Atlassian Corporation (TEAM) Earnings Call Transcript & Summary

Welcome, everyone. Thanks for joining us for this investor forum here at Team 23. It's really good to see all of you who have joined us here in person in Las Vegas. And for those of you joining virtually, thanks for tuning in. So for agenda today. First, you'll hear from customers at United Airlines and Reddit about how Atlassian products drive cross-functional productivity across the organizations. And then co-founders and co-CEOs, Mike Cannon-Brookes and Scott Farquhar; President, Anu Bharadwaj; and CRO, Cameron Deatsch, will join me for a moderated Q&A session. All questions must be submitted via Slido. [Operator Instructions] As a reminder, due to the proximity to earnings, we won't be taking any Q3 financial performance questions today. And finally, we will conclude the session with a short social reception at the end. Now before I hand it off to Cameron, I need to mention our typical legal disclaimers. This session will include forward-looking statements and is likely to include discussion of non-GAAP financial measures. Information about risks involving forward-looking statements and limitations regarding the use of non-GAAP measures is discussed on this slide. Please take a moment to review this information. Okay. Now with that, let's give a warm welcome to Cameron Deatsch, Chief Revenue Officer; Sean from Reddit and Oxana from United Airlines.

Okay. Thanks, Martin. The energy is palpable, man. So pretty difficult to pronounce last names is Oxana Trotsenko.

Yes.

Sean Joerg.

Yes.

Cameron Deatsch. I dare you to spell any of those. Thank you. We have about 30 minutes here. He's a 2 of our fantastic customers, and we're going to go through largely their journeys with Atlassian. They've made some pretty significant choices over the last year or 2 with Atlassian tools. And I think it's highlights a couple of the key strategies that we've been pursuing over the last couple of years at Atlassian. So first off, let's just quick intros. Sean, can you tell you everyone a little bit about yourself and your company.

Sure. Yes. I'm Sean. I lead IT at Reddit. Reddit is a community of communities. So we're the place where you go to find out how to raise a dog or how to build a house or kind of product allows individuals to kind of share those comments, vote on them, share stories, et cetera. We have roughly 50 million daily active uniques. So we're pretty cool, we've been around since 2008 -- or pardon me, 2005, there we go.

Great. Oxana, a little bit. I think we all know of United very well, but give us...

Yes. So United Airlines, we are in the business of connecting people and uniting in the world. So if you have not flown United recently, please join us, we would love to have you as a customer. I lead Digital Transformation and Agile Transformation for digital technology of United Airlines. I have been with the company for 4 years now. I am behind United's first Agile Center of Excellence. And now I'm working on building out the center of enablement, what that means for the folks in the room is we're focusing on how to scale a method for scaling our digital products, right? Scaling the delivery and really working on finding efficiencies and shaving off time of the cycle times of how we do it. So before leading digital transformation, I was leading United's largest portfolios of products, delivering industry-leading innovations like Agent on Demand, if anyone has heard of that. And prior to United, I spent about 13 years in continuous improvement. So I kind of bring all my experience doing the transformation in public and private sectors to United today.

Okay. Well, thank you both for joining us today and, of course, taking your time sharing your story. So Sean's dig in on you for a little while here. So tell us a little bit about Reddit's history and use with Atlassian like...

Yes. We're a fairly old customer and company I think Reddit is -- or pardon me, Atlassian is kind of the underpinning for tech and product at Reddit. So it's how we track work. It's how we measure success and output. And it's been -- I think Reddit is an interesting company in the sense that we are fairly old in the tech scene being around since 2005. We've gone through various stages of our evolution. And I think Atlassian Jira, Confluence Jira, software management specifically has kind of been a part of how we've grown the product over those years. Now kind of my focus within IT is really transitioning the organization from being more of an ops kind of focused support organization to a true enablement organization. And for me, Atlassian has really helped us start to bridge out and make those connections within the business units, start to pull together kind of the technical and non-technical pieces together and help us look at our strategic commitments as a company and talk about and measure how we're actually succeeding or failing against those or tweaking or changing how we go. So kind of that connective tissue between the organizations between tech in product and kind of the business units that run the rest of Reddit.

I think this is a common -- pretty common story with Atlassian customers. We kind of land somewhere in a technical team, finding track and managing software development work, then we kind of grow and expand from there. Can you talk about -- I'd consider you a modern technology company. But can you talk about like what collaboration means -- like what's the culture like inside of Reddit? Is it -- was it the tech teams and the business teams or like everyone is showing up the meetings also like if you had to define what collaboration at Reddit looks like? Like what it was and what it's going to be?

Yes. We -- Reddit, working at Reddit being like Reddit is almost like being on Reddit as the product. We like to -- there's not really a hierarchical conversations that happen. We all kind of get together and talk independent of org chart or structure where you sit. And interestingly enough, we're actually really a relatively small organization part of the pandemic. We're roughly 400 people. Now we've grown to a little over 3,000 with contractors. So it's been through that growth over the pandemic figuring out how we actually work together, how do we maintain culture and connection. And for us, a lot of that has rolled into how do we use the tools effectively in order to maintain those connections. When I started about 2 years ago, the majority of Reddit at that point hadn't met their fellow Redditers or their fellows [indiscernible]. So we're using technology as this function to kind of get us to stay in touch and talk and not necessarily as a crutch or cudgel for culture but as a way to enhance culture and collaboration. And the things kind of grew organically out of that. So a lot of work that I've been also focused on now really to as kind of NxStage in collaboration beyond just the asynchronous stuff that you do in Slack or the document that you write up in Confluence or the meeting that you have in Zoom is focused on how we think about our large-scale company initiatives, which is a new muscle for Reddit prior where we were a little bit more siloed. And then how do we get all of those teams to talk about how they're succeeding, how they're getting their road map planning together. And Atlassian is helping us do that through a variety of ways. Jira Work Management recently we were 1 of their early customers in the Together program has really helped some of our teams to kind of take that line between Jira Software Management, service desk can start to create some really powerful forms through some of that injection and then start to track their work in a way that maybe before they really weren't. They were using tasks in Google or something like that in order to take a look at like what they were doing. And now they've got a cohesive way cultivating a backlog, powering our team meetings, things like that, so.

Got you. So over the last few years, we had, I think -- do you feel like, well, a different technology go from 400 to 3,000. So hired a bunch of people, had to do it all remotely due to the pandemic. And I also believe you had kind of a proliferation of tooling as well, right? Let's collaborate more. So we just get more and more tools to go collaborate. Can you walk through -- like it's all great, but like some of just the how did the challenges of all of that massive scale and all these new tools coming in? How do they actually like show up in the business? And what did you decide to do about it?

Yes. How they show up in the business is they just appear, right? And I say that somewhat jokingly, but also seriously is without strong guidance or controls, teams will go out and buy software. And the beautiful cloud world that we live in today and SaaS, it makes it really easy for teams to say, "Hey, I have a problem. I saw this thing. I'm going to put a credit card out there, I'm going to buy it. And without centralized controls, which we really didn't have at Reddit, we were small and scrappy, teams were just going out and buying whatever kind of fit their need that day with no real forethought around okay? Does this something -- is this something that someone else could use, et cetera? So that was the problem statement. And when I arrived, it wasn't anything that anyone was really thinking about. We knew that we were spending a lot of money on software and various tools, but no one was kind of thinking about like, hey, well, how can we do this better? What can we consolidate on? Can we pick a method or a process that will help us speak the same language so that we can move more effectively together. So some of the things that I've started to tackle and we put in place or one, understanding what's out there from a software perspective and then being strategic about which ones have the most impact, which teams are using what. Also working really closely with like our TPM organization, who is responsible for helping tech and product work very closely together cross-functionally, helping them set guidelines within how we work within Jira Software, kind of archetypes on projects, moving the company from team-managed projects, the company projects in our cloud journey was really helpful in getting at least everybody talking about things in the same way, measuring things in the same way, having the same components that they're using there. And then more broadly, we've started to institute a software review council. So we're actually now going through and evaluating software requests as they come in and then sitting down with stakeholders when they have a new request and figure out, hey, how does this fit into the overall picture, how can we help enable you there? What tools do we have in our backlog that can actually do what you're looking to do here and start to curtail that. So it's a journey. We're still early on in that journey, but we're making headway.

Great. And obviously, over this last year, something that we've attempted to do is help organizations at least simplify their decision-making with Atlassian. So 1 of those, especially in this broader work management space is our offering of Atlassian Together, bringing all of our work management products, Confluence, Jira Work Management, Trello, Atlas and Access got all 5 into 1 offering. And obviously, you guys went forward with that. Can you walk us through the decision around that? Is that going to coexist with other tools? Are you shutting down other applications? And then how is -- how are you convincing the rest of the business to actually standardize on the Together tool set?

