From startup to enterprise, Atlassian provides powerful tools to help every team unleash their full potential. In this episode, Atlassian COO Anu Bharadwaj chats with show host Avanish Sahai about the incredible growth of the Atlassian platform, all while staying true to the people the products are designed to support.
In this episode, Anu Bharadwaj shares her tricks of the trade as one of the most impressive product and platform minds we know. She draws on her time at Microsoft, where she launched several products for Microsoft Studio, as well as her current role leading operations for Atlassian, to provide some really fascinating insights.
Guest: Anu Bharadwaj
Anu Bharadwaj is the Chief Operating Officer at Atlassian. Anu leads transformative projects across product lines at Atlassian, including the enterprise business, cloud platform teams and operations.
An accomplished executive with a track record of growing $1B+ businesses, building great teams and shipping blockbuster products, Anu joined Atlassian in 2014 as the Head of Product for Jira, and has held a variety of roles at the company during her tenure. Prior to joining Atlassian, Anu was at Microsoft, launching several products for Microsoft Visual Studio for over 10 years. Anu is passionate about making the world a better place through technology and effective philanthropy. She is a member of OutSystem’s board of directors and serves as the Director of Atlassian Foundation.
Host: Avanish Sahai
Avanish Sahai is a Tidemark Fellow and has served as a Board Member of Hubspot since April 2018. Previously, Avanish served as the vice president, ISV and Apps partner ecosystem of Google from 2019 until 2021. From 2016 to 2019, he served as the global vice president, ISV and Technology alliances at ServiceNow. From 2014 to 2015, he was the senior vice president and chief product officer at Demandbase. Prior to Demandbase, Avanish built and led the Appexchange platform ecosystem team at Salesforce, and was an executive at Oracle and McKinsey & Company, as well as various early-to-mid stage startups in Silicon Valley.
About Tidemark
Tidemark is a venture capital firm, foundation, and community built to serve category-leading technology companies as they scale. Tidemark was founded in 2021 by David Yuan, who has been investing, advising, and building technology companies for over 20 years. Learn more at www.tidemarkcap.com.
Links
You can find the the full transcript here.
Avanish: Welcome, everybody. Today I’m really happy to have another fantastic guest: Anu Bharadwaj, Chief Operating Officer at Atlassian. Anu, welcome. Thank you for making the time. It’s really good to see you.
Anu: Thank you, Avanish. Wonderful to see you again. Thanks so much for having me over.
Avanish: Let’s just dive in! Atlassian has become, in some ways, a household name, but some people still don’t know a lot about it. Let’s start with the company’s history: how it started, where it originated, and so on.
Anu: Sure, I love talking about Atlassian’s story. Atlassian is an Australian company; we started in Sydney back in 2003. The founding story is fascinating: a couple of Australian guys, Mike and Scott – both really strong product people – decided, “We’re not going to work for an existing company, because we see people around us wearing these suits and ties and not having a good time.” They wanted to create a place where everyone would love to work, and more importantly, where people would have fun.
Originally, they tried dabbling in a couple of different services, and they also built a bug tracking tool for themselves, to manage their projects. Then they figured, “Why don’t we just sell this? It seems like it’s really useful for a lot of the clients we talk to.” That’s how Jira, our first product, was born. Even now, I think Jira is a more well known brand than Atlassian!
Jira was really our flagship product, and from there, the company has gone through a number of different milestones and created a platform with multiple product lines behind it. I think one of the key strategic decisions in the history of the company was adding a second big product soon enough after Jira became really successful – that was Confluence, the internal knowledge-sharing product. We’ve also made a few acquisitions, and now we are one of the biggest builders of digital collaboration tools, with over 200,000 customers worldwide. We’ve had a very successful run as a public company so far. It’s been a fun ride for those of us that have been around for a while.
Avanish: It’s amazing, truly. A lot of companies aspire to have their product become a verb – and Jira certainly has become a verb, right?
Talk to me about your journey. You’ve been with the company for a little while; you just took on this broader role. Give a bit of that background.
Anu: I had started my career at Microsoft as a developer. Microsoft hired me straight out of school and my first job was building video games. I couldn’t believe my luck! I was like, “You’re going to pay me for this? I would do this for free!” I did that for a year or so.
