud upsell. Thanks.

Scott Farquhar

Hey, Scott here, Adam. Great question. And so just reminder for those who haven't followed you asking stories for such a long time that back in the 2008, 2009 financial downturn, we introduced three versions of our behind the file products back then. And $10 starters, I was really -- what we introduced back then where you had to pay $10 and you got 10 uses of our products down from what was the couple of thousand dollars with our lowest price at the time. So very disruptive pricing changes. And that was a long-term play for us because we worked a lot of money on the table. People were previously paying us thousands of dollars, were now paying us $10 to get a version of our software.

And the good thing with that is it opens the funnel up to a whole bunch of companies that otherwise would not have considered our products and over the long term they go from 10 to 12 users and then start paying us on a different pricing curve.

So, echoing that, in the COVID time period a couple years ago, we introduced basically free versions of our products down from $10 a month in the cloud down to free. And soon we saw that open the funnel at the top of our customer acquisition side of things. And so natural questions asked, well, then how do you convert more of those free customers to paid? And how do you change that cycle? That is something we do look at. But honestly, most of our free-to-paid is based on usage and based on customer value and changing from 10 to 11 users. And one of the advantages of the Atlassian business model is that customers pay us as they start using more of our features, and we don't expand a huge amount of energy trying to call them, trying to encourage them. You could waste a lot of money in a sales motion doing that. And so what we do is spend a lot of time looking at the levels at which we charge for things and make sure that is the appropriate gates, and make sure we provide value to our customers. And so as they use our products more and more. We get the value as they increase their usage.

Now, not only have we done free -- at the start of COVID, when we saw a couple of months ago more interesting macroeconomic times, we decided to go harder on ITSM, on our JSM offering where we made it cheaper at the low end. And for competitive migrations for people that were switching out from other higher priced offerings. And so, at Atlassian, we talk sometimes about the price rises that come through at various parts of our cycle, but we also spend a lot of time working on how do we make it cheaper in other areas like in order to more aggressively get market share. And so for us it's a long-term play about getting market share and not really a short term sugar head [ph] of trying to convert those free customers to paid in any particular given time period.

Adam Tindle

Makes sense. Thank you so much.

Operator

The next question comes from Peter Weed from Bernstein. Peter, please go ahead.

Peter Weed

Thank you. I guess, there's two parts to this kind of both focused, maybe on a fall on even with the new customers. You certainly had a nice uptick this last quarter relative at the prior one on net new customers. And I guess there's two parts of it. One is we were backing out in our model, what portion of the -- kind of new revenue growth was coming from the two? And it looked like you were probably seeing kind of flat quarter-over-quarter revenue contribution from kind of new customers over the trailing 12 months. And most of the headwinds are really coming from existing customers. Does that seem about right? That the new customers were probably kind of like flattening and it was mostly the existing customers expansion that that had been the issue.

Cameron Deatsch

Yeah. This is Cameron. I'll speak to the net new customer number. So, as we said before, the new customer number does jump around from quarter to quarter for a variety of reasons. That said, very glad to see the increase from Q2 to Q3 with over 6,500 net new customers. It's showing that there's continued demand for what we have, that we still can convert those free customers to paid customers. And it's nice to see the quarter on quarter increase. However, I do want to call out that challenge that we've been mentioning for the last few quarters of our conversion rate from free plans to paid plans is still lower than it was historically before we saw these macroeconomic headwinds.

As far as net new customer's impact into our short term revenue, it is minimal. And today where we always focus on the net new customer, overall number is really a better guide for our long-term portion of our business.

Operator

Your next question from comes from Ari Terjanian from Cleveland Research. Your question please.

Ari Terjanian

Hi, all. Thanks for taking the question. Just a question on the cloud growth and expectations here. One, given the strength in data center, you called that out as better than expected in the quarter. Does that suggest potentially that future migrations to cloud may not be quite as pronounced, given customers are renewing on data center more than you expected?

And second, I believe in the shareholder letter it was called out for the first time there was an impact from seat count reductions. Do you believe that's more reflective of layoffs that are being -- that are occurring today or more reflective of the layoffs that we saw last year? Thank you.

Joe Binz

Yeah. Multi-part question here. This is Joe. I'll take part of it and then Cam will chime in. I'd say in terms of the data center to cloud migrations, not at all. We expect those to continue to be strong for the foreseeable future. Cam spoke earlier about the multi-year journey we're on. We continue to add a ton of value in the cloud. We continue to invest in migration, tooling and customer support. We continue to invest in data residency and scalability and certifications and extensibility and all the things, that are going to enable more customers in that data center category to move to the cloud. So, we remain very bullish on that opportunity.

Cam, you want to take the next part?

Cameron Deatsch

Yeah. Yeah. I'll -- just the -- we see customers choosing data center as further investment in Atlassian. And our multi-year journey of getting customers to the cloud is very much on track with our expectations, and we increasingly are improve our ability to migrate customers every single day.

As far as you mentioned on the seat count reduction, in general I say is the biggest issue that we focus on is the paid seat expansion and customers are still expanding their seats. It's just not at the same rate that we saw before the macroeconomic headwinds. Of course, there are a subset of our customers, a very small subset, that have reduced their employee count over the last few months. And when their renewals come up, they're renewing at a lower tier than what they previously renewed, because they have less the users in their business. But overall, that is a small overall percentage of our customer base. The bulk of where we focus is really the paid seat expansion.

Operator

Your next question comes from Jake Roberge from William Blair. Jake, please go ahead.

Jake Roberge

Yeah. Thanks for taking my questions. Just wanted to double click on that the data center strength you saw in the quarter. Is that more a result of new customers starting to land there or upsells within existing customers? Or is that really just the server migrations going more towards DC than you expected to cloud?

And then more of a high level one. But what's driving the seat expansion for DC versus the headwinds for cloud? Is that primarily a result of the enterprise focus in DC?

Joe Binz

Yeah. Thanks for the question. This is Joe. As we discussed earlier, when you look at that DC strength and resilience, it's really driven by two things better than expected renewals and then paid seat expansions at existing customers. Cam?

Cameron Deatsch

Yeah. I have to say is this -- in the quarter, we have to remember we did do a price change on data center and that causes customers to make a choice whether they -- if they're thinking about adding users in the future there, it's a good compelling event that for them to choose data center. But once again, we see that path to data center. We prove it again and again. We have the ability to move data center customers to the cloud. Over the last year, half of the seats we've moved to the cloud are from data center customers. So, we see this as within the quarter. It's fantastic. People choose data center, but that's great. We can eventually get them to cloud and we're proving that every day.

Operator

Thank you. And that concludes our question-and-answer session. I will now turn the call over to Mike for closing remarks.

End of Q&A

Mike Cannon-Brookes

I just wanted to say thank you everyone for your questions. Thank you to those who came to our analyst function at Team '23. And I hope you all have a fantastic rest of your day.