Sure. Yes. I think the consideration there was project management tooling actually is 1 of the most prolific that just kind of spreads throughout an organization. Every team likes to measure or track work in a slightly different way. So for us, it was one, having some resourcing already existing around the Atlassian space; and two, having such a strong technical and product bias around that. It made sense for us to think about, well, how can we bridge that gap and get Atlassian and Jira and other certain, Confluence into the hands of more people and maybe address some of those smaller one-off equations and start talking about, hey, this is what it can do for you. I think one thing that I've noticed as well as Jira kind of can sometimes be intimidating to a lot of users. So a lot of our conversations there are oftentimes taking the place of educator and explaining and showcasing the products. And -- this is what we can do. A lot of people may have perceptions around what the product used to be versus what it is today. So it was a little bit of product show and talent times around, "Hey, you're saying you want to do this. Look at what we can do here and here's how quickly we can do that." So that was really 1 of the big draws for us was the ease of use of kind of Jira Work Management and that Together program as a whole, showcasing things like, hey, if you're doing simple task management, and you want to frame your meetings better. Hey Trello is a really great way for that -- for you to do that. Like look, you can set up a basic Kanban board. You can have tie-ins to some of the other services and things that you're looking at and you can quickly orient your conversations around that. Or hey, you have a basic form that you'd like folks to fill out, you don't really have like a backlog of work. But look, let's put this together, we'll throw it in Confluence, you can get some basic information and then you can start to cultivate some work that you need to do that might be more of a planned work process for you. So it's actually, in a lot of cases, the conversation that we've been having with folks, the product somewhat sell themselves once we show what they can do based off their needs. But we are taking kind of that step of educator in a lot of cases.

Got you. And then just 2 questions on [indiscernible] Oxana, which is like, one, any significant pushback or hurdles that you've seen as you've tried to roll this out. And then the other side of that is benefits, hard benefits, soft benefits like what are the -- as you started to roll out with Together offering for the rest of the business, what are the value props you hang your hat on?

Yes. I mean I think it's always a conversation, and it's always a negotiation. People have very strong feelings about things that they've chosen. So I think some of the conversations there have been more focused on trying to really pick apart the output that you're looking for and then removing the emotion and having a more natural conversation around tool consolidation costs, resourcing associated with some of that as well is a really big lever to talk through. But it's also kind of the intangibles of like I fortunately have a really wonderful leadership team that wants to have Reddit focused on more things cohesively help us work in very similar fashions. So having that backing also help those conversations be a little bit easier, like -- we get it. We all want to track work in a very specific fashion. But if we're all moving in 10 different directions where none of us are moving in the same direction, how can we start to get people kind of in line and talking and about things in the same way. So it has been a give and take. It has been my job, I don't -- 1 thing that I really love about kind of the work management and the Together piece is I don't really want to be the organization that takes things away from people. What I'd like to do is let people work kind of the way that they want and give them a tool set that gives them the flexibility to inject Google Docs into the equation or use spreadsheets to do things. Or there's this very specific tool that they can just inject into whether that's Atlas or Trello and kind of work in the same way that they need to. And the product gives us that capability, which has been really great. And then in terms of what we're seeing on the back end, the value back, the soft kind of intangibles, I think 1 selfishly for IT. We're gaining trust with the organization. One of the things that's always a challenge for IT is getting folks to come to you with requirements and not just solutions. And now we're seeing more of that as we're deploying services and tools that people are really excited about or that they see applicability towards. So it's been extremely useful for that, for my team to gain that trust and credibility of like, hey, we know what you're looking for, and we can help provide that solution for you.

Okay. Thank you, Sean. Oxana. I'm sure 3,000 people is a drop in the bucket at United, but...

Yes, we're growing.

A different level of scale. So you mentioned like you're going into this digital transformation, you deliver digital products to United. Can you help like explain what are digital products for this -- for the audience here? Like what does that actually cover across your organization?

So to give everyone a perspective, right, where United is at today, we are on a trajectory of aggressive growth. So we made a historic order to purchase 700 aircraft of our -- by 2032. What that means is that we're delivering 2 aircraft per week this year and about 3 aircraft per week next year. We're also planning to increase our employee pool by 50,000 employees over the next 5 years. So you can imagine that the growth and the need for scale is massive. So when they talk about digital products, there are 2 sides to that. One is the employee-facing side because we have a ton of employees, right? You guys are all using the airline as a product to get here or anywhere else for that matter. But those employees use specific digital products to meet your journey successful. There are people on the ramp. There are people in the lobby, there are people at the gate, they all use digital product ecosystem to make sure that we're actually effective as an airline. The second piece is customer-facing. So customer-facing digital products, the easiest is probably the app that some of you might be familiar with. So that's just 1 example, but other examples also include our MileagePlus offering, right? So some of those things, how we engage directly to consumer and also B2B pattern, they all include a form of a digital product. So when we look at this massive growth opportunity and how United is planning to scale, we quickly realize that, oh, we need to catch up quickly because a lot of things at United have been monolithic in the past. And so we went on a quest to figure out how can we actually move from that framework into micro services. Some of you are familiar with AWS that allows us that ability to go to cloud, right? So we are in the process of doing that, but much quicker than we initially anticipated. The second piece is really going after that time to market and transparency. So where the product teams who participate in the development, right, my business needs to know what's going on? How quickly are we moving? What needs to be done? And my dev teams need to be -- have guardrails in place. We have to continue to shave off of time of how people develop product. So it's this massive ecosystem of things that we develop within United, but ultimately, they all represent that digital product space.

Got you. And I say we were going to go from monolithic, we want to go more [indiscernible]. It sounds like these digital products are core to how the whole organization runs other than the actual planes fly off in the air. But the -- how do you frame the challenge and say, "Hey, we need to do something different was it taking too long where your competitors are innovating you? Like what were the core challenges that you're trying to address through this major transformation you've been going through?

So during COVID, we went into something we call [ scrappy ], right? So it's a quick cycle to innovation, right? We are able to gather small teams together very quickly. We're able to what we're looking to do, right, to design thinking session and then develop products, super fast. Well, coming out of COVID, as you understand, the need for more products is increasing. So now we go back into what Sean shared, right, communication, how do people talk, how do we provide transparency. How do we show to leadership, hey, this is where the portfolio is at? This is the progress because typically, I would produce a deck, right? Can you show us the status, lovely slides are awesome, everything looks green. It's a cold watermelon approach when it's green on the outside and the red on inside. So a lot of that stuff going on, right? That's what we were doing pre-pandemic. But now we're like, hey, we really want to be good at this because we want to excel for our customers. So taking that, right, and looking at opportunities to really how do we scale is what drove these decisions, right? So digital transformation continues because we're trying to deliver much better products, much quicker for our customers.

Got you. And United has obviously been a long-term customer in Atlassian. It's like 1 of actually our early 2000s, I believe, the first license of Jira is bought at United. But you largely kicked this off as almost a greenfield brand-new project. You're going to say, hey, we're going to operate in [indiscernible] we're going to look at all the tooling that's out there, in-house out of the house, and we're going to standardize on a new way of working. Can you walk everyone here like through that journey? Because it's interesting that you're a legacy customer, but largely treated this as a -- almost like a net new customer would. So can you walk us through your decision criteria, how that process works?

So we set on the quest first, right? I mentioned moving from monolithic to microservices, we set ourselves on a quest to see -- to find a platform that would support DevSecOps. We want to be secure, right? There is a lot of threat in the risk and digital space. And we wanted to make sure that our DevSecOps teams were set with the right platform. We found Harness as our key partner to do that. And then we started to say, okay, we covered the DevSecOps process. Now what tool will it help us enable the product side. So the product side is where you get your business stakeholders, your leadership engaged right, UX teams, people who do design of experience and interaction. And we've got those folks together. So we said, hey, Atlassian might seem like a great fit. And some people are like, yes, a lot of Atlassian like let's do it. And a portion of those people said, oh my God, I as your DevOps is everything. I mean I'm not going to even talk to you because you're talking to me about Atlassian. So I quickly realized that that's almost like a religious question, right? Like what do you believe in? What do you really believe in? And so people get super passionate about that. And I was like, okay, guys, I'm not playing this game, right? Let's actually look at facts. Okay, product owners want something. Flowmasters want something, leadership wants something. I want to manage effective cost, right, change management to be easier. So I talked to my developers. I talked to my product owners. I talked to my Flowmasters, and they had this operational side/leadership angle as well. We put together a 58-point comparison. We took all the features, all requests of people who are really passionate about, put them on a spreadsheet and we ranked them. And actually -- so 3 products that we were evaluating. One was Monday, which quickly fell off the track because just they don't provide us ability to scale. The second piece was Harness integration was critical, right? So that was almost a deciding factor for Atlassian suite. Azure DevOps and Atlassian actually came really close, so we were like 74% coverage of what we want, 75% for Atlassian. And my DevSecOps team, they're like, we need to scale at such a rapid pace that the integration between Microsoft and Harness is not robust enough. There is no guardrails. There's no automation, request for change is a little bit different when it flows through Microsoft suite. So we looked at it and we said for this group of stakeholders, right? This is a much better solution because we can provide transparency, we can cover our DevSecOps team, and we can tackle this conversation in a cohesive manner, which is -- so right now, we're really happy that we made that decision.

