Episode 227: How Atlassian Uses Behavioural Science and Data to Power Distributed Work (Interview with Annie Dean)
What does it really take to make flexible work succeed - at scale and over time?
In this episode of the Digital HR Leaders Podcast, host David Green sits down with Annie Dean, Vice President of Workplace and Future of Work Transformation at Atlassian, to explore how one of the world’s leading tech firms is boldly reimagining work for the long term.
Sharing insights from Atlassian’s five-year journey as a fully distributed company, Annie unpacks how her team is using behavioural science, asynchronous collaboration, and AI to design a more human-centric and productive way of working.
Join them as they discuss:
The biggest lessons Atlassian has learned from five years of distributed work
Why they built an internal behavioural science function, and how it drives their work design
How asynchronous collaboration is redefining productivity and employee well-being
How AI power users are saving up to 7 hours a week - and how Atlassian is enabling that shift
Proven strategies to integrate AI into distributed team workflows
Guidance for HR and people analytics leaders defending flexibility with data
A sneak peek into upcoming research from Atlassian’s Teamwork Lab
Whether you’re refining your flexible work strategy or looking to future-proof your organisation’s operating model, this episode, sponsored by Worklytics, is packed with practical ideas and forward-thinking insights.
Worklytics helps leaders understand how work actually happens with data-driven insights into collaboration, productivity, and AI adoption.
By analysing real work patterns - from meetings to tool usage - they empower teams to work smarter, not harder.
And here’s something special: Worklytics is offering Digital HR Leaders listeners a complimentary AI adoption assessment to understand how your teams are really using AI - and where untapped potential lies. But don’t wait - spots are limited.
Learn more at worklytics.co/ai
[0:00:00] David Green: If you're anything like me and follow the ongoing research and commentary around remote, in-office and hybrid working, you'll know just how polarised the conversation still is. Some companies are doubling down on flexibility, others are pulling people back into offices full-time, but very few firms have taken a long-term, data-informed and research-driven approach to rethinking how work happens within their unique organisation. Atlassian, however, is one of the companies that has. Since 2020, Atlassian has operated as a fully distributed company, and instead of reverting to old ways of working, they've leaned into new ones, built around behavioural science, asynchronous collaboration, and now AI.
I'm David Green. And today, on the Digital HR Leaders podcast, I'm joined by Annie Dean, Vice President of Workplace and Future of Work Transformation at Atlassian. Annie and her team have done and continue to do some amazing work, shaping Atlassian's approach to distributed working. From building a behavioural science function, to redefining productivity through asynchronous work, to helping teams adopt AI in ways that are both practical and people-first. And today, we are going to be looking into how she is seeing all the facets influence the way work is designed. So, grab a pen and paper, or your digital notepad, as this is a conversation packed with insights that could help spark some new ideas. With that, let's get the conversation started.
Hi Annie, welcome to the show. To kick things off, could you start by telling us a little bit about yourself, your background, your role, and Team Anywhere, perhaps, and its context at Atlassian?
[0:01:57] Annie Dean: Thank you so much for having me, David. Yeah, so I would love to kick off by talking about my sort of non-traditional role and non-traditional pathway to getting there. Today, I am the head of Team Anywhere at Atlassian, which is a software company that you may know for products like Jira, Confluence, Loom. Basically, we help teams collaborate wherever they are working from. And in my role as the head of Team Anywhere, I'm really responsible for helping the company continue to excel as a distributed company. So, in 2020, like many other companies, our founders announced an intent to become distributed first. And that means that while we do have 12 global offices, folks are not required to attend them, and yet they do. And I oversee a nontraditional org that includes our real estate and offices, so I'm the head of real estate, as well as a research team, known as the Teamwork Lab, run by behavioural scientists that study how work happens today. And I also oversee a set of talent programmes, including the Team Anywhere policy, which sets forth the idea that we are a distributed company, as well as some important talent programmes, like onboarding and what we call intentional togetherness, another key programme that we can talk about throughout our conversation.
I started my career as a corporate real estate attorney on Wall Street. So, think about representing institutional lenders in mega deals of hotel portfolios or lending to whatever Blackstone is buying, type of thing. And then, because I had a very challenging experience as a young mother early in my career as an attorney, I really was curious about how to make work better. It just seemed like it didn't make sense to me that I would be at the office at 4.00 in the morning and then miss the hours that my son was awake during the day. I felt like there could be a different and more effective way of working, that ultimately would solve some of the challenges that I was seeing play out in my workplace, like the fact that there were no women partners. And that led me to found a startup, which was a people analytics platform to help the Fortune 500 wrestle with the opportunity of the flexible workplace. And that was all pre-pandemic. So, that was this the time period from about 2016 to 2020. And the company was actually acquired in January 2020, at which point I thought, "You know what, I've done my work, I can see that this is the future, but CEOs are not fully aligned with it. I'm tired of banging my head against the wall".