For the next few years at Microsoft, I got to try out building different kinds of products and software, and eventually found my way to product management. What moves me is solving customer problems and seeing the impact of my work on people, so I naturally gravitated towards that. I built a few products and businesses for Microsoft in the visual studio division, primarily building for developers: tools like Visual Studio Team Foundation Server, stuff like that.
Fast forwarding to 2014: When I arrived in Atlassian, I have to confess I did not know very much about the product and company, but they had convinced me to come out to Sydney! (Sydney is now my favorite place in the world to live in.)
You know, a lot of companies talk about values, company values and “living the values.” With Atlassian, I thought at first that was just some marketing bullshit. Every company says this! But when I met the team, I was taken by how much they really lived the values. The two founders set that down from the top. It wasn’t about putting the newest, coolest technology on their résumé and moving on – people were (and are) there to really solve customer problems! I felt like these are the kind of people I want to work with; I really felt like I belonged with them. They felt like my tribe.
I joined Atlassian as the head of product for Jira. The first two years were really just going around the world convincing people to move to Australia to construct a product management team, because we had very few people at that time – about 400 people. I constructed the Jira PM team over two years, and led the platformization of Jira. We went from being a single product to a platform that supported three product lines: Jira for software, Jira for IT, and Jira for business. I always thought of Jira as this bug-tracking, project management tool used by software developers, but in customer interviews, over 50% of our users came from non-technical teams – people who were not involved in writing code. That was really amazing for me. It was an entirely new set of use cases that we could unlock because we have a generic horizontal capability that’s so great across so many different use cases.
After the first couple of years with Atlassian, I decided to take a year off. I had worked for 12 years in tech, and I thought I wanted to go do something completely different. I took a sabbatical and went and worked with lions in Africa, and chased around cheetahs in Namibia, putting GPS collars on them as part of an anti-poaching activity. I went to the Antarctic and worked with penguins, setting up rookeries to track their population. I’m a bit of an animal geek, so I did all the wildlife conservation projects that I’d wanted to do field work on, that I’d only been able to support financially before. That was a lot of fun.
I did that for a year, but then I’d blown all my money and needed a real job. [Laughs] I came back to Atlassian, and one of the first things I did was work as the head of strategy. A key project at that time was how we would shift our entire business from being on-premises to being a SaaS company. That was a huge, pivotal moment for us. It involved redoing our technical platform, retooling our GTM engine and how we sell – reeducating our channel on how they sell, resell, and offer services – and also explaining the financial impact to the street. As a public company that is undergoing a transformation, it’s never easy to explain the full picture and articulate the small details of how we go through that transformation. I’ve led the cloud transition of Atlassian; we’ve gone from a company where over 90% of our earnings were on-premises to a company where over half of our earnings come from the cloud. Over 90% of our customers now choose our cloud products, rather than on-premises. It’s been a fantastic journey.
Avanish: That’s awesome. I didn’t know about the year in Africa and Antarctica!
Let’s dig into the platform strategy. You came in as product manager/owner for Jira, and you talked about a platformization of that across different use cases. What was the genesis of that idea? Was it recognizing a certain pattern? How did it come about, from a strategy perspective?
Anu: I have vivid memories from that time. I recall very distinctly that in my first year at Atlassian, I made a presentation to my Jira team showing Ford Model-T pictures, and how Ford thought about platformization!
A few things really led to the confluence of this decision. One, our customers really pulled us in that direction, like I said. We had launched this product thinking it’s useful for workflow management in the context of software building, because that was what we were using it for. But well over half of our customers were using it in completely different contexts: hiring, tracking recruiting processes, help desk, customer service tracking. These were use cases we hadn’t quite envisioned the product to service, but customers were already doing it. That was a clear signal that the product was capable of solving multiple problems that we hadn’t really built it specifically to do.
Avanish: Auto-productized, right.
Anu: The second thing is that Jira is an enormously successful business. In fact, I was shocked at how successful Jira was early on. We kind of pioneered product-led growth in the industry; we were generating revenue with very low operating cost because the product was selling itself. But the thing with a successful product is, what are you going to do to take it further on the path to success? For us, a platform was an easy choice. If you look at technology history, a lot of successful products start off as single products and are able to then hit the next inflection point of growth by converting themselves into a platform, and spawning off multiple other products.