Great. And I think that just supports the -- our entire strategy around largely open DevOps. The reality is that there's always going to be a next set of great best products and tools to help deliver software, monitor software, operate your software, and we're trying to integrate across everything. So it sounds like that integration focus is really a key element your teams.

Definitely. And I think once we build the pipeline, right, so we have infrastructure as code concept at United is what we practice. So we ingest architecture and we split out the infrastructure. So automated developers don't need to do that work. It can be automated if you set that architecture in a proper way. So we shaved off from -- we went from 22 minutes per deployment to 5 minutes per deployment. And that was actually super critical for us because we integrate with AWS, GitHub, right? We have Harness, we have ServiceNow, my DevSecOps -- SecOps team are begging me to switch to Jira Service Management.

We can help with that.

And so there's a lot of opportunity, right, as we build this infrastructure, and we realize that, that's delivering us what we're getting -- what we're hoping to deliver these integrations, right? Atlassian is not locking us into 1 ecosystem. And actually, that was another differentiator because Microsoft was more like, hey, you guys are a Microsoft shop, Azure DevOps is a product, but it's actually a DevOps product. It's not a product management tool or platform by any means, right? So for us, it was a big, big win in terms of Atlassian like this is what really we should be going with to scale.

Great. And I would consider a very large enterprise, but you started -- you basically -- this entire thing, you chose Atlassian Cloud from day 1, I know you have a large [indiscernible]. But I mean, was it just not even a choice like we're going with the cloud offerings? Or was it an actual decision that you had to make along the way to choose our Cloud offerings?

I think for us, again, 2 things, right? We're trying to scale, we need speed to market. So speed to market is critical. The last piece is transparency. So it's almost like a no-brainer if you're building a modern setup and a pipeline, how you develop software, you cannot go with anything other than cloud, unless you have to have full control over supersensitive data, which is for the pipeline build out, right, the mechanism, the sausage-making thing. It's not the area where we needed that control. And so after talking to a few folks and hearing out conversations about, hey, you know what, we had outage for 4 hours was not fun. I would never want to do this again. I would never want to deal with us at a scale of United, cloud is we must go cloud. I mean there was actually no convincing anyone. People that are like, yes, cloud enterprise, I mean this is great. What else is out there, right? So it was the first choice, not much discussion. But now we're taking all those other teams similarly to what Sean was sharing about. If you don't have guardrails, right, Atlassian provides flexibility for smaller companies to just use your credit card and pay for a company like United, you can imagine 95,000 employees, if each of them creates a team where they're paying with a credit card, it's a disaster for a corporation to manage to consolidate, right? How do we bring all these folks together to talk to each other. So I started this journey and I'm like, hey, I have this 1 team that does most of the software development within United. And then my procurement team sends me a list. I looked at the list and I was like, okay, well, 5 more teams I can combine them fairly easily, right, go from on-prem to cloud, not a big deal. And then Atlassian was like, hey, Oxana, there are another 30 teams out there in United paying with a credit card. So at which point I was I was like, okay, maybe I was not fully prepared for this emotionally, but nonetheless, I think we have to have a form of control at United just because that content role or what people are doing, right? I don't like taking things away from people, but I do because it's such a large scale, you can -- it can be like, hey, Cameron is going with something and then Sean is like on different products. We need to talk to each other versus being siloed when we scale to the scale of the size of United as a corporation.

Yes. That's -- it's great to hear that like SaaS [indiscernible] and credit card, the ease of that is a problem even in an enterprise as large as United, so.

Okay. Last question for you, you attended the keynotes this morning. Have you seen some of our presentations and announcements?

Yes.

Okay. So you're going to leave here today and seeing some new things. What's the most exciting thing that you're going to take back to your office tomorrow or the day after when you head home, Sean?

The new AI functions are really exciting. I think that's a no-brainer. I'm also really excited about the improvements in analytics. I think that's been something we've all been looking for, for a long time and having that capability now directly within Atlassian is really exciting, so.

And the combination of the 2, using [indiscernible] to get better...

I know...

It's a brave new world. Last, Oxana?

I think for me, personally, the difficulty of using JQL is something that you have to overcome as a large organization. There are a lot of users who have no desire, no need to learn JQL, and so I think the plug-and-play version of, hey, feed me a question, I'm going to give you an answer, and I'm going to populate the JQL code. I think that's great. . The second is discovery, product discovery. We actually -- the intake, right? Again, large organization, and you want -- we support everyone and then you want who has an idea, right? We want to hear it. We want to make sure that -- we're actually doing something about it, and maybe that's the next best product for our customers. So we want to make sure there's a good funnel. And when you funnel a large corporation like United, there is no good way to gather ideas, right, who is managing that? Who is looking at that? Where do reviews happen. That has been a struggle in many other project management solutions. And I'm actually super pumped that there is a consolidated way of taking that and placing it on like an impact matrix view and say, let's get a conversation about this, what makes sense for us. I can tell you 1 thing. At United, we don't do prioritization because there's 100 priorities on the list. So we started doing -- and I have a former CEO here of United, who introduced this concept, who introduced the concept of picking the top 3 things, right? So not just working through this prioritization exercise but actually picking the top 3 things and saying, we're going to just hit these things off. When we're done, we're going to focus on the next set. While that product is great to get us closer to like what are the top 3, I think you have to have an internal mechanism for all these things Atlassian, how to work with culture in the process, right? Tools don't give you agility, tools don't give you a new culture, tools don't enable you to talk to other people. You actually have to go and talk to other humans. So I think I'm super happy that we started the transformation with integrating the culture piece and the process piece and then following up with tools as a fast follow to that because just taking a tool or taking an ecosystem or platform does not make you any better. At the end of the day, you're dealing with human behavior. And so I'm excited about the product discovery for sure, and then AI capability and shaving off of time of JQL.

All right. Got it. Couldn't agree more or Jira product discovery, It's -- I mean it's going to be just anyone who's managing products and product delivery and Jira software, it's a complete no-brainer. You're going to have a product manager needs to put in their backlog and prioritize...

You know what's my most favorite part is? It's free.

There you go. Yes. Okay, for a time being. Hey, All right Sean and Oxana I could not thank you enough for taking time out of your day to come up here and share your stories with all these individuals. Can we give Sean and Oxana a round of applause please?

Thank you for having us.

Martin, you're coming back up.

All right. So on the Q&A panel today, we have Co-Founders and Co-CEOs, Mike Cannon-Brookes; Scott Farquhar; president, Anu Bharadwaj; and CRO, Cameron Deatsch. As a reminder, we'll be taking questions via Slido, and another reminder that we will not be taking any questions about Q3 financial performance today. So with that, let's get into it.

So the first question comes in from Michael Turrin of Wells Fargo. Intelligence product announcements captured a lot of attention today. How should investors think about monetization of added value functionality that enables given current pricing is mostly on a per [ stage ] basis? And what drives differentiation for Atlassian and leveraging AI plus large language models? Anu, do you want to start with that one?

Yes. I can take that. Thanks for the question. Let me answer in the reverse order. In terms of differentiation, at Atlassian, the way we think about infusing AI into our products is really how can we use the transformative power of AI across all of our different products. So we can rethink how collaboration is done. So today, this morning, you saw the demos we had around how we bring generative AI into our products, which is step #1. But if you think about it as how do you strategically play ahead, the vendors that have a real advantage in the AI wave are those that have a large swath of customer base, large swath of customer data to play with. And at Atlassian, we have a unique advantage there in terms of we have 250,000 customers that we serve across SMB as well as enterprise. So all kinds of natures of customers as well as industries, which positions us in a really unique differentiated place of being able to have the broad surface area of different kinds of customer data. Second, we also have very unique shape of customer data in that we do both service-based work, project-based work across different kinds of teams across developer teams, operations teams, business teams that does -- that do non-technical sorts of work. So as a result, we have a unique data shape that we can form into what we call the teamwork graph, which effectively connects different kinds of work to different personas across multiple kinds of teams. As a result, the richness of the teamwork graph and 1 that actually works across different types of teams gives us the ability to have that kind of graph across a very wide swath of customers, which is a strategic differentiator. So positioning the orchestrator in between the teamwork graph and the wide swath of customer base that we have really places us in a very unique place to say, "Hey, let's go figure out how deeply we can intervene AI into each of our experiences." You want to take the monetization question, Mike?