I decided to go to Deloitte to their workforce transformation org within human capital. And I thought, I'll put 10% of time into this kind of flexibility concept. And of course, I joined Deloitte on February 24, 2020. And very quickly, I was a leading point person to help the Fortune 500 adopt the mandatory work-from-home orders that were unfolding at the time as the crisis of the pandemic was developing. And after about a year of working with Deloitte on setting up new models of working for large corporations, I then became the first Head of Remote at Facebook, where we sort of designed the set of talent policies that would really define that era. After about a year-and-a-half there, I came over to Atlassian, where I had this really cool opportunity to design a much larger organisation. And also, it's been very exciting to do this work at a company where the business is really oriented towards succeeding at distributed, because it is our business to help teams work on the internet. And that's what this world looks like. Distributed work is this non-controversial idea that work happens on the internet, and Atlassian really is at the forefront of that. So, being here has been an amazing laboratory to play with these ideas.
[0:06:02] David Green: Just for context, most of the people that listen to this podcast are based in HR. Are you in HR or do you work tangential and have close partnership with HR in Atlassian?
[0:06:12] Annie Dean: It's a great question. And as I've done this work, I've been in different parts of organisations, both as an advisor and as an employee. So, when I was working at Deloitte, sometimes it would be advising the HR team, sometimes it would be advising the CIO org; when I was at Facebook, I reported through the Chief People Officer; and when I joined Atlassian, I was hired by the COO and reported to the COO, and then eventually shifted into reporting through the Chief People Officer. What's interesting about my remit is that, although I do report through the Chief People Officer today, we have a really close partnership and our work often sits outside of what would traditionally be in a people remit. We have a lot of close partnership with our product teams, our sales teams, and our marketing teams. And there's a lot of support and space created to make that happen.
[0:07:10] David Green: And is there an element of what you and your team are doing which is like customer zero effectively, you're trying out products for Atlassian before you potentially roll those out to customers?
[0:07:22] Annie Dean: Exactly. In our teamwork lab, we really focus on this concept of Atlassian on Atlassian. And what we sort of did as we were developing our thesis about how modern work unfolds for everyone, is we really looked at these kind of magical pockets of Atlassian, where in particular, asynchronous work was really flourishing and in ways that felt very new. And we have an experiment sort of process by which we recruit, you know, 5,000 Atlassians have participated in our experiments, and we sort of A/B test what we have decided is kind of an ideal way of doing a workflow versus a control way. And it helps us understand and gain validation that it's worth it to drive the change towards the "better workflow", or to test new workflow ideas. And so, not only is this all routed in what we learn at Atlassian, but it's also tested at scale within Atlassian, which means that there's a lot of experimentation happening. It doesn't always mean that everything we're recommending externally is perfectly adopted within Atlassian, but that there is this really rich testing ground that we are playing with as a laboratory.
[0:08:40] David Green: Yeah, and maybe when we talk about some of the learnings, I think for listeners later, that A/B testing, we need to do more of that in HR, particularly with topics such as ways of working and everything else, I think.
[0:08:56] Annie Dean: Well, it's interesting, though, David, before you move on from that, I was having a conversation with customers not long ago, and we went around the room. It was about 20 large enterprise customers, and every single leader indicated that their organisation was going through a significant transformation, a lot of it due to AI being an opportunity to transform various business models. But every leader also said that they were struggling with how to drive experimentation and change how work happens. And yet, across the board, there was no owner for that work. So, I do think that's an opportunity for HR organisations to really establish leadership about driving the operational efficacy of working differently, which I do think is kind of the opportunity of our time with the advent of AI. AI not only gives us new opportunities as we pursue new business models, but it also gives us a true transformational opportunity in how work gets done. And I do think that's kind of where the next level of effectiveness in our corporations can come from.
[0:10:01] David Green: How does work really get done in your organisation? Worklytics helps companies measure collaboration, productivity, and AI adoption, using real data, not guesswork. With insights into meetings, tool usage, and work patterns, Worklytics helps enterprises optimise team performance to get more done, faster. They're currently offering podcast listeners a complimentary AI adoption assessment, gain insight into your organisation's AI usage, and unlock its full potential. Limited spots are available. Learn more at worklytics.co/AI.
Let's turn to Atlassian, Annie. So, you mentioned that Atlassian adopted a fully distributed workforce model back in 2020. How has that journey evolved over the past five years?
