Atlassian

The most consequential decision in Atlassian's history was made by two 22-year-olds who simply refused to wear suits. In 2002, Mike Cannon-Brookes and Scott Farquhar, fresh out of the University of New South Wales, looked at the graduate jobs on offer in Sydney and made themselves a modest promise: match the corporate starting salary of AUD $48,500 without putting on a tie. They charged $10,000 to their credit cards, wrote a bug tracker, and named it Jira - after Gojira, the Japanese name for Godzilla, because the tool it replaced was called Bugzilla. The monster ate the market.

Twenty-four years later, Atlassian is a NASDAQ-listed company with fiscal 2025 revenue of $5.22 billion, more than 300,000 customers, and a product suite - Jira, Confluence, Trello, Bitbucket, Jira Service Management, Loom, and the AI assistant Rovo - that functions as connective tissue for how modern companies plan, track, and ship work. In its most recent quarter, revenue grew 32% year-over-year to $1.8 billion, and cloud revenue alone passed $1.1 billion. The suits never happened. The salary target was exceeded by a comfortable margin: the founders became Australia's first technology billionaires.

The company that fired the sales pitch

To understand what makes Atlassian unusual, start with what it didn't have. For most of its first decade, the company had no traditional sales force. No cold calls, no golf outings, no "contact us for pricing." Prices were published on the website, entry tiers were cheap enough to slide under a manager's expense threshold, and the product was expected to sell itself. Where a legacy enterprise vendor might quote $80,000 and send a rep, Atlassian charged a few hundred dollars and sent a download link.

This was not frugality. It was the whole strategy. By stripping out the cost of selling, Atlassian could underprice competitors, spend the savings on engineering, and let adoption spread laterally - one team installs Jira, the team next door sees the board, and six months later the whole department is on it. The company calls this its flywheel. Analysts call it product-led growth, a category Atlassian was running years before the term existed. The financial signature is distinctive: Atlassian has historically spent a fraction of what enterprise peers spend on sales and marketing, and poured an unusually large share of revenue into R&D instead.

The model has since grown up. As Fortune 500 accounts got large enough to need hand-holding, Atlassian added an enterprise sales motion for its biggest customers - but as a layer on top of the flywheel, not a replacement for it. The result is a customer base that runs from three-person startups paying nothing on free tiers to the majority of the Fortune 500 signing multi-year platform deals.

"Don't #@!% the customer."Official Atlassian company value, profanity intact

What it actually sells: coordination

Atlassian's product catalog can look sprawling, but every item answers the same question: how does a group of people agree on what the work is, who is doing it, and whether it is done? Jira, the flagship, tracks units of work as they move through a workflow - originally bugs, now everything from sprint tickets to marketing campaigns. Confluence, launched in 2004, is the shared brain: documentation, meeting notes, plans, and institutional memory. Jira Service Management turns the same machinery toward IT help desks and incident response, a business that has quietly become enormous - Atlassian's Service Collection passed $1 billion in annual recurring revenue this year, growing over 30%.

Around that core sit the acquisitions: Trello, the friendly kanban board bought in 2017 to reach non-technical teams; Opsgenie, the on-call alerting tool added in 2018; Loom, the async video messaging platform acquired for roughly $975 million in 2023; and Bitbucket, the code-hosting service that anchors the developer side. In September 2025, Atlassian agreed to acquire The Browser Company, maker of the Arc and Dia browsers - a signal that its ambitions now extend to the window through which knowledge workers see the internet itself.

Year	Acquisition	What it added
2010	Bitbucket	Git code hosting for professional teams
2017	Trello · $425M	Visual boards for non-technical teams
2018	Opsgenie	On-call alerting and incident management
2023	Loom · ~$975M	Async video that replaces status meetings
2025	The Browser Company	Arc and Dia browsers; the desktop for work

The pattern across every deal is the same: acquire habits, not just revenue. Trello, Loom, and Arc were all products people chose for themselves before their companies ever wrote a check. Atlassian then wires those habits into its platform, where the switching costs do the rest.