I was actually going to give 1 example. I think because people talk about differentiation, I think our differentiation, if we assume large language models are going to be free, freely available, anyone come up in ChatGPT and go talk some stuff in. Your differentiation comes in how do you use that. And it's not just having a customers' data in terms of their content, it's actually knowing about their pattern of usage. So a feature we didn't show today that's coming as well is the opposite of the JQL feature I showed where you write what you want, and it gives you the JQL, we have the opposite where you can hit JQL and you can hit explain, right? And it will tell you what that JQL is doing in human language it sounds silly when JQL simple. We have migrated customers with 10 years' worth of data. Someone wrote this query like years ago, it's 5 lines long, it takes me 10 minutes to figure it out and I hit explain and it will tell me what it is. You can do that actually pretty well. If you go to ChatGPT and say explain this JQL team. What we can do that's differentiated. It's understand that user when they're hitting that button. So they're still just hitting one button. But I can say, "Hey, explain this to me like I'm a 6-year-old because I know that person is on day 3 in Jira versus explain this to me like I'm Jira admin, who's been doing this for 10 years and just need know it, right? So that is our ability of knowing is that user is 1 day old or 10 days old. As an example of the type of data we have that we can use the technologies much more effectively to give a great customer outcome than otherwise are totally generic. Migration -- sorry, monetization question. Yes. Look, we believe fundamentally, these technologies are going to appear everywhere. And it's a huge advantage to us in being a content-driven company, right? We deal with a lot of techs and these give us huge advantages to build 5 better products for customers. Again, we're chasing the Fortune 500,000, as we've always said, and that will -- we think these can give us a huge accelerator to get there. We're really excited about the potential of the features. There's also an obvious migration story to continue driving migrations. It's a huge carrot for the cloud for customers with more differentiation. And specifically to the example I just gave, the more data you have from your years of history on-premise that you move to the cloud, this unlocks a lot of that data, right? The knowledge graph features we showed, explaining things like older JQL, et cetera, adds up to, I think, a significant continued advantage to migrating customers.

So staying with the AI theme. We have 2 questions of the same vein. With AI enabling so much developer productivity, how do you view the seat demand outlook for developer suites? And what does this mean for the value and people intensity involved in creating software and what does this mean for Atlassian? And then similar vein from Alex, you can -- Atlassian Intelligence is highlighted to bring 50% efficiency improvements, clearly, very powerful. How are you thinking about the impact on agents, a company with staff before and after such AI capabilities.

I can take that. It's a good question. I think it harkens back to a fundamental belief as a company. We believe that every company is going to be a software company and the demand for digitization. The demand for software is only going to go up, not down, right? So fundamentally, what we're seeing with the AI transformation is it's easier to make software. And therefore, we are no longer constrained by the supply of developers, because what the code-generating ability gives us is more and more people can actually now be creators can write an app without having to know how to code. And overall, that's a fantastic thing for Atlassian. That's a fantastic thing for overall, the amount of software that gets generated and deployed in the world outside. For us, specifically, as we think about what it means for Atlassian, more and more people will be able to create software and therefore, will need to collaborate with each other. And we don't just operate with specific developers. We don't just build tools for developer specific personas, we build tools for teams. And therefore, the number of teams is going to actually expand as more people are able to build more software. And having the tool set to serve entire teams around that and overall expanding the number of people who can be called developers, who can be called creators, is a net-net benefit for Atlassian and for software manufacturing tools, I would think.

I just wanted to correct. I think I heard the 50% improvement to start. I don't know if that was from keynotes that we had, just to be clear that our virtual agents internally now handle about 50% of our internal IT in other cases. So that's not necessarily a 50% improvement to handle 50% of the cases, right, and save a significant amount of time. What that does lead to is, I believe, more fulfilling jobs for the people doing the rest of the work. They don't want to be doing password resets all day long, right? They get the more complex and interesting and advanced cases because the virtual agent is able to handle more and more and more of the simple cases. And as they get more volume with those advanced cases, they'll automate them and build the tools and then it will keep moving up the stack. So it's a different example of more fulfilling work perhaps.

Okay. So the next question comes in from Ryan MacWilliams from Barclays. It makes sense that Atlassian Intelligence will be exclusive to cloud-based offerings. But can we expect most new feature development to expand off team's cloud platform from here. Mike, you just kind of touched on this a little bit earlier as a carrot for migration. Do you want to start touching on this and talk about the power of the platform.

Sure. I mean I think the Atlassian Intelligence features for sure, highlight the power of the Atlassian platform. So I'm not going to say that 5, 7 years ago, we saw this all coming, that's bulls***. But it's certainly -- you see the more we have centralized data, the more we build platforms and connected tooling, the more we work on our internal architecture. You're always asking why we spend so much money on R&D and where does it go? These sort of transitions are where that really starts to pay off we have to execute on that, but that's the theory is we can build a lot of these things. We can show a lot more connected value to customers. The cost of building those features for us in the cloud and the speed of building them with the platform is just significantly higher, right? So that's sort of the reason that you're seeing those features there. Besides, obviously, most of the large language model stuff is pretty tied to the cloud by default. But I do think it's a validation of a lot of the platform work we've done in bringing things together. You saw in some [indiscernible] is the tying of automation in, you saw how analytics is feeding in and having all the cross-product story, you saw us using a lot of third-party data from open tool chain and other areas and in work management, this is all kind of having it all there is 1 thing, being able to now build automation and intelligent features and other things on top of that is a unique advantage we have, I think, in the platform.

Great. I think 1 last AI question is from Ittai Kidron from Oppenheimer. How would you expect the use of generative AI to change over time, i.e., how do you move from more content discovery to more content creation flow automation?

I think it's different in each of our markets. So when we talk about -- like in the software space, and DevOps, I do think we're fundamentally supply constrained on developers. So it will enable us to unlock more of the development in terms of generation of things. That leads to farmer software being created, which is generally good for us, especially for things like Compass because you have far more of a management problem here. I think in the broad generation, what does it change? I think there's a human skill change needed, right? You have less -- fundamentally less journalists and more editors, right? People need to be good at asking questions, defining what they want to get back and then editing your response, not necessarily like an editor is doing, but you're kind of checking, auditing whatever the response is, right? You can ask a question of Jira, and we'll give you the JQL and run it for you. You've still got to do a little bit of editing and auditing, right? Like you shouldn't just randomly expect that to be correct. You've got a little bit of checking. So I think this changes the nature of work. It's definitely huge for productivity. But I think that changes the nature of what people have to do, right? They have to know what they're asking for when you're starting out on a task. I think we'll all get good at working out as this stuff appears literally everywhere, like it will be in every single company and every single piece of software, how the vendor implements it, et cetera. Maybe an example you can see in our early thinking here in a lot of the design aesthetic. You've got to remember that these features have to deal with. It's not just generating text and splitting it out. It has to fit in. So you see us using a lot of the rich text and features that we've built in the editor, again, the power of the platform. But secondly, you have to deal with the fact that this stuff takes a bit of time, right? There's a -- it's not an [indiscernible]. You have to deal with the fact that answer might not be correct. So the user has some editing capability and you have to make that clear that this feature, this button is different than that button, right? This 1 gives you a definitive and deterministic answer, and this 1 gives you probably [indiscernible] that will take time, but can do super powerful things. I do think that's a design aesthetic. The world hasn't standardized on how that will be yet, but I think we've had a really good shot so far.

Okay. So switching gears -- we have a question from Keith Bachman of BMO. For Jira Service Management, how do you think your features and functionality has changed in the last year versus the competition?