[0:11:13] Annie Dean: I think one of the things that is most interesting about Atlassian's journey is the very strong executive sponsorship we've had for this model. So, what struck me when I first arrived at Atlassian, I was so used to having conversations with people and all kinds of companies all over the market who were really struggling to grapple with core ideas about what it means to be a distributed company. And when I got to Atlassian, it was clear that Mike and Scott, our co-founders, were among the most sophisticated people thinking about this topic. And I think that sophistication has led clarity in their thinking and given us a strong platform to continue to experiment over this five-year period.
Remember also that the rubber doesn't hit the road on this stuff until offices reopen, right? So, we all kind of get this two-year period where we're trying to figure it out and experiment. But then, once you get the offices back in play, there's a lot more pressure to start to figure out how an organisation needs to be designed differently to thrive in this distributed world. And when I got to Atlassian, by the way, there had been this great cross-functional team, basically of volunteers, that decided many of the key elements of what this policy would look like. And so, I was able to arrive with really clear thinking already on the table that I just needed to operationalise and push forward in other ways, and I felt like the decisions that had been made were so spot on.
One of the things that we did to make sure that this programme would have longevity, is we really spent time getting data on what was happening at the company. How was this policy impacting things like our ability to hire? Was it having a negative impact on performance in any way? Did we have evidence that innovation was still thriving? And by taking a research-based lens to this, it gave us the clarity to say, "This is something we want to continue to lean into". And so, by having this data, it gave us the opportunity to have a very sensible conversation that enabled us to say, "Instead of relitigating this concept, let's instead focus on what's different about how shit gets done at an organisation today, that requires us to change the design of our offices?" How do we get more budget to support what we call intentional togetherness gatherings, which is the opportunity for people to come together onsite for a couple of days per quarter. And those become substantial operational questions, but they are questions that are driven by a clear sense of what needs to happen and a good set of guardrails about what value those types of investments or operational changes will make. And that just is a totally different kind of conversation than having a circular discussion about trying to bring value to offices by having people attend them without addressing what challenges the organisation is truly facing, which is one of the sort of catch-22s that I see continue to dominate our conversations about this topic today.
[0:14:30] David Green: And I guess, as you move to a distributed model, whether it's a fully-distributed model or a partially-distributed model, you're learning all the time. It's probably more complex than having either a fully remote organisation or a fully onsite organisation. The way you manage teams is different, perhaps. As you said, when people are together in the office, you need to think intentionally about how are you going to use that time effectively. And again, we're going to talk about what you've learned in those five years, some of the biggest lessons. I think you've talked about executive sponsorship, you've talked about experimentation, you've talked about good business questions, so thinking around innovation or the impact is on hiring, positive or negative. You've talked about using data to ground the conversations.
Now you're five years into it as an organisation at Atlassian, what are some of the other big lessons that you've learned about making a distributed-workforce model successful?
[0:15:35] Annie Dean: I love that question, because so much of this work, David, is about transformation and business. And so, I think some of the things that you just mentioned that I've talked about are really business transformation questions, and how to build influence and how to continue to have sponsorship for the decisions that you're making; how to make sure that experimentation drives a design of this programme as opposed to just winging it. And yet, then there's actually, how do you really do this? And what is it that we're doing? And it takes time, by the way, to figure out how to tell that story, because any story of change has to be a really simple story in order for people to get it. And people like Christine on my comms team and Liz, who leads our storytelling efforts, have been instrumental in taking the clarity that we've slowly gotten to and making it into a few set of simple messages.
So, let me tell you what I think this is that we're doing and why I think that actually it's an easier model to react to what you just said. You said, "This must be harder than doing hybrid or fully remote". I actually think this is an easier model. So, first of all, a lot of the discussion about these ways of working has been "about remote work". And that is a 'where to work' conversation, right? Remote means there's a home office and somebody is not working there. And I think we need to break that whole frame. Distributed work just means that work happens on the internet. It's non-controversial that most work transactions happen in the confines of our computer screens. And that's a really good thing, because AI is the most transformative tool we've ever encountered related to how we work, since we basically invented mass production back in the late 1880s. And AI does not know what's happening at the water cooler, it needs to know what's happening inside your computer. So, it behoves companies to make sure that information is in the digital ecosystem to power the assistance of AI. Okay, so that's one thing. Distributed work, it's work that happens on the internet.
The second thing is that the conversation has really been dominated about where we work, but it's how we work that matters. And so, when you start to work in a way that is digital-first, which means that you have fewer, better meetings, because you value the time you spend together more; you make it possible for people to retrieve the information they need to get their job done from AI on their computer instantly, instead of having to wait for a meeting; you make sure your team has a lot of clarity about what their goals are, and you make goals visible to everyone; these are the ways that we need to work differently. And evidence about how work is happening shows us that these are the real problems. 65% of knowledge workers in our State of Teams report said that they think it's more important to respond to a notification than to make progress on their goals. 98% of Fortune 500 leaders, executives, said that their teams could be just as productive in half the time if they could collaborate differently.
