Oct 9, 2023, 4:34 AM ETFireside Chat, Interview

Scott Farquhar: Founding Atlassian; How We Scaled to a $50B Valuation; The 4 Jobs of a CEO | E1070

Scott Dunn and Mike Cannon-Brookes co-founded Atlassian in 2001 during the "nuclear winter" of the dot-com crash and post-9/11 market, bootstrapping the company for 15 years without venture capital.
Atlassian grew to serve 250,000 customers with a staff of 11,000 over two decades before accepting external funding.
The founders rejected the myth that startups must raise early capital to survive, noting that in niche markets with low competition, bootstrapping remains a viable path today.
Scott Dunn identifies "startup glamour" as a pervasive BS myth, attributing it to survivor bias while emphasizing that most startups fail hard and the reality is often unemployment rather than prestige.
The company's mission evolved from solving bug tracking to "unleashing the potential of every team," a shift that Dunn believes generates more passion than the original product problem.

Atlassian raised its first venture capital round 10 years into operations when the company had a ~$60 million run rate and 20-30% profit margins.
Dunn utilized a "closed-envelope bidding" strategy for the $60 million raise, rejecting indicative offers to prevent VC collusion on valuation.
Excel Capital emerged as the winner with a $405 million valuation (5-6x revenue), injecting $60 million, significantly higher than the average $200-$300 million range from other bidders.
The company ran a similar bidding process for a pre-IPO liquidity round, securing valuations up to $4 billion.
Scott Dunn identifies his biggest regrets as "mistakes of omission," specifically failing to invest aggressively enough in new opportunities like Bitbucket (GitHub competitor) and Stride (Slack competitor) despite having strong products.
Dunn admitted that for the first VC round, if Benchmark had matched Excel's price, Atlassian would have split the deal to accommodate two partners, a decision Benchmark reportedly regrets given their eventual 100x return on investment.

Atlassian shut down its Stride product (formerly HipChat) after a full rewrite, a decision driven by the inability to scale the $2/month pricing model against Slack's freemium model and the realization that doubling the team size did not yield proportional productivity gains.
The decision to kill Stride was framed by Dunn's criteria for team impact: "Could I look my staff in the face and say that the next five years of their life was going to be the biggest impact to humanity working on this product?"
Atlassian is launching a new internal incubation program called "Point A" to ensure aggressive investment in new categories, such as Atlas (OKRs) and Compass (developer portal), to avoid repeating the under-investment mistakes of the past.
Jira's complexity is acknowledged as a trade-off for scalability; while the company continually simplifies the UI, some complexity stems from user configuration rather than the core product.
Pricing models are expected to shift away from per-seat licensing as AI agents enable customers to solve problems without proportional increases in human users, potentially leading to "free" software or value-based pricing.

Atlassian treats AI as a necessary strategic shift, noting that while "commodity" AI features (text summarization) are ubiquitous, the real opportunity lies in integrating unique customer data sets to solve complex workflows (e.g., auto-fixing bugs based on log files).
The company employs hundreds of engineers on AI initiatives but avoids building its own massive Large Language Models (LLMs), focusing instead on integrating external models to access customer data effectively.
Atlassian operates as a fully remote-first company, a "one-way door" decision made in 2020 that allows hiring from rural areas and countries without local offices, resulting in lower attrition and 92% of staff believing remote policies aid their best work.
The company found that "intentional togetherness" (gathering staff periodically) boosts team connectedness by 30% for 4-5 months, negating the need for daily office attendance to maintain culture.
Dunn notes that while small companies can move faster, Atlassian leverages its 250,000 customers and terabytes of historical data to run AI experiments at a scale that can actually accelerate learning.

Dunn identifies his core strengths as "systems thinking" across disciplines and his primary weakness as prioritization, requiring a support system to focus on high-impact activities.
He cites the four critical CEO responsibilities as hiring/firing the management team, setting vision, establishing culture, and allocating resources (the latter ideally on a quarterly cadence).
The "Team Anywhere" policy is considered a critical, non-reversible decision that burns employee trust if reversed, having placed half the workforce outside office proximity zones.
Dunn's relationship with money involves a conscious trade-off between time and cash, using wealth to purchase time for family, though he admits the challenge of raising wealthy children with humility is unresolved.
Key personal principles for success include putting one's partner on a pedestal and the habit of immediately apologizing when wrong, a behavior learned from his wife over 15 years.
Dunn's philanthropy stems from a 2005 experiment to raise charity funds via a $5/user, 5-day pricing model, which raised $100,000 for Room to Read and inspired the "Pledge 1 Percent" movement.

Dunn predicts that the list of top technology companies will look very different in five years due to the disruptive potential of AI, similar to past shifts from mainframes to PCs to mobile.
The long-term vision for 2033 focuses on reducing "busy work" and "work about work," utilizing AI and emotional intelligence training to help teams collaborate more effectively and enjoy their roles.
The company aims to maintain its position by integrating deeply into workflows where unique data and multi-product ecosystems create value that standalone AI tools cannot replicate.