Yes. I'll just go back to the -- are you seeing us? [ On the screen ] we're going to just go and beyond -- its 3 years ago now. Over the last 3 years, we've largely done -- completed or significantly increased our IT service management capabilities across the Jira Service Management offering. We went from largely a simple help desk to including incident management, configuration management, automation, analytics, messaging integration with Slack and Microsoft Teams and has been much more as you saw today. Virtual agents, the pro forma capabilities for having a much richer portal hot offerings. So if you largely come down to anyone who is looking for enterprise-grade IT service management solution for the largest company in the world and wants to go down the checkbox to list of capabilities required, we checked just about everyone in the boxes, right? And you see that largely the best version of that is to get into the leadership quadrant in the Gartner Magic Quadrant, like that basically shows that we hit all the capabilities and that we're executing. What makes us differentiated are the same things that always made us differentiated with Jira Service Management. One is the fundamental belief that legacy IT service management products that are completely separated from your development in DevOps tool chain, as like Oxana mentioned earlier is a disservice to the organization. We believe the teams that are building the software and the teams running the software are largely getting closer and closer together and not just the cultural change to make that happen. That is modern DevOps and IT but also the tools that can support it, we can do that on 1 platform. And I talk again and again to customers who might have a legacy IT system, they get tickets in that system. And someone an IT takes that ticket and turns, goes into Jira and enters the ticket again, development team solves all the work, and they take the ticket, solve, then they go back to the other ticketing system, like it's almost a 0 value process, we can turn that into 1 system. So can dev team and the team working on the same platform as our core value prop around Jira Service Management, working with Jira Software and of course, cost. We are a fraction of the price. We had a European customer, saved 70% on license costs moving from their legacy vendor to Jira Service Management. And on top of that, they can basically have way more configuration set up help desks in no time, set up for more use cases beyond just broad IT. And that's where we're going. So on top of that, we're layering all of the AI capabilities you saw this morning, and you'll continue to see us differentiate going forward.

And I'd add just from speaking to customers, the other additional thing we hear more and more of is just time to value because the benefit of us having used the same consistent philosophy in JSM, we want this to be broadly appealing and easy to adopt and implement is a real advantage in this particular economic climate where people are looking to consolidate more and more and also implement systems that are easy to see value quicker. It's become a lot more important to the question of, over the past year, what has changed, we see more and more customer demand in that area.

So we're hearing a lot that JSM is being used a lot outside of ITSM. How much of that usage does that represent? More importantly, how does that proportion change over time?

Yes. So this largely is going from IT service management to enterprise service management or as what we call outside cost call go/help, which is, hey, I'm a team. I'm someone I just need help from a legal team, from a marketing, HR team. And that's something that is definitely part of our strategy. Of course, we want to have the IT team standardized on our process. But you can set up a desk with just about any department. It becomes that key contract between I'm a user in the business, and I need to work with any department very simple to set up those service desks. We haven't quantified. I couldn't give you like, hey, is that 30% of the overall user base today. I can say -- and if you talk to any enterprise customers who use Jira Service Management out there in the audience, I'd say almost every single one of them will speak to one or multiple use cases that are outside IT when they're using Jira Service management. Simply because it's so easy for them to get it set up, like they can basically do it in an hour or 2, they can get a whole new help desk, customized and routed to those users. So I'd say it is very prevalent across the board, and an area that we, through automation, through AI, through the Slack integrations, we're making it easier and easier for more and more business teams to address. So it is definitely a core part of our strategy going forward.

And then while you're speaking Cameron, I think investors oftentimes have the question of how do you reach deeper into the enterprise in ITSM? How does the sales strategy need to change? And do you need more of a rep presence to speak to that buyer?

Yes. Absolutely not, the end the credit card joke. No, that's the -- we have invested in our enterprise advocate team over the last -- ever since we launched a data center in 2014, and actually, the first data center is actually out there here today. We've grown that same time. So we do have account executives. We call them enterprise advocates, covering largely our top multiple thousand accounts, and they're focusing on migrations, enterprise agility, IT service management and broader work management with Atlassian together. We do have dedicated ITSM specialists in that motion that we've invested in. And what we are going into is existing Atlassian customers. So we -- that we know are using Jira Software, largely standardized on our products, and we're coming in and looking for key use cases with those development teams and IT teams where we know we can add value in a very quick period of time. So we have done that engagement. You see us investing getting the Gartner Magic Quadrant as a leader is bringing people in, and that's just within my teams. We also have to remember, we have hundreds of certified Atlassian solution partners out there. We're also running around this big event today. We stood up dedicated certifications for IT service management. So we have certified trained IT service management professionals out there in our channel. And we've also increasingly partnered with global system integrators to do the same. So yes, we are increasing our relationships even at these events today, like we have -- like I just continue to amaze like we have our enterprise cab. We have CIOs some of the largest companies in the world attending these events, building relationships that honestly, 5 years ago, like beyond my hopes and dreams. So -- we are getting that presence. But more importantly, when we get into those conversations, our value props resonate, save money, move faster, Devan IT working closer together, just about every CIO I meet with, like nods their head at that.

Okay. So the next question is coming in from Keith Bachman. At the previous analyst event, you made comments on your views on margins and the desire to invest for longer-term growth. Can you provide your thoughts on optimizing for longer-term growth versus margins? And how economic weakness may shape your views. Scott, do you want to...

So we've been public for about 8 years now, I think, as a company. And before that, I resonate 3 years' worth of accounts. So you kind of got a decade worth of Atlassian accounts. If you want to go back and throw them all into your model. And over that time, we've proven that we have a very profitable business model. Our go to market strategy, what we do, who we serve, like we've done a great job of kicking off EBIT margin, but free cash flow as well over that period of time. For those of you new to Atlassian story, which I see a lot of familiar faces in the audience. So for those who are new to Atlassian story, about this time last year, we sat in a room pretty similar this, maybe the same room, and we talked about that we saw so many investment opportunities in the near term that we wanted to make a deliberate choice to invest behind those opportunities, deliberately take our margin percentage profile down in order to go off those opportunities. Those opportunities were and still remain migrations, ITSM, the broader work management opportunities that we have across that and enterprise. And that will be the same ones we talked about last year. And the reason for that is that the ROI for all those motions for all opportunities was so high. We've only increased float migrations. People are banging on our door saying, I want to move, I want to move across, but I just need some handholding or I need you to improve the migration tooling behind that or your cloud needs this one extra certification that we don't yet have. So when customers are banging on your door with revenue opportunities and you just need to tactically go through those things, it makes sense to invest behind that opportunity and not just wake over time. So that was last year, and obviously, what has happened in the capital markets in that last year. And we've made choices around rebalancing even further behind those opportunities and taking other opportunities down in order to go behind those ones we talk to you about. So as an investor, you like, okay we've taken margins down, what does that look like going forward? And there's some things that I think are important. One is that we made a very deliberate choice to take margins down. We told you ahead of us doing it , like this didn't happen to us -- this is not because our cost of acquiring customers went up or our cost to provide support to those customers when obviously very deliberate these the areas and are the same areas a year later that we're still investing in. So then, okay, when does that investment year? What does that look like? And when we look ahead through FY '24, I still see opportunities for us to continue investing behind those areas through FY '24. I would back and then say, okay, well, what does that look like in terms of returning margins to what the historical averages look like and I feel very good about that. There's nothing fundamental in our business, in our COGS, customer, anything in that that makes it hard even to get back to those historical margins. At the moment, it just doesn't make sense for us on a short, medium or long-term view to be doing that when we can invest to get our customers across the cloud and enterprise and IT and all the other opportunities we've talked about. So I hope I answered your question.

So I think another topic that's top of mind for a lot of investors is cloud migrations. So a question coming in from Gregg Moskowitz. Despite all the success you've been having with cloud migrations, a lot of your customers are still on server. Do you and your partners have enough capacity to help all server customers who want to move to cloud over the next 10 months?

Yes. It's been a big part of my job over the last 2 years talking to server and data center customers, getting them to start the journey towards club. Many chose cloud immediately, and we're happy out there and give references and CSAT surveys and out there talking about all the benefits and making all the other customers jealous, which is great. It's the best thing we can have. We have many that are in their journey. They've chosen cloud, but it's their large organization, they're doing in steps month over months, potentially over years as they're moving their organization along. And then we have customers, as you can imagine, that are going to wait until the last minute. And the last minute is February of 2024. So we have a good 8, 9 months to make that choice. The reality is -- the good part is every one of those customers knows, no one is surprised. So we've been sending about 2.5 years since we made this announcement. We've been sending them e-mails every single week and talking to them. So everyone knows this is coming, they're planning appropriately for it. And almost every single one knows that we provide cloud migration trials, different licensing capabilities, extended evaluations and migration tooling available to them. They have to self-serve their way through migrations. Bulk customers on server, our migration toolings for the bulk of those customers, relatively small, not too customized implementation systems, our migration tooling they can simply migrate to the cloud in an afternoon in a week with the same amount of plan. So I'm not too concerned. On the bigger side of town, it's one of those customers, if they are on server and they have 10,000-plus licenses, and there are those customers out there. They're going to make a choice in February. We are going to obviously get them to choose cloud, first and foremost, of a camp and data center as a backup, either of those options, we'll make sure that regardless, they will be able to continue to use our software going forward, but they'll have to make that decision. It does not mean they have to migrate all of their data to the cloud that day, where you can be a -- make a license decision. If you buy cloud today, we let you continue to use your on-prem for roughly a year as we run that migration, makes sense as a migration program. So for the customers that wait, that's fine. We're ready for them, and we'll help them through that journey. We're going to do what's right for them to make sure we get them across the line, but they do have to make a choice of data center or cloud February of next year and while these purposes, we're making it really appealing to choose cloud.