So, those problem spaces are not solved by working in an office. They are solved by working differently. And once you figure out how to work differently, using the digital ecosystem as the way to manage information and get work done, then offices are still important, right? They're a great place to go, they communicate the value of your company, they bring people into an incredible community, they make it really easy for people to get shit done. But they are not how work is orchestrated, they're not how teams are designed. And when you free yourself from the office as the design of work, then things start to get a lot easier and more interesting. And that's why I think it's easier. These are the challenges that we need to make to meet the moment of AI. And it just so happens that as we make those changes, where we work matters less. And by the way, that's a huge advantage to people who for the past five years have reintegrated a lot of different values, being healthy, being connected to their families. Certainly, this is the highest representation of women in the workforce that we've seen, because it gives them the opportunity to balance their caregiving needs. So, this is a positive thing for business and a positive thing for people.
[0:20:15] David Green: Yeah, so as you said, and I've heard other people who I respect in this space say, "More focus on the how, less focus on the where". And as you said, the 'how' means work is done differently from how it was done in the past. Yeah, very good.
[0:20:33] Annie Dean: Yeah. It's funny too, David, because like I mentioned, I oversee our onboarding programme. And so, we're going through an onboarding transformation right now, where we're thinking about different formats of content and different types of content and what type of story that we're telling. And one of the things that we see new hires struggle with at Atlassian is that we genuinely do work differently and it can be uncomfortable. I'll never forget that when I first joined the company three years ago now, one of the co-founders, then one of the CEOs, asked me to put together a study on this very specific thing. So, just like I would have done in my previous jobs, I sent an email to 20 different people. I laid out a specific question, goal, next step, suggested a call time where we could all come together in a stand-up. Nobody responded to me, not a single person. And it wasn't until a day later that a member of the team, who later became one of my direct reports said, "Oh, Annie, nobody saw that because we don't email at Atlassian, and we also don't have a strong meeting culture. We would set this up in a page for people to review asynchronously by a clear deadline".
Those are the reminders that when you come to Atlassian and when you come to this opportunity of working differently, you have to be sceptical of your toolkit and be willing to do things differently, and to be comfortable knowing that you're going to send an email for an urgent task for the CEO, and it's going to fall through because you haven't figured out the new way of working yet. And that actually led me to build the research team that I did, because I realised if we're going to succeed in this way, we have to study what's happening and make some hypotheses about what good looks like.
[0:22:21] David Green: And you set that up perfectly, Annie, because that was my next question. I'd love to drill down on the team of behavioural scientists. A lot of the guests that we have on are people analytics leaders or professionals, and I know a lot of our listeners are working in people analytics. And certainly, when we've researched leading people analytics teams, and when we mean leading teams, we mean those that tell us that they're consistently adding value and having an impact, one of the key roles that they invest in to develop and retain is behavioural scientists. And obviously, you have a team of these. Can you tell us a little bit more about this team, why it was established, what its mission is, maybe a couple of examples of some of the work that they've done?
[0:23:02] Annie Dean: Yes. So, the Teamwork Lab is led by a brilliant behavioural scientist, named Molly Sands. And Molly has also a background in product management, which is great. It makes her perfectly suited to navigate a complex tech organisation and figure out how to add value in our product roadmaps and the way that we tell our story in our sales and marketing channels. And so, that gave us the opportunity to add to the team. I think it's a group of about ten folks, and we do a variety of different things. On the one hand, we have our kind of experiment funnel where we're constantly testing new ways of working on Atlassians, using our products as some of the root of what we're focusing on. And we also do external research, like the State of Teams report that I just mentioned, where we're trying to understand, we built some hypotheses about what we are experiencing internally, and then this gives us an opportunity to go and get the macro story out in the world to say, "Is everyone struggling with these things in the same way that we expect they are?" And then, we do discrete research or analytics projects, where we're looking at a specific issue and we want to understand how location impacts performance, or we do a lot of work with the team to actually understand the efficacy of our offices, which has been incredibly valuable to us because we really take the hypothesis or the ethos that our offices are a product and require the same level of research as any other product would.
What we're really doing in the Teamwork Lab is we're understanding what modern teamwork looks like, and we're trying to build consensus over what the best behaviours really are. And that's a challenge. I think all companies are kind of hesitant to say, "Here's the best way of doing things". But there are some really clear things that we've learned that I think are applicable to all companies and teams, regardless of which tool they use to implement them or what size they are or what they're trying to achieve.
