top of page

Software's Infinite Leverage

Archimedes, one of the greatest scientific minds in antiquity (c.287-212 BC), was the first to describe the mathematics of levers. Given a fulcrum point, a small mass can balance a much heavier one at the opposite end of a lever:

Given a long enough lever, anything is possible. “Give me a place to stand, and I will move the earth,” is Archimedes’s enduring epithet.

In the year 2020, the equivalent of the long lever is software.

Anyone in the world can head to, type some code into their browser, and instantly have thousands of machines working on their behalf, serving billions of customers worldwide.

Jeff Lawson, CEO of our portfolio company Twilio and one of the first product managers at AWS, said it more eloquently: “Writing software is one of the cheapest and most scalable forms of innovation humanity has ever seen. Think about it: you have a blinking cursor, and you can type some magical code, and suddenly you make a global network of thousands or millions of computers do your bidding. Why wouldn’t you do that all day, every day?”

Despite the incredible power of software, there are only a couple dozen million software developers worldwide:

Not surprisingly, ask any tech CEO what is the biggest constraint on the growth of their business and the answer is unanimous: they can’t get enough software engineers.

Our portfolio company Shopify has even announced a program, Dev Degree, in which students spend 25 hours per week working at Shopify while taking three computer science courses per semester at university. Shopify picks up the tab for the tuition and training, and the results so far have been very encouraging, with higher retention and quicker time to productivity. (If you have teenagers who want to study computer science, this sounds like a no-brainer deal!)

Given the strong demand for software developers and above-average pay (>$105k per year compared with a national median of $39k), the Bureau of Labor Statistics estimates the occupation will grow by 20 percent per year over the next decade, much faster than the 5 percent growth estimated for all occupations.

The increasing power of software and its practitioners has led to the creation of several developer-first companies. Twilio, which we have described in detail in past letters and blog posts, creates application programming interfaces (APIs) for communications and customer engagement.

Another company that started out with a focus on developers is Atlassian, which we began buying in January 2019. Bootstrapped by 23-year-old Australian students Mike Cannon-Brookes and Scott Farquhar with $10,000 in credit card debt in 2002, the company has reached mythical status for its incredible business and financial success, as well as for the laid back and down to earth attitude of the founders.

Their first product was Jira, a team planning and project management software, followed by Confluence for team content creation and sharing. Delivered over the internet, they were early adopters of the software-as-a-service business model. Relying on Google AdWords to sell their products, by 2004 Atlassian had 2,000 paid customers. Six years later, the company was generating $100 million in yearly revenue from its 20,000 customers—despite not a having a single salesperson.

Growing through new product development as well as through smart acquisitions, Atlassian this year is on track to generate nearly $500 million…of free cash flow, on a revenue base of nearly $1.6 billion, serving 171,000 paid customers.

As it builds and acquires more tools for developer and team collaboration, Atlassian is penetrating a growing market.

It is, in effect, giving the power of software to an ever-wider audience. Earlier this year, Chief Marketing Officer Robert Chatwani explained:

“When we look at the numbers globally, roughly, there's about 25 million to 28 million software developers. When we think about the teams that work with software engineering directly, technical team members, there is about 100 million of them: data scientists, product managers, designers, and folks running experiments. And then when we think about the broad base of knowledge workers, it’s roughly 800 million. So, we're looking at this total global market of 1 billion potential end-users who can benefit from teamwork and collaboration software.

“When I think about how penetrated we are, I would say that we are in very, very early days of the potential that we have with all 3 segments. Now with software developers, 85% of software developers have unaided awareness of Jira. And so there's hardly a software developer in the world who won't know about Jira. That said, we are seeing that segment grow, and that's being really driven by companies who are operating and adopting more software principles across their organization.

“Even though we have a great foothold with Jira, what we see is Jira's vector of growth is expansion with more software developers and then adjacent teams as others start to unify their work with that core group, and similar trends in IT and knowledge management for collaboration. We have 170,000 customers at Atlassian. Globally, there are 1 million companies with $10 million or more in revenue. We think every one of them can be an Atlassian customer. And so relative to the potential, we're still quite early.”

One can think of these customers and the potential market as being “17 percent penetrated,” but it is important not to forget that the market itself is expanding. Not only that, but the base of users in Atlassian’s free tier is in the millions; these are all potential paid customers down the road. Atlassian’s goal is to have every knowledge worker on the planet use at least one of its tools once a day.

The company is well-protected from disruptive entrants. Using the Innovator’s Dilemma playbook and harnessing the zero-marginal cost of distributing software, Atlassian has free tiers across many of its products.

This makes “cheaper” entrants impossible (nothing’s cheaper than free) and protects the low-end. As teams expand their usage of Atlassian and require more extensive features, they can start paying a low monthly fee and grow from there.

This is the Atlassian flywheel; note this was before they created the free tier at the low-end:

Because of Atlassian’s extremely efficient (self-serve, online) model, it has a remarkably smooth sales pattern, adding new customers and expanding with existing customers every minute of every day:

With tens of millions of visitors to its website every month, and the frictionless nature of trying its products, customers will buy and try by the thousands. As one example, on August 8, 2018, Atlassian sold $3.5m in software to over 8,000 unique customers; another 20,000 customers began free trials. This is an enterprise software company operating at consumer scale.

On top of this sits an entire economy. Globally, Atlassian hosts city meetings, roughly two every day. There are over 25,000 developers on the Atlassian platform, 76,000 people on LinkedIn with “Atlassian” as a job skill, and an app marketplace—for apps that complement Atlassian’s products—that has already generated $500 million in sales since inception.

With its strong engineering culture and continuously improving software (the company deploys updates 2-3 times per day, many invisible to the end user), Atlassian is aiming at reaching $10 billion in revenue.

With an exceptional founder-led culture, great products and sales motion, and a growing market, Atlassian has the wind in its sails. The company exemplifies what is possible with the leverage of software: serving an ever-growing base of customers with frictionless digital products. We are happy to partner with this team as they develop their business over the next decade and beyond.

[Note: This post is an excerpt from our Q1 2020 letter to investors.]


bottom of page