The third and final thing that really brought this decision into focus for us was our aspiration to be a multi-product company. We are very much focused on teamwork. That’s our mission: unleashing the potential of teams worldwide. And teams require multiple products and services! Our philosophy is that there is never going to be a single vendor or product that works for every single team. We were sure that we would exist in a multi-product, multi-service universe. The strategy that really helps you leapfrog in a world like that is having a common platform and set of shared services that allow you to service multiple verticals together. It was an amalgamation of these three reasons why we chose platform.
Avanish: Fantastic. Let’s talk a bit about the aspects of being a platform. You mentioned that one is the different internal products that are going to sit on top of that common platform. You also mentioned customers extending it and building new use cases for functional areas that were not part of your product strategy, so that’s a second: customers doing custom work. And then there are third parties building with or on top of your platform.
How do you think about those groupings in terms of prioritization? What was the team’s perspective on internal Atlassian use cases, versus the customers’ custom development, versus the third-party ISVs building around it?
Anu: That’s a great question. In terms of a platform that’s extensible, two things were salient for us in the journey for Atlassian. One was that there were already customers, as well as independent software vendors, that were building robust businesses with extension points on top of Jira. This wasn’t necessarily something that we had thought was going to be a substantial business, but it grew organically because of the stickiness of the product. Existing Jira APIs were both customers as well as second-party vendors that were building extension plug-ins like time sharing, or software testing using Jira.
Second, one of the technical reasons that caught fire was that we built with the model that we would want any internal Atlassian team or external developer that wants to extend to have the same experience. We really thought about API-ifying each of our components from first principles, from day one. With any new thing that we built into Jira, we would automatically ask whether it was easy to extend. That really set up the product in such a way that a lot of power was exposed through extensibility. Just like how use cases turned out to be a lot broader than we thought, the ecosystem was exponentially broader than we thought, because our extensibility piece got consumed in various ways.
Avanish: Awesome. Let’s talk a bit about the third pillar, then: the ISVs. You have a very active marketplace. The Atlassian Marketplace is probably one of the most successful marketplace examples out there from a growth, consumption, and adoption perspective.
How did that come about? How much of that was because of forethought, and how much was because it started happening organically and you realized there was something powerful there? What’s the story behind the marketplace?
Anu: Like all successful company stories, I think it’s a bit of both. It’s a little bit of strategic insight at the time, and also a good deal of the community making us successful.
Our marketplace currently has both on-premises and cloud apps, and apps that work across those platforms as well. It also has applications for each of our different products – there are Jira apps, there are Confluence apps, there are Bitbucket apps. When we platform-ized Jira, there was already a functioning, well-used marketplace with our customers, so most of them installed an app from our marketplace. The vast majority installed more than one. We’ve crossed about $2 billion in lifetime sales on our marketplace so far. We have about 5,000 apps right now, from well over 1,200 partners.
Avanish: Amazing! Amazing.
Anu: To answer your question specifically: when we started out, we were very clear that we would not be able to address the entirety of use cases that our customers required. We wanted our customers to be able to fulfill those through custom extensions themselves, and for ISVs to be able to build independent businesses on top of that. We saw early on that there were ISVs building businesses on top of the platform – they would come back and excitedly tell us about the next set of extensions they needed because they could envision a really big business built on top of them. That was a great virtuous loop. The ISVs acted like our extender team.
We also had teams working very closely to make sure that our extensibility points and distribution network scaled, because one of the unique things about our marketplace is our distribution network. Atlassian is unusual in that we have 200,000 customers. For an enterprise software company, that’s a really broad base of customers. And we self-serve – our products are bought, not sold. Therefore, the extensibility plug-ins, the marketplace apps, are also bought, not sold. For ISVs to have that broad of a customer base, we really had to support them with self-serve. The underlying billing engine – where customers could come and buy a Zephyr plug-in, or a Temple plug-in, without ever having to contact either Atlassian or the ISV themselves – turned out to be a key foundational block for our marketplace.