[0:25:18] David Green: I want to take a short break from this episode to introduce the Insight222 People Analytics Programme, designed for senior leaders to connect, grow, and lead in the evolving world of people analytics. The programme brings together top HR professionals with extensive experience from global companies, offering a unique platform to expand your influence, gain invaluable industry insight and tackle real-world business challenges. As a member, you'll gain access to over 40 in-person and virtual events a year, advisory sessions with seasoned practitioners, as well as insights, ideas and learning to stay up-to-date with best practices and new thinking. Every connection made brings new possibilities to elevate your impact and drive meaningful change. To learn more, head over to insight222.com/programme and join our group of global leaders.
How closely or not does your team work with your people analytics colleagues at Atlassian?
[0:26:35] Annie Dean: Quite closely. We have a new People Analytics Leader, who I know knows many of the folks in your world, David, Emma Crockett, who formerly worked at Meta, and I was a colleague there as well. And she is rebuilding the people analytics team having only joined about six months ago. And it's been such an incredible partnership to really build distributed norms into our stable org analytics. So, one challenge that I see people analytics teams have when I speak to folks out in the field, is that we haven't integrated the norms of distributed work into the way that we report our people data, using for instance, this is a small example, but when we are looking at office attendance, we want to understand where people are coming from. Because if you're coming from two hours away, it's very different than if you're coming from 30 minutes away. I want to understand, are the whole group of people who live within 30 minutes coming in frequently, because that would indicate it's easy for them and I'm setting up a great service for them, so they're going to keep using it? Whereas, if people are coming from outside a commutable distance, I know that they're visiting from out of town, so I need to have a different level of service provided to them. And location and office attendance have never been a thing that was really a concern in the past.
[0:28:03] David Green: I'd love to talk a little bit more about asynchronous work. I managed to say it properly, that's good. Could you explain what asynchronous work means to Atlassian and how it supports productivity and work-life balance?
[0:28:14] Annie Dean: Yes. So, first of all, again, these words are so hard. Remote work, asynchronous work, we're saying we can't even pronounce them. It's like we're doing this synchronously and this asynchronously, we can barely get out all of the S's. But asynchronous work basically means that it's work that's not happening when you're together. So, right now we're together, we're in a synchronous environment. And our premium on synchronous work should be higher than ever. We want to do the best, most valuable things that we can when we're spending time together, because now we have technology to manage a lot of the other things, which gives us the opportunity to do tasks a bit more independently or asynchronously.
So, I guess our biggest hypothesis is that the most important thing that companies can do is spend the time they spend together in a valuable way, and that means cutting out a lot of the noise from the way we spend that time today and replacing it with technology. And to us, that comes down to a few clear, new practices. The first is that teams need to be really clear on what matters, they need really good goals, and that helps them direct their time, which is their biggest asset. So, one, being super clear on goals is an incredibly important part of this formula. The second is that teams need fewer, better meetings, and they need to use the meetings that they spend for a different purpose. So, when I say fewer meetings I mean, instead of finding a meeting to get up to speed on something, you ask your LLM or your AI agent to search the corporate documentation and surface for you the context that you're looking for. That's instant, instead of having to wait three days for a meeting to get that information. It also means to use different formats.
So, we have an incredible tool called Loom, which is like a video message that's incredibly intuitive on a response basis. You can have a great asynchronous conversation with it, you can use emojis, it titles the video for you, it creates a summary, it's an incredibly intuitive tool. And if you're looking to improve just one thing about asynchronous work, go try making a Loom. But that can replace meetings, because I can just have a thought bubble that I want to go out to my leadership team and I can record it on Loom, and all six of us on the team can participate in that conversation in a non-urgent but highly productive way. And then, also try things like voice notes and just these asynchronous methods where you don't need to call a meeting to have it.
Then you think about what it means to have a better meeting. So, any meeting that you're having to report status updates is something that you can assign to your AI. Don't have meetings unless you're doing work together or driving momentum or solving a big problem. And so, I would say I actually spent a lot of time in meetings every day, but they're all working meetings. They're not talking about getting context on something or assigning the project plan. I do that all in Jira, I manage that in our goals platform. I look at project status updates in our goals platform, which provides a tweet-sized update in every single project in the company. And I can see every project in the company that I have permission to see from a permission standpoint. So, the internet does that for us, right? But when I'm in a meeting with my team or a cross-functional partner, or whomever it might be, we're leaving that meeting solving a problem and we're spending our time together creating something, as opposed to just, I guess, reflecting on something.
One way that we drive effectiveness in those meetings, because one important use of meetings can also be to make decisions, although I would say that we probably make a higher degree of asynchronous decisions at Atlassian because our documentation is so good, we have something called page-led meetings. So, we have a tool called Confluence, which is a documentation tool. It's incredibly intuitive and you can put any kind of media into it and it's very visually easy to navigate. And people create sort of three- to four-minute read time pages that are extremely effective at communicating what the reader needs to know. And when you join a meeting, you spend that three to four minutes reading the page. Now everyone's up to speed. Anyone who isn't in the meeting who might want this information can get up to speed in three to four minutes. And you have the balance of the time to talk and figure things out, as opposed to verbally presenting information, which would take 17 minutes to present that three- to four-minute read time. So, it's just a way better use of time, and it forces the person who's leading the meeting to make it very, very valuable.
