How Atlassian’s dual-CEO structure has helped the company thrive

In this weekly series, CNBC takes a look at companies that made the inaugural Disruptor 50 list, 10 years later.

In early March, collaboration software maker Atlassian published a blog post titled, “Atlassian stands with Ukraine,” laying out the company’s plans to support employees and customers in the region and announcing it was “pausing the sale of all new software to Russia.”

The post was signed by co-CEOs Scott Farquhar and Mike Cannon-Brookes. They went back and forth on the content and the main points. But Farquhar did most of the work, freeing up Cannon-Brookes.

That’s one of many conveniences of keeping two people at the top of a company. The atypical structure has helped propel the Australians’ 20-year-old firm into the top tier of the competitive software industry, with products so well known that big companies might find it difficult to move away.

In 2013, Atlassian landed on CNBC’s inaugural Disruptor 50 list of private companies worth watching, ahead of its 2015 Nasdaq debut. The stock has risen almost 1,000% since then, compared with 124% growth for the S&P 500 over the same period.

The duo have had the same job at the same company for 20 years, they were born one month apart, they became parents three months apart, they were best men at each other’s weddings, and they own property next to each other in Sydney. “Our stock ticker is TEAM, and so, yeah, that’s what we’re about,” Farquhar said.

But they’re different people. Cannon-Brookes is a long-haired idealist, who became an owner of a U.S. basketball team and attempted a takeover of Australian utility company AGL Energy. His comments are sprinkled with obscenities. Farquhar is clean-cut and careful as he speaks. Early investor Rich Wong of Accel calls Farquhar more analytical.

“Mike is kind of the quintessential unreasonable man,” Farquhar said. “‘The world should work this way.’ ‘Mike, it doesn’t yet.'”

Farquhar and Cannon-Brookes became friends in the late 1990s after taking the same course at the University of New South Wales. The technology bubble broke around the time they graduated, and facing a dearth of job prospects, they formed a business. Initially it offered support for another company’s application server. Then it changed direction and started building its own software. The first version of Jira, a tool for tracking issues and projects, appeared in 2002.

Within about five years, many Accel-backed start-ups had embraced Jira. “It was already the standard that you had to integrate your software with,” Wong said. The company widened its portfolio with the 2004 launch of document-collaboration service Confluence and the 2012 acquisition of team messaging app HipChat. Along the way, Atlassian released versions of Jira for different types of workers.

Today Jira is a market leader, transcending its status as Silicon Valley darling and overtaking heavyweights with decades of experience selling to enterprises. Atlassian controlled greater share than any other company in the market for software change, configuration and process management tools in 2020, ahead of Microsoft, IBM and Broadcom, according to an estimate from researcher IDC. Atlassian’s revenue in the market grew about 22% year over year, faster than the overall category, which expanded almost 15% to $4.8 billion, based on IDC’s data.

Part of the momentum derives from programmers getting to try out Atlassian’s software for free before they pay for it. The strategy goes back to the founders.

“Our exposure to software started with things like games,” Farquhar said. “Back then, games had different business models. You could buy your PlayStation ones shrink-wrapped. If you look at Id Software, they came out with a shareware model, sort of try before you buy. We thought that was a great way to sell software, because of course you want to try before you buy. At SAP, there’s no trying. You get to see what it looks like, because it takes that long to implement it.” (SAP does offer free trials for some of its products.)

Atlassian was either the first or very early to sell software with a freemium offer, Farquhar said, adding that cloud file sharing app maker Dropbox made it more popular. And in the late 1990s Red Hat, which IBM later acquired, gave away CDs containing its distribution of the open-source Linux operating system and permitted people to download it free of charge.

Lacking a pile of money from venture capitalists for its first eight years, Atlassian skipped the custom of assembling a squadron of salespeople to score deals. Now, though, there are a few on staff who pursue select business opportunities, Farquhar said.

Focusing less on selling hard and more on delivering products people actually want to use has given rise to a robust financial profile. Atlassian enjoys the fifth widest gross margin of all 76 constituents of the WisdomTree Cloud Computing Fund, at 83%.

That status has caught the attention of investors.