The AI wager: more work, not less

Every software company now claims an AI strategy. Atlassian's is at least a distinctive one. While rivals pitch AI as a way to shrink headcount, Atlassian argues the opposite: AI generates more output - more code, more content, more decisions - and therefore more coordination for someone to manage. Its answer is Rovo, an AI teammate launched in 2024 that searches across a company's tools, chats with context, and runs agents inside Jira and Confluence.

The early numbers back the bet. Rovo's monthly active users are counted in the millions, usage of its AI credits is growing more than 20% month-over-month, and customers who use Rovo are expanding their spend at roughly twice the rate of those who don't. On the strength of that engine, Atlassian posted its first billion-dollar cloud quarter in fiscal 2026 and followed it with 29% cloud growth the quarter after.

Annual revenue, selected fiscal years (US$B)

$0.6BFY17

$1.2BFY19

$2.1BFY21

$3.5BFY23

$5.22BFY25

Company reports. FY26 pace: $1.8B in Q3 alone, up 32% year-over-year.

Everyone complains about Jira. That's the moat.

Spend five minutes on any developer forum and you will find Jira jokes - about ticket bureaucracy, about workflows with more states than a federation. Then notice that the people making the jokes are still using Jira. This is the paradox competitors keep underestimating. Jira's configurability is exactly what lets a bank, a game studio, and a hospital all bend it to their process, and every ticket, custom field, and integration a company adds makes leaving harder. Atlassian's true moat isn't any single feature. It is the accumulated organizational history stored inside its products.

The competitive field is crowded from every direction: Monday.com, Asana, and ClickUp in work management; GitLab and Microsoft-owned GitHub in developer tools; ServiceNow, Zendesk, and Freshworks in service management; Notion in docs; Linear among newer issue trackers beloved by startups. None of them fights Atlassian across the whole board at once. Atlassian's position - a single platform spanning developers, IT, and business teams, priced to spread bottom-up - is the thing each rival matches only in part. Its closest structural analogue is Microsoft, which is precisely why Atlassian has pushed so hard into the cloud platform, System of Work messaging, and now a browser.

Two founders, equal stakes, twenty-two years, zero public feuds - a co-CEO structure most investors call a red flag ran longer than most companies last.On Cannon-Brookes and Farquhar, who split the top job until 2024

Culture as a published document

Atlassian's values read like they were written to fail a legal review, which is rather the point. "Open company, no bullshit." "Don't #@!% the customer." "Play, as a team." "Be the change you seek." "Build with heart and balance." They have survived, profanity intact, into the official filings of a public company - a small daily test of whether the culture is decoration or infrastructure. The company runs quarterly 24-hour "ShipIt" hackathons, co-founded the Pledge 1% movement that commits equity, product, and employee time to philanthropy, and operates under "Team Anywhere," a permanent distributed-work policy covering roughly 12,000 employees across 13 countries. A company whose entire product line is remote collaboration decided to eat its own cooking, and kept growing 20-30% a year while doing it.

Leadership, too, broke convention quietly. Cannon-Brookes and Farquhar ran the company as co-CEOs for over two decades - a structure conventional wisdom says cannot work - and ended it the rarest way possible: calmly, with Farquhar stepping down in August 2024 and Cannon-Brookes carrying on as sole CEO. Along the way the brand has picked up unexpected real estate, from the timber Atlassian Central tower rising in Sydney to the nose of a Formula 1 car: since 2025, Williams competes as Atlassian Williams Racing.

What you can do with it

For a reader deciding whether any of this matters to them, the practical version is simple. A small team can start free: a Jira board for the work, a Confluence space for the thinking, Trello for the lightweight stuff, Loom for updates that would otherwise become meetings. A growing company can graduate tier by tier without ever talking to a salesperson. An enterprise can consolidate service desks, incident response, agile planning, and documentation onto one cloud platform, then point Rovo at the whole pile and ask it questions. And for founders, Atlassian is a case study worth stealing from wholesale: publish your prices, make the first purchase trivially easy, spend the sales budget on product, and treat acquisitions as a way to buy your customers' existing habits.

The company that began as two graduates dodging neckties now describes itself as the teamwork platform for the AI era, where humans and AI agents deliver work together. Strip the language back and the through-line from 2002 holds: software's hardest problem was never the code. It was getting people to agree on the work. That problem is not going away, and Atlassian has spent a quarter of a century compounding on it.