That leads me to the third thing, which is you've got to communicate differently and you need to be able to use writing extremely effectively. Smart brevity is my favourite tool for this. This is a framework that was developed by Jim VandeHei, who is a founder of Axios. And it's just about how to create, at a super-scannable, easy-to-read message that just says, "TLDR, here's what's important, here's why it matters, and here's a few more details". This is incredibly effective at sharing information around your organisation. I always talk about this example where somebody was trying to communicate to me an extremely detailed regulatory issue related to real estate that caused real risk, so I needed to send it up to the CEO. I originally got a lengthy, lengthy email. And the process of me converting this email into this set of bullets that could be instantly and intuitively understood by a CEO took me 48 minutes. And so, by having my team start with the format of smart brevity, it might take me eight minutes, it might take me one or two back or forths to make sure that the argument or the information or context is fully there.
But it's not the exercise of taking raw information, finding the argument and delivering it up the chain. One of the things that I like to say on my team is, when we communicate, we don't pass on leadership debt. We figure out what needs to be communicated and we deliver that communication. And that's very different than just kind of writing what you think, right? So, that's an essential async tool.
Then, the last major skill I would say is that information has to be available within the digital ecosystem so AI can access it. That means every goal, every project, all of your strategy documentation, your deliverables, what your org looks like, and all of that information needs to be instantly retrievable by anyone on the team that should have access to it. And AI helps that happen. So, that makes an organisation move way, way faster because today, people play telephone instead of using technology. They get into a meeting and it's incredibly ineffective. So, the future is information online, searchable by AI.
[0:35:23] David Green: You've recently released some research on AI collaboration. One point that really stood out was how AI power users can typically save up to seven hours a week by effectively integrating AI into their workflows. Now, if you work a 35-hour week, which most of us don't, even if you work a 42-hour week, that's still nearly 20%. That's a big saving. How is Atlassian fostering this kind of advanced AI adoption within its own workforce?
[0:35:50] Annie Dean: So interesting how fast this is moving, because when we took on AI and worked towards releasing Rovo, which we released for general availability back in October and announced the previous April, and Rovo is a set of agents that we can create for any purpose. It's also an enterprise-level AI-driven search that can look across all of our enterprise service areas. And there are many other applications as well. And when we were first talking about it, not being on the AI team myself, I'm sure that they may have had a different view, but I was like, "This is going to take so long to figure out. This still feels pretty conceptual". And then it got released. And then, I started using the ChatGPT app, and I was like, "Anyone who is not using this is already behind". And I'm learning at such a rate that I can't believe how much faster I'm moving. So, and in part, Molly had told me and Ben Ostrowski, who is leading some of this AI research, Molly Sands being Head of our Teamwork Lab, that one of the big insights was that there's a mindset shift that's required where you need to think of AI as a co-creator as opposed to a delegee or delegate. And I didn't understand that at the time, but what it's meant for me is that, I think that we need to start crossing the chasm on AI using consumer use cases.
So, once I downloaded the ChatGPT app, which I use only for personal reasons, and I don't use for corporate reasons, I just have an endless set of conversations with it. Like today, I was going to talk to ChatGPT, and I pulled up my text app, because in my mind, it's now graduated to a friend that I'm texting, as opposed to a tool that I'm using. And I do everything from, I have celiac disease and I've been kind of figuring out how to heal from that, and so one of the things that I've been doing is I ask ChatGPT every single day if what I'm planning to eat for lunch or dinner or breakfast is a good thing, and it will give me all of these tweaks and recommendations and new recipes. Or I will ask ChatGPT what to do with my kids on a Saturday. Or I will think about a challenging personal conversation that I need to have and ask it to make me an outline. I know there's plenty of nerds in this viewership. So, for anyone interested in a great book, I just read Notes on Complexity by Neil Theise, over the holiday break, incredible, incredible synthesis of the three pillars of science and complexity theory being a way that the universe self-organises in all different kind of formulations and sizes.
The final part of that book talks about different spiritual realities and how these things intersect with the limitations of empirical science. ChatGPT and I can have so many conversations about this stuff. It is just so fun. So, again, these consumer applications help you realise that it's not just about writing an email better or doing something you don't want to do, it's about getting incredibly value-additive information. I can't even tell you how informative it's been about my health. I just got off a call with my nutritionist and she was like, "Oh, you should try this thing and this thing". I was like, "Yeah, ChatGPT already told me to do that". So, I don't want to say it replaces her expertise, because it doesn't, but it just helps get me so much more informed.