So next question kind of going on the same theme from Brent Thill. Do you believe cloud migrations could be a long-term growth driver beyond the server deadline in 2024. And what are your future plans for data center. Scott, do you want to take the first part of it and then Mike, maybe the second part of it.

Yes. I think like Cameron is actually -- to answer that question in terms of migrations from data center of FY '24 in.

I knew it was actually jump in over answer.

So we're all very eager to talk to you about migration. So that should tell you something. So in terms of -- is it a long-term migration? Is it a long-term growth driver. The thing to remember in terms of the bigger picture for cloud migrations is that though February 2024 is the date for server end of life. Over the last quarter, about half of our cloud migrations came from data center. And like Mike was saying earlier, last week, I met with one of our enterprise customers, who is a very, very stong data center customer that we've been talking to him about cloud. And this time, he came in and he said, "I need to go to cloud." I nearly felt a lot of my share was like, oh, okay, what happened? And he's like, "Well, I know the AI features that are coming in, and this is a big draw for me to come over to the cloud. So we see clearly that the acceleration that data center customers experience, the draw that the experience to cloud is actually helped by underlining the AI capabilities. But in terms of the time line itself, like Cam was saying Feb 2024 is the server EOL date, but then customers have the option to say either data center or cloud. And we will continue to see data center migrations to cloud for the next few years to come. So this is like we have been saying repeatedly more of a multiyear journey to think about in terms of how much growth will it drive? So yes, it's definitely a multiyear growth driver and not something that's going to suddenly stop February 2024. Do you want to take the second part?

What was the second part? You did such a good job answering. I don't know.

What is the future plans for data center?

Look, the -- our aim over the long term is like how do we get customers to cloud, like that. We're going to continue to try and do that. But the practical reality is there's some customers who may never move either for regulatory reasons or their own internal reasons to do that, and we're going to continue supporting those customers for as long as -- in the foreseeable future. And so we're been trying to keep enticing them with carrots and sticks they got to move across the cloud. But there's no plans to end of life data center like we believe that, that business line still has a long way to go.

I think it's worthwhile, I just want to add one thing from my customer review, especially last week. One of the things that I do think is resonating with customers is our commitment. It's sort of a weird thing you're all looking for what do you put in cell D7, I get it. I don't know how you put commitment into one of these cells. But the customers have seen for years now, a commitment to building an enterprise cloud solution that's enterprise grade. So the slides with every compliance stick that we add every time we add scale, every time we add data residency, every time we improve the migration assistance, there is a really gratifying sense. That's just a lot of f****** hard work over a long period of time by huge engineering teams. But when you talk to the customers, they're like, "I get it. I'm not sure I believe you 4 years ago, they don't really say that clearly. But like every time I see every 6 months, you're getting better at all of the hard nilly expensive engineering that underlies building sort of a true world-class enterprise cloud. I do think for most of the data center audience that's making a huge resonate distance, right? So I don't know how we measure our commitment levels to that. But I would say that every time I meet with customers, maybe that customer who didn't say it, but it's part of that theory, that's really making a difference and again that is fundamentally a lot of hard graph from a lot of very fine engineering teams.

And then this is something that you think a lot about new. How do you lower the barriers to cloud adoption given cost of migrations, technical hurdles and the need to retrain users?

Yes. And the answer kind of evolves over time because I feel like I've spoken to many of you about this every quarter or so. When we started out the cloud migration journey, we had a set of -- a set of capabilities that we had to build in cloud just to get to parity with behind the firewall solution. And that was our first phase of migration. We built that out. And then we had a second phase where we said there's a set of capabilities that are blockers for people to come from behind the firewall to cloud because there are now additional requirements that you have on the cloud when somebody else is hosting your software that we will need to check off in order for customers to come from behind the firewall to cloud. We've been working on those blockers through the second phase and we've put out a public road map that's available to everyone on Atlassian.com/roadmap where we've been making continual progress over the last couple of years knocking out these blockers quarter-by-quarter. And right now, when I think about, okay, what else is remaining in terms of friction for customers to come from behind the firewall to our cloud products. We think of that in 3 categories. And like Mike talked about every quarter, we go out there and we say, here's the progress we are making in those 3 categories. One is around scale and like -- you've seen we've started from 1,000 users, for instance, all the way up to 50,000 users, for instance, now. And we also have our enterprise edition where we have multiple unlimited instances of users that customers can deploy. So we've gone to a great place on scale that this is not really a problem that we hear from customers anymore about, if only you had more scale, we could come out of the cloud. So we've lowered the barrier on that. A second category of work there is around data residency, data management investment. So you heard from us this morning about BYOK, which is actually a pretty big deal for regulated customers, especially data residency. We have more regions come online at a regular cadence. We are working on FedRAMP for our federal customers, which is also an important category of data management capabilities they want in the cloud. So that's a second category I would describe where we've made a lot of progress unlocked folks, and therefore, we've had large German bank migrations, large regulated customer migrations to the cloud. And the last category, I would say, is really around extensibility and ecosystem. And many of you have asked me pointed questions around specifically that there really are 2 options today are Forge and Connect. And as you saw again this morning, we continue to make progress there around fundamentally the question, especially larger customers ask us is, hey, if I use cloud marketplace apps, how can I trust that the app has the same enterprise-grade capabilities that Atlassian has. And one way to do that is to build it on forge where we host our own infrastructure, and therefore, we can make certain guarantees about the data access policies and security posters of the app. The second way is what we introduced this morning as a privacy and security certification of sorts. Right now, marketplace app vendors can self-declare their security posture and data access policies there. So enterprise customers have even more ability to say, "Okay, cool, I'm making a decision based on what I'm reading about the app itself." So we continue to make progress in the third category as well. And migration tooling experiences have gotten substantially better over the last couple of years. I'm sure you've heard a lot of customers talk about it over the last 2 days as well. So we think of this as a continuing journey where we're pleased with the progress that we have made so far, and I would characterize it as we are at a good point where we expected we would be.

Can I speak to the cost side of that. So 3 big costs. One is the license difference and then the actual cost to migrate and then the cost to retrain users. The license difference, we've gotten very good at. We have ROI models spreadsheets you name it. We can just buy the license difference largely based off of savings in cost of having admins, run boxes and the harder cost of SaaS, just the hardware, all that type of stuff like we can basically do that and largely justify it, that's before we even talked about unlocking new productivity, happier end users and so on, which we can also talk about through all the great capabilities. So that one we've gotten very good at largely comes down to timing and budgeting and then using loyalty discounts that wind down every year, we give people kind of a big compelling reason to choose by June 30. As far as the cost to migrate, absolutely right. Like if you have technical people in-house, and you want to work with our cloud migration managers on migration tooling, we're happy to run a project with you. If you want to pay one of our partners, x amount of dollars, whatever our partners go do it for you, that is an option. Either one of those is going to cost a whole lot less to migrating to an alternative vendor in every situation, migrating to our cloud will be less expensive. But that is a project, you should decide if you want to do it in-house or with our partners. We're here to support either one of those. Our tooling has gotten much better. We know that for our ability to like sell a deal and get it through is well instrumented throughout. We know if a migration is going to south, we can basically go off and run intervention very quickly. Cost to retrain. We got training modules and all of that stuff, but the reality is everyone thinks that. It is a much more modern, intuitive and better experience. So yes, when people log in, there's a little bit of who move my cheese, right? But it's a little bit different. But the reality is it's much better in like a much richer experience and most people get it really quickly. So we can see how CSAT might drop down a little bit. They figured it out, but within a few weeks or a couple of months, the CSAT goes up into the right as everyone gets familiar with the modern experience. So all things we address with every one of our customers who go through this journey.

Great. This is -- I thought was an interesting question from Michael Turits of KeyBanc. How do you see the combination of Jira Work Management and also Jira Service Management templates combining to drive Atlassian usage beyond software developers and IT across the enterprise? How do these tools complement each other?

We should spin some random....