Avanish: I think that level of “buy, not sell,” is amazing. The fact that you extend that to your ecosystem adds another dimension of uniqueness, and it probably also adds some complexity that you have to hide within the workflows and the processes to make it work.
Does that model require some self-selection among ISVs? Some ISVs are built around the sales model, the POCs, the complex-solution-selling approach. Do folks who are coming to the Atlassian ecosystem and marketplace already have that “buy, not sell” mindset, and therefore fit into that business model?
Anu: You know, that’s fascinating, and I used to have a thesis similar to yours when we were scaling up our marketplace. It’s easier to fit into our marketplace if you’re a product-led growth company, because it naturally matches the funnel that you’re used to. However, we saw a lot of companies who were used to the traditional top-down enterprise selling model adopt this as a secondary channel – which then quickly became their primary channel! There were plenty of companies that were able to make that switch because of the broad distribution that we had. We’re talking about tens of millions of monthly active users – not total users – and 200,000 companies (when we say customers, we mean companies). The overall denominator, the distribution base, was really attractive for companies to come and try out something new, and it was broad enough for people to build a substantial business on top of it.
There’s also this interesting nuance of third parties. You were asking which segments we think about in our ecosystem. We think about first party, second party, and third party. When we say first party, we mean Jira, Confluence, Trello…
Avanish: Your products, yep.
Anu: Add-on products, yes.
When we say second party, we mean people who extend these first-party products. So time tracking built on top of Jira, smart tables built on top of Confluence, or potentially a vertical Jira for legal teams.
Third parties we think of as an interesting newer use case over the past few years. Third-party companies – partners – like Slack or Snyk, who builds security software, typically come with an enterprise top-down selling model but we use our in-product distribution channel to distribute them. Inside of a Jira ticket, we have a place where you can try out Snyk, or something else, right inside the workflow. It offers an adjacent channel for a lot of these traditional, enterprise-bound companies to expand the top of their funnel. That has really helped us evolve from PLG companies to traditional enterprise companies trying to attach a secondary funnel.
Avanish: Phenomenal. I mean, fascinating, truly. I think you’ve implied some of the key success factors: there was internal alignment, there was understanding of the use cases and the thought process that dictated API-ifying and platformization. I would assume a lot of it was also alignment across products and the core marketing function – how you brand the company.
How about the other side of it? What were the challenges? As you went through that journey, what were some of the roadblocks? I’m wondering what that was like.
Anu: Yeah, we’re operators. We know this is not as rosy as it sounds. [Laughter]
I’m going to divide my answer into two parts, since we’ve talked about both platform and ecosystem. The platform shift, in my mind, is a little different from the ecosystem shift. The platform shift involves building a set of shared services, on top of which you can quickly build verticals and stamp out first-party products because you have a strong foundation underneath. That was a hard shift for us. There’s always tension between product teams and platform teams about how much to centralize and put in the common foundation, versus how much to build inside the product itself. It’s a good thing to have that tension, but boy, it can be really hard to execute with it. Product teams are always thinking, “Platform’s slowing us down because they want to build all these things centrally”; and the platform team’s thinking, “Product is trying to build their own thing! It doesn’t make any sense for us to have four notification platforms, four engagement platforms, and three services that do the same thing in different ways.” There’s a natural tension to balance there, and it took us a couple of years. We swung the pendulum too far into platform, then too far into product, and eventually learned to balance in the middle.
In terms of the ecosystem shift, it’s another layer of complexity on top of that. With the platform shift, it was all internal developers. Once you externalize the platform, there are people building on top of your APIs and you’re stuck in a place where any breaking change is expensive, but you want to move fast. There is tension between innovation and maintaining backwards compatibility.
That was an interesting challenge for us to solve, especially in light of our cloud transition. We had this framework called P2, which worked well in an on-premises world, and we had also completely rebuilt a new extensibility framework called Connect in the cloud world. Our external developers had to choose between the two. Andrew Chen describes this really well with his new book, The Cold Start Problem. For a network-based product like a marketplace, there’s always that chicken-egg problem. A lot of our business was on-premises, and plenty of our extensibility developers and marketplace partners had built on top of the P2 platform. Then, Atlassian said the cloud is the future; here’s this brand new Connect framework that you have to build on top of. The inertia is in P2, but momentum is gathering in Connect. We’re asking folks to invest in this new platform and make this jump, and for some time they may have to actually maintain both.