Then we think about what you can do in the corporate setting. And in the corporate setting, where obviously there's a lot more guardrails and privacy, and you would only ever use your enterprise setting for any work information that you're trying to deal with, but you can just move so much faster. You don't need to ever ask somebody to send you a document again. You can just ask in plain English, "Hey, can you find that document that Miriam and I were working on last week?" and up it will pop with a summary. One of the great tools that I've used it for is I will kind of say, "Hey, I have this thing that I'm trying to do with this team, and I think here is the outcome, and here are the different places, surface areas that we're playing in. Can you turn this into a goal?" It would have taken me three hours to come up with a goal as succinct as AI can give us. So, play with it as a consumer and see how fast it makes you move as a leader, because I talk to people who are unwilling, for whatever reason, to engage with AI, and I'm just like, "It's so easy, and it's so transformative". Now is the time to experiment.
[0:40:44] David Green: Yeah. As you said, just experiment, just try, just see how you get on.
[0:40:49] Annie Dean: Right. And before you know it, you'll be thinking it's a friend in your text inbox.
[0:40:54] David Green: Replace my dog, but never replace my dog! And what kind of practices have been most transformative for helping teams seamlessly integrate AI into their workflows, especially maybe from a distributed-work perspective?
[0:41:12] Annie Dean: Well, I have to give all credit to our technical teams and our engineering teams, who have been very thoughtful about integrating AI opportunities into especially our internal tools. So, we have this incredible tool that manages our performance management process, called ELEVATE. And even this, we'd go through that process every six months, and new this six months in this half has been new AI features that give us the opportunity to really use AI in a meaningful way, while also being thoughtful and ethical about how performance reviews and AI come together. And so, again, I think that type of exposure and helping people learn from it has been a really embedded part of what we're doing. Also, Atlassian is an incredibly business-focused culture. And so, again, all credit to our executive leadership team and to our founders and CEO, who are constantly talking about AI and are just genuinely excited about AI. And that excitement and priority rubs off on everyone.
I would say the next generation of practices is the development of agents. So, it's incredibly easy to develop an agent through Rovo, our AI tool. You literally just open up a screen and say, "I'm building an agent on new hire onboarding. And here is the folder of information that I want you to be an expert on. And make sure, on a scale of ten, make sure that your niceness is a seven, professional, but not overwhelmingly friendly, and set those parameters", and instantly you have an agent that can do what you're looking for it to do. And I think that is going to be profoundly transformative, especially as we think about HR organisations, where a lot of this one-to-one help is really needed and can be, at least initially, facilitated through an agent.
[0:43:10] David Green: So, we're coming to, got a couple of questions left, Annie. Now, I'm sure anyone listening to this is convinced, okay, distributed work, moving everything digital. But a lot of times, sometimes the people listening have to convince senior leaders within their organisation. I'm not going to get into all the politics of stuff, but it's almost like the 'where we work' question has become heavily politicised, certainly in the US, but not just in the US, to be fair at the moment. We've seen high-profile examples of CEOs and companies bringing back employees five days a week, again, focusing on the 'where' and not on the 'how' again. What guidance would you offer to HR, people analytics leaders, maybe teams a little bit like the Teamwork Lab at Atlassian, if there are any out there, on how they can evaluate and, depending on the data, defend flexible working, distributed work?
[0:44:04] Annie Dean: I think when we think about distributed work or the opportunity to have some choice in where a person works from, there is starting to be a fairly good consensus from the industry on industry-wide research. So, I'm thinking specifically of Nick Bloom, and Raj Choudhury over at Harvard. And I think the success buckets that you want to investigate are, are we hiring people faster? Are we retaining them longer? Do we have less churn? Is the quality of hire at least as good or better? I think one thing that's not discussed as frequently, but I think is relevant, is can you attain that same talent at a cheaper rate? You know, certainly hiring for instance in San Diego versus San Francisco has the same collaboration impact, but can be a different compensation zone, and I think that's a nice incentive for business. And then, so I think you want to look across those metrics. And again, the industry research is showing that having the opportunity to choose where you work just a couple of days a week, or just sometimes, drastically reduces quit rates. And by giving people heavy-handed in-office mandates, the best talent leaves. At Atlassian, I can say that the number one reason why top performers say they want to stay at Atlassian is because of Team Anywhere, and that's a great thing for companies. I mean, I think it's non-controversial that having great people is what makes a company successful.