Yes, totally, totally. It's a great question because if you look at -- I think about work, this work happens within your teams and there's work that happens between teams. And for work that happens within teams, we have things like Jira Software, we have Jira Work Management, where you allocate work amongst each other and you know the people that are involved. If you look at work that happens between teams, there's almost some sort of interface that needs to happen between teams not out there. And that interface for IT teams is often ITSM, how we ever help desk for advanced companies or companies that think about it well, you have sort of enterprise service management, where your IT and your HR and our legal when any service team internally sort of publishes their interface in the form of a service desk. So I do think those things pay real, really well with each other and that pairing works nicely. I do think that over time, there's opportunities that Atlassian have that other people don't in order to make that health experience different to just a service desk because if you talk to a legal team, they don't think, "I provide 4 services, and I should create a help desk around that." The legal team just to answer questions or they write docs or they say, get in touch with us. And because across the last thing, you have Confluence and you have the help desk, can we have the concept of teams in our product and all the stuff that we can do with large language models now, we have a huge opportunity to actually change the way that help happens across an organization. And have all got other vendors out there, if you adjust the help desk product, you can't know where the experts hang out. You don't know what docs they're creating Confluence and you don't have that expertise that you can tap on to. If you're a search vendor, you might be able to find a doc, but you -- if the answer is not already there, you can't route someone to an expert to get an answer. So there's a lot of things that we at Atlassian do across our products and with the network and graph of work that we have, that we can really solve that get help across the organization problem in, I think, unique ways that haven't been possible until now with some of the live language models that can bring it all together. So really a good question. It is amazing for customers using within their team. There's a customer here from Breville in Australia, and Breville is the great toasters and microwave ovens and apparently really good coffee machines, but I haven't used them myself. And they want to transform their organization because they wanted to move from selling products to selling solutions. And so they don't sell coffee machine. They sell the coffee machine and they said your subscription for the beans and the subscription for the things you need to clean it over time for their ovens and the micro ovens, they'll send your recipe book that then talks to the other. And so it's like, oh, I want my chicken and it actually has 17 different temperatures that cooks over the life cycle of the chicken, you'd never enter them in yourself. But if you press the recipe number, it all works. So that's sort of move from a kind of a product company selling one-off things to a long-term service company. And I'm sure the stuff happened in engineering, but this deployment of JWM happened in the marketing organization and they have 100 people in marketing and the head of marketing operations said, "Hey, we need a different way of doing things." And they had a smattering of other work management products across their organization, they wanted to standardize on 1 thing across the company. I'm sure there's some benefit of JWM talking to Jira software at some stage, somewhere in their organization. But largely, it was like the benefits of JWM that sold them on that. And again, it's a transformation, the whole company had to change the way marketing work had to change, but they want a technical company. They are building software or the closer they get this microwave oven. So I thought it was very interesting in terms of JWM, the opportunity we have across organizations. People sometimes pigeon hole that is like, well, it's just near the software teams or only the software teams that expand out and say, well, not all company needs to transform to be a software company, but the marketing team aren't thinking in terms of building software and sprints and stuff like that. They just need to get their work done. So it's really example to JWM.

From Adam Tindle of Raymond James. Atlassian together, can you talk about the decision to bundle these products, which seems like an opportunity to consolidate share onto the platform that you may not have had in some of the individual products versus threat in the sense that it appears to be deflationary on a like-for-like basis. Mike, do you want to take this on Atlassian together?

I was going to throw the deflationary part they can, first time we've been accused of being deflationary in the world. So I love it. Cam, you are first I can follow up.

Yes. I think the conversation with Sean at Reddit is a perfect example of what we're trying to solve with Atlassian together is over the last 5, 10 years, you name it, there's been a proliferation in new ways of working and new tools. Then like when we try -- and we're largely trying to embrace that. And you see actually Atlassian participating in that. We have Confluence. We have outlets. We have different ways to track, manage, communicate your work for different teams. And that's all of our work differently together. It's a part of what we believe there won't be one project management tool to rule them all. We think there's going to be many, many different capabilities out there. But we want to do 2 things. One is we want to simplify the decision that customers needed to make when they came to Atlassian and looked at our portfolio of offerings. And they said, we like these things, they work well together. We understand different teams are going to want to use them for different sets of capabilities, and we wanted to make that decision very easy for them. So instead of 5 individual decisions, we can make 1 really easy decision. And that's where Atlassian together came in. The second part of this is -- the other side of this is -- there is, especially over the last 6 to 12 months, a desire from on-high to control costs and consolidate vendors. Now we believe long term, there's going to be a whole lot more software companies than there is today. But that doesn't matter. You still need to control costs when you can and consolidate vendors when you can. And when they come to Atlassian, another thing we simplify that decision. Hey, we have 15 different project management tools, we're looking to save money, why should we standardize on Atlassian? Well, because you get 5 incredible products from Atlassian, largely bundled at a price that is less than any individual product of one of our competitors making an easy financial decision, and you still get that flexibility of decisions for your end users. You don't have to rip away all the tools that they want to go use. That said, it only makes sense for us from a pricing perspective, if we know we get significant user expansion during that conversation. And I'm happy to say that that's -- and that something we want. That was the bet we win, right? We're going to consult like offer a bundled price, we need to make it up by getting way more wall-to-wall adoption of our project management tools. And that's actually exactly what's happening as we've gone with the Atlassian together deals. We watch it like a hawk to make sure that original hypothesis worked, but that's exactly what has happened with customers. They decided to consolidate, standardize on our tools and of course, start purchasing much larger users, and we are always focusing on a wall-to-wall conversation at that time with our teams.

Next question is coming from Arjun Bhatia of William Blair. Can you talk about the Beacon product you launched today and your longer-term aspirations in the security or SecOps market?

I can take that. So Beacon, the product that we launched today is another offering from our Point A set of products. It's basically threat detection and monitoring tool. The need for Beacon was really born from the desire to protect IT organizations from suspicious user activity and possible malicious intent across all of Atlassian products. So you'll probably find it familiar from the Atlassian Access value proposition where essentially, we think of this as what can we do inside the Atlassian platform to help secure our users and admins across our entire set of products. So with Beacon essentially a natural starting point was how do we help users that are implementing DevSecOps life cycles with our products, help track some external threats that might get mitigated at very much the entry point. Having said that, the DevSecOps life cycle itself plenty of our customers implement that with our open tool chain. The first point of entry is very much partners like sneak, the developer security solution, where we very much intend to continue telling that story. We believe in the open tool chain philosophy. We want to integrate with several other security point vendors that can help construct the DevSecOps workflow for our customers and provide different forms of security. So think of Beacon really as an extension across the Atlassian platform in terms specific security scenarios that we can help build for our customers, and it's very much an addition to the overall open to chain philosophy that we will continue to pursue with multiple security partners that they can integrate our ALM life cycle change.

I want to give you an example of -- I want to give you examples of Beacon detections that only we could do that an outside vendor can't do. So I think the security market is really interesting because there's sort of you get generic broad security tools that can secure everything. And then we have customers that are asking us for very specific things that they're looking for. So for example, we have detected a number of suspicious activities, including busting some of our customers' red teams, which I thought was hilarious who were trying to attack our products in a way that only our products could do. So this is not a generic security tool, trying to secure all sorts of other things. There are things that only we know about activity within the product like what they are searching for, what they are clearing, what they are trying to do spaces that become open to anonymous access or something like this that is actually relevant to our data and our products. Now we don't know it's a security floor, the biggest thing we can do is highlight it to the right group of people who can then either put it into an automated tooling or something else. But these are very specific alerts to our product landscape that I think you're going to see more of that now. Again, the pricing model, as me mentioned, with access is quite unique because like we do with Access, we sell it effectively to the IT administrator to say, let's take that off the table as a concern set to unlock all of the Atlassian family of products for your organizations to purchase independently, right? The access original thing came from the Trello world, which was people saying, "I want to let my whole organization use Trello and I want each different group to pay for their own thing because I don't want to do this. but I need to make sure it's secure." So we take the use of security and content security off the table. IT 10 can be comfortable, so well at night that, that's covered, and then give the business the ability to purchase whichever Atlassian product and whatever quantity they want, to separate the concerns. It's a unique ability of ours economical it's done very well with access for us and for the customer. And so that's what we consider to do is a similar model.

So similar way on the new product side, how would you compare the long-term opportunity from a product like Jira product discovery to something like Jira Service Desk or other products. How do the number of addressable users compared to your other offerings?

Okay. I'm just recited by Jira product discovery, as you probably saw from the keynote this morning, if you were around then Jira product discovery solves a very clear market need that there is already defined market. It's for product managers that want to capture insights from their customers and want to use that upstream of actually development. And it's interesting because it's a defined sort of product category out there. There's a handful of smaller vendors, but it hasn't yet really been something that every single product manager uses largely, I think, because it's a bit of a niche industry, the pricing packaging hasn't worked out particularly well. And then if you look at the way we've implemented, we've actually done some innovative ways in what that product does to solve that problem. So I believe that, that product in over time will be something that can attach really well to the Jira software user base now will be the same in size. We're going to have as many product managers as developers and so forth, like I wouldn't be putting in spreadsheets as sort of a one-to-one uplift across every single seat that we have. But because of the defined market, I do think the attach rates happen faster than, say, a product like Atlas or Compass, where we're defining new market categories out there. And ultimately, those ones probably have larger addressable markets, but it takes an effort to educate the market and the market understand and so forth. So I think JPT very defined market will attach at a faster rate, some of the other products that are defining markets will take a little bit longer to ramp up.