Making that transition was a definite challenge we had to solve on the ecosystem side. We had to work through how we incentivized more of our marketplace partners to build on the new platform. What could we do to help our partners? One example: we have a market developer fund to help kickstart development on the cloud platform.
Avanish: Fascinating. I mean, talk about replacing the engine while the plane is flying! But I think you’re one of the best examples of the transition from on-premises to cloud. Kudos to the team, because those of us who have been through that know that it’s not easy, and not all planes keep flying. You’ve given us a fantastic picture of the transition, the journey, and how you’ve defined success.
If you were to talk to someone else thinking about this platform-ecosystem mindset, what kind of guidance would you give them? What are Anu’s top two or three lessons? How should people embark on this journey, if at all?
Anu: Having done this a few times, it’s fascinating to sit back and reflect on what the lessons are from that whole experience. One lesson is to be very clear why you want to build a platform. I see a lot of products aspiring to be a platform. That’s not necessarily the right answer for every product; there’s no need to platformize every single product. Ask yourself if this is really an area that requires horizontal capabilities that you can build vertical applications or use cases on top of. Is there a long tail that requires self-serve or other people to build on top of your platform? Platform is an expensive and patient strategy. It’s not for the faint of heart and it’s not something where you pivot in a year and start seeing results immediately. It’s helpful to be really thoughtful about the why. Is there space for platform here? And is that the foundation of your product strategy?
A second lesson is that strategy is great, but execution matters a lot. No strategy is better than its execution, and executing on platform comes with its own set of problems. It’s one layer of abstraction that you add on top of your current execution, because you’re not just building for your own use cases and internal developers; you’re also building for an external set of use cases and developers, and for your end consumers. The complexity grows exponentially, so it’s really important to think about how you will handle execution while implementing strategy. Beginning the grind is not easy.
The third lesson is on building an ecosystem itself. Like I was saying, marketplaces come with a unique set of problems. If you want to build a marketplace product, you’ve got to think about how-many-sided a marketplace you are trying to build. Who are the different consumers? What does the supply/demand equation look like? Which is the hard side that you want to develop for? All those classic problems that come with a marketplace.
Fundamentally, as a leader, I would say the crux of it, that many people really have to viscerally internalize when they build a marketplace, is that you have to lead with heart. On a successful platform and marketplace, developers are betting their careers and livelihood on you. It’s helpful to remember that these are actual people. There’s a real human on the other side of it. It’s a big responsibility, and you’ve got to take that seriously – but at the same time, that’s what makes working on a platform so rewarding. Constructing an ecosystem is rewarding because you’re fundamentally creating a win-win situation. I really get energized by that.
Avanish: That warms my heart. I think you’re a hundred percent right on that. It’s very hard and it’s not for the faint of heart, but when it comes together, it’s pretty magical. And it creates tremendous economic value, for sure, but also business value and happy customers, which is the end goal, right?
Anu, this has been fantastic. I truly can’t say enough how much I appreciate this.
Anu: This has been fantastic, Avanish. Thank you for the chance to talk to you about this. I remember a couple of years ago coming to you for advice on what to do with our marketplace, and you’ve built this a few times at mind-boggling scales so it’s truly been an honor to chat with you about this.
Avanish: Well, thank you, Anu. It’s fun to see how far Atlassian has come.
The views presented here are those of the individual speakers and not necessarily those of Tidemark Management Company or its affiliates. The content is provided for informational purposes only and is not an offer to sell or the solicitation of an offer to purchase an interest in any private fund managed or sponsored by Tidemark or any of the securities of any company discussed. Companies discussed in these posts may include current Tidemark portfolio companies and/or prior investments made by Tidemark employees while at other investment firms. These companies identified above are not necessarily representative of all Tidemark investments, and no assumption should be made that the investments identified were or will be profitable. The information in this post is not presented with a view to providing investment advice with respect to any security, or making any claim as to the past, current or future performance thereof.