Then, I think you want to look at what performance looks like. So, I would say that the number one thing hampering performance for companies right now is the ability to work differently and use the internet to its advantage and to harness AI. So, anybody who is considering a return-to-office mandate, and in this current round I would say that takes a lot of time and energy, we don't really have a lot of evidence that that's going to solve our problems. How close are we to adopting AI? Because AI is going to drive collaboration, save time, increase quality, give us greater innovation. Let's start focusing on how we work versus where we work. And what happens over time is, once you start figuring out how to work differently, it just doesn't matter where work happens as much, the office will still be an important part of the overall picture, people want to come together. And
I think that's a big misnomer here, that the idea of distributed work is that we're going to become part of this corporate video game with NPC co-workers, and everybody wearing a VR headset all the time. That's not the case. People want the office, they just want the office to deliver value for them and they've learned something, which is that they value the opportunity to eat a healthy meal at home and be home when their kids get home, and go for a run in their neighbourhood in the morning instead of putting on business formal wear. And doing that sometimes, that's improving how they work, not taking anything away. And taking that really valuable thing from people without improving any business outcome, I think is a real struggle, people really struggle to digest that, and it loses trust with your organisation and causes them to leave. So, I think it's just about being sensible, focusing on the outcomes and really driving the conversation toward it's how we work, not where we work that matters.
[0:47:44] David Green: Yeah, and I think one of the key things you talked about there, early on in that answer, was talking the language of the business. If it's senior executives, quantify. Talk about it makes hiring a lot easier. It's actually cheaper to hire the same talent. We've got data around quit rates and stuff like that, so we can do it, but it's going to cost us a lot more money by doing it, and then people might think again. And I think, yeah, right to obviously highlight Nick Bloom, Raj Choudhury, Brian Elliott, I think, along with yourself, Lynda Gratton, others like that are really producing some great research about this way of working, which I think helps HR professionals that are having those conversations with senior leaders.
Annie, I'm conscious of time. We've got a couple of questions left. One is about upcoming research that you're doing. The next one is around the question of the series that we're asking everyone. So, looking to the future, is there any upcoming research that you're particularly excited about within the Teamwork Lab?
[0:48:49] Annie Dean: We're really focused in this next few months on the concept of information. So, the way people are getting information today tends to be meetings. Yes, there's one-off information that's available in different places; and, yes, information silos are a thing that's been around since the 1990s, right? But AI is completely transforming how we access information, and that's going to cause us to really deliberately change our behaviour to use our time better. So, that's something I'm really immersed in and interested in right now.
[0:49:21] David Green: Great, well, we look forward to hearing more about that in future months. Annie, this is a question we're asking each guest on this series, and appreciate that you're in HR now, we've worked in and out of HR moving forward, but I think it's a topic that you're an expert on, the team-effectiveness part. How can HR help the organisation understand and improve team effectiveness?
[0:49:45] Annie Dean: Great question. I think that first of all, you can't be a credible advisor on improving the effectiveness of a team, which causes friction to drive any kind of change, unless you deeply understand the business. So, I spend a lot of time staying up on what we're saying internally, where we're going, what external strategists say about where we're going, just really sitting in the middle of what each of our executive team leaders are trying to achieve, and where the challenges might be in that or what blockers they might face. So, I make sure that if I'm recommending something, it serves them. And then, the next thing is, how do you actually impact team effectiveness? And it's the centre of what we've talked about today. How we work matters, it's the opportunity of our time. And so, let's start owning that conversation and help driving change, because we're not going to be able to solve the big problems that we're facing as companies, as teams, or in the greater world, unless we can think about solving those problems differently, and I think it's time for that.
[0:50:59] David Green: Yeah, very good, and I think you're right. I mean, be a good business leader first, be knowledgeable about the business. If you're in HR, you're a businessperson who happens to work in HR. It's not HR in the business, you are a businessperson.
[0:51:15] Annie Dean: Exactly. I think a lot about if a CFO is allocating capital, an HR leader is allocating time. And time and capital are the only two things we've got. So, let's make sure that we're investing our time wisely through smart org design, through people that have the right skills and through ways of working that makes sense.
[0:51:34] David Green: Perfect ending. Annie, I've really enjoyed this conversation. I've learned a lot. I think you're doing some fantastic work at Atlassian with Team Anywhere, and obviously publishing some of that externally as well. How can people stay in touch with you, learn more about Team Anywhere, learn about the work you're doing at Atlassian, and learn more about Atlassian as well?
[0:51:55] Annie Dean: Please come find me on LinkedIn. I post all of our research and commentary about what's evolving out in the industry. And you can check out our Teamwork Lab website, which we can make sure that folks get, David, where we post all of our different research reports.
[0:52:16] David Green: Annie, thank you very much. Look forward to hopefully seeing you again in person. I think the first time we met was at the New York HR Analytics Meetup last September.
[0:52:26] Annie Dean: That's right.
[0:52:27] David Green: Hope to see you again there soon.
[0:52:28] Annie Dean: Bye David, thank you.