Great. The next one is a competition question. I lost that question. Okay. Another topic that couple of...

Competitions disappeared?

Of the panel, who would win in a foot race, I would come off.

Vendor consolidation came up in an early response to a prior question, and it's been coming off a lot more often as of late. Is there a consolidation play for us in the DevOps space or something like the Together bundle that we think about in other product areas as well.

Yes, the vendor consolidation, I think as we went into COVID, it was like vendor proliferation. It was like whatever the CIO wanted to make people productive, go crazy, and that's how Reddit got for 403,000 employees, and then Sean had to go clean it up last year. Thank you, Sean. So we saw it across the board. And then all the renewals came up over the last 12 months. And every renewal gets a conversation with someone at the company of like are they using it? Are they getting the value out of it we've -- that we were promised, right? And then can we -- do we have to have it? The good part about Atlassian products broadly, and then I'm going to talk about agile DevOps in a second here. But the good part of our products is -- for the bulk of our customers who start on free licenses, then go to monthly per user licenses and then go to annual licenses, largely in that journey. They're buying more and more users as they're using more and more standardizing. So very rarely that someone come out of the gate and go, go from 0 to 5,000 users and all of a sudden it's shelfware, like we are very largely protected from that simply because of our business model and how we price and license to sell our software. So when the consolidation conversation comes, we are pretty damn confident that one, they're using our software and two, it's highly competitively priced. They're not going to save a ton of money by going to any other vendor in the market, with lots of in. Interesting in the Agile DevOps market versus the work management for all market, yes, we have a variety of software. We have a big bucket. We have Jira Align and we have now Jira product discovery. You name it and can you just bundle that all up and call it a day, and we haven't. Largely because when we looked at the DevOps market core to our strategy in the open DevOps market, it is one of flexibility and choice and as what we fundamentally believe that if you're trying to be an all-in-one vendor in DevOps, do everything from ideation -- AI ideation, like we showed at JPT down to deployment and monitoring like that is a broad set of technology that's constantly changing, and there's a new vendor every 2 days. And he's like, just go to our show flow. There's a bunch of new vendors out there just over the last 12 months. So our goal was to one, make our products awesome and easy to use together in that space, but always preserve the optionality for customers to choose what makes the most sense for their development teams. So that's why you haven't seen us go with a bundling effort there. Mike, Scott, you've been...

I was just going to add one flavor that wasn't in there. With Atlassian together, it sells wall to wall because those are all the products that every single person in your company can use. That's why the SKUs can match in terms of quantum. You've got 1,000 people in your company, buy 1,000 Trello seats, 1,000 confluence seats, 1,000 JWM seats, 1,000 Access seats, it lines up. In ADO, it's very different, right? You have a few product managers, lots of engineers, lots more seats at Jira software than you do have seats a bit bucket because 25%, 30% of people in Jira software are actual coders. So the codes not is more, but like that's -- it's a very different sort of set of user personas in that space, which makes that idea much more complicated at pricing package for a customer, then harder to explain, that creates higher fraction.

Can you help quantify the market opportunity for new products such as Atlas and Compass? Are you expanding your TAM or simply addressing the TAM? How do you think about go-to-market as you add new products? And how do you balance the sales and marketing resources with the strong number of products?

I'll do to TAM and Cameron can think about the sales and marketing on. I guess though Atlassian's history, if you're doing something innovative, you often are doing something into a TAM that doesn't exist and you steal it from other areas or the market categories evolve. And I think when you look back at Jira Software, when we started 20-plus years ago, the TAM for that was probably the number of companies that use rational software, which is about 1,000. And they probably paid a couple of hundred thousand dollars each for it, the TAM that was tiny. And because of the way we did it and the way we licensed it and what we did for companies, we totally changed the TAM for those products. And so it's not like a semiconductor industry where the number of computers sets the TAM and you're just fighting in a zero-sum game. Almost all of our products are in some ways, expanding TAM from what you would say the traditional competitors are, JSM, for example, the most number of customers don't switch from some competing solution. They come from spreadsheets. They come from using e-mail. They come from using some open source tools that they've cobbled together they didn't pay any dollars for. And so all across our entire product set to today -- what we have today and historically, we've always been done a great job of expanding TAM, whether we expand TAM on a dollars per seat basis or a seats per customer basis or a number of customers' basis like break it down, we actually expand TAM in many different ways. And so that just specific things around Atlas and Compass. Atlas broadly fits into, hey, project management, communication, teamwork, directory. We've put a whole bunch of things together that no one else has really put together before, which is that, hey, work is done by teams that doesn't fit in your org chart. Hey, if we put those 2 concepts together, it really changes the way that people contract and communicate work around their organization. I could probably ham it into some existing like Gartner, Quadrant, or Forrester Wave if I squint to do that. But at the end of the day, I'm really creating a new way for teams to come together and solve a problem which is communicate work, which isn't really a TAM there, but broadly in sort of the project management space. On Compass, I just -- for those who aren't familiar with it, if you look back 15 years ago, developers did a very few handful of things. They would wake up in the morning, they would get their ticket, mostly using Jira within right their code and committed in their IDE, they'd put it on some sort of repository like a Bit Bucket, and then they would build it, you see something like [indiscernible] and that was their day. These days, writing code and developing is way more complicated than that and developers are on call. So when something goes wrong in the middle of the night, they get woken up. We have a product for that. Once you develop stuff, you no longer just writing code using open source libraries, you're talking to third-party APIs. And you want to know if the Amazon API is up or down when something goes wrong. And your code depends on other code because you're building micro services versus monolithic things. And so when we look at these changes in how code and things get delivered, we've worked at Compass and said, hang on, tracking microservers tracking, running product is a thing that no one does particularly when it was totally greenfield out there. And if you ask customers what they use today, it's like, well, we sort of use spreadsheets. We try and use asset management from like the old way of doing it. It's usually used for tracking laptops or physical hardware. And we kind of make it work like there's a huge opportunity to track it and work in that way. And so again, that is a totally new market category that doesn't exist in at least to my knowledge in this Forrester, Gartner, sort of Quadrants and markets that they have defined out there. And I feel for both those products, I feel very excited about the possibilities to change the way that work gets done, but I can't tell you what the TAM is at the moment because we're defining it.

Just quickly to Cameron's -- sorry, Cameron's point on activity, I wanted to give a compass example and my team. So everyone has their favorite among the new things. Compass is a good example of how we build products in new spaces. We have tracked the activity and usage. As you know, we do philosophically with our customers very, very detailed to Cameron's point, about shelfware. Internally, our usage of Compass has gone from hundreds of users to thousands of users over the last 1.5 years. And we track ourselves. We have a large engineering team, as you're probably well aware. We don't force them to use our products. We try to make sure that the products deliver value to them in the building of those products because that has to work for the rest of the market for it to take off. That takes a bit more time. That takes thought, that takes rework and moving around, but we are incredibly detailed about tracking activity, especially for those products in compiling new spaces and innovative areas. And if they don't work internally, can get used by large numbers of people and sort of the adoption flows around our organization. And we're big enough now that it's beautiful because you can watch whether it's working or not working. And that's pretty core philosophy, especially for those new things that they have to work for the audiences that they're intended for in an organic way, which Compass is doing really well and started always doing really well on that front as well.

Just on go to market across the board, whether it's a new product or old product, we pride ourselves and on the stong defender of one of the, if not the most efficient go-to-market model in enterprise software. And anything new we do has to sit within that world, which means you're not going to see a Super Bowl ad for Compass next year, I'm sorry. But without -- we're not going to go burn a bunch of money to go drive awareness of a product and lose money over time. The advantage we have is we can be patient with flywheel model. We know how it works. I got 250,000 paying customers. I have hundreds of thousands more free customers, and they all talk to each other. They all go to community events and they all go watch these shows. And once they're successful, we get them in the user base, largely for free. We start them in small teams, we start transacting. We see what works and we double down, we double down and we double down.. The next thing you know we have Jira Service Management. That's how the model works with new products at Atlassian. But it's something that we know those models. We get better at it every day. We have machine learning models that actually calculate what people have historically done in our base, and then we basically can put them in automatic segmentation for the next campaign for the next of frame. So these are the whole things that we've been -- got very, very good at over the last decade to ensure that we maintain the efficient go-to-market that we have today, even as we release new products.

Great. Well, that's all the time that we have for today. Thank you so much for everyone joining us today, both in the room and virtually, and hope you can join us for a brief social hour after this.

Thank you.