“In my history of 33 years of doing this, I have seen more than a handful of companies that have tried to do it without an internal salesforce, or an external salesforce, either. The thing I would say about Atlassian is they’re the most successful at it,” said Brendan Connaughton, founder and managing partner of Catalyst Private Wealth, which held $91 million in Atlassian stock at the end of 2021, its largest position at the time.

Like many other cloud stocks, Atlassian isn’t actually profitable. Connaughton said Cannon-Brookes and Farquhar would find it easier to turn Atlassian into an actual moneymaker than its peers, thanks to its relatively sparse sales team.

A more prominent feature of Atlassian’s 7,000-person organization is the group that actually builds the company’s wares. Engineering, product and design report to Cannon-Brookes. Farquhar supervises legal, human resources, finance, sales, marketing and customer-support teams. “I’m sort of the grandparents,” Farquhar said. “I leave him to deal with the temper tantrums and the screaming.”

When they talk about responsibility, they consider both skills and enjoyment. You don’t want someone who’s good at handling a task but doesn’t like doing it, and vice-versa, Cannon-Brookes said.

Marketing and sales reported to Cannon-Brookes for 15 years, and engineering once reported to Farquhar. And they’ve both run the entire company at different times. They’ve gone on sabbaticals. Last year Farquhar took three months off to caravan with family around northwestern Australia. “We sort of got to travel unencumbered,” he said. “I think other CEOs would have to retire or quit to be able to take a break that long.”

The structure has contributed to Atlassian’s success, said Gregg Moskowitz, an analyst at Mizuho.

“I think it has helped, having two strong executives at the very top who see eye to eye, at least on all the important issues,” he said. Other technology companies have employed CEOs in pairs, including Autodesk, Ceridian, Oracle, Salesforce, SAP and Workday. Alphabet’s autonomous-driving subsidiary Waymo recently went the co-CEO route.

The strategy has a mixed history, Moskowitz said, saying it didn’t work well at all at handset maker BlackBerry. The relationship between co-CEOs Jim Balsillie and co-founder Mike Lazaridis “had gone cold,” according to one account, and the two stepped down.

What’s different for Atlassian is both Cannon-Brookes and Farquhar are founders, said Wong, the Accel investor. Their combined knowledge helps them move faster, he said.

Wong pointed to Atlassian’s 2017 acquisition of task-management app Trello for $384 million, still the company’s largest deal to date. At Trello it was a shock, because Atlassian’s Jira was viewed as a competitor, said Stella Garber, who ran marketing at Trello at the time.

“I think it took conviction of the founders to say, ‘I know we could have built it, but it would take us time, and it would really expand the organization if we make the choice now and pay what it takes to get the acquisition done,'” Wong said.

When there’s an issue on Cannon-Brookes’ turf, it’s his decision to make. But when it’s something big, he consults with Farquhar, because it’s almost certainly going to affect them both. There are plenty such examples in and around the company right now, and it’s natural that they divvy things up.

“The pandemic and Russia and Ukraine — at the moment Sydney is under floods,” Cannon-Brookes said. “Put it all together, and there’s a lot of things you need to deal with in a growth biz that aren’t just the product.”

Farquhar said he and Cannon-Brookes had long conversations about what to do with their team-messaging app Stride, which arrived in 2017 as Slack and Microsoft Teams were gaining momentum.

“It was weird, actually, because everyone was talking about how good Slack is. We were using Stride internally,” Farquhar said. “The product was actually better. The Slack thing is amazing. It’s actually not as good as what we had. We had to make a decision.”

Ultimately, Atlassian shut down Stride and HipChat Cloud and sold the intellectual property to Slack. It also bought an equity stake in Slack, which shot up in value as Slack stock appeared on the New York Stock Exchange in 2019.

When Cannon-Brookes and Farquhar were younger, they could close the office door and have a conversation with each other about a crisis, and for fun they might go mountain biking or drink beer together. The pandemic stopped them from seeing each other in person so often. They’ve gotten good at connecting on Zoom, Farquhar said.

Cannon-Brookes doesn’t need to massage what he says to Farquhar. Without prompting, he imagined what would happen if Farquhar were to leave.

“I’d be constantly explaining things, which would feel like I was talking down to someone,” he said. “‘Good idea, but let me tell you what happened in 2012.'”