Atlassian & Freemium Framework

Systems thinking and why revenue-first models are the future.

Editors note: Below is an exploration of Atlassian & freemium business models. We wanted to go deeper into why Atlassian chose to go freemium & the often overlooked partners that played a massive role & helping them sell to larger cos without the overhead & headcount of hiring & training.

The essay is written by Sebastián Cuervo (Research Lead at 42 Agency). Illustrations are by the super talented Alejandra Céspedes Cárdenas.

Atlassian’s Freemium Framework - Systems thinking and why revenue-first models are the future.

Freemium, the magical pricing and revenue model that revolutionized the world of startups, can arguably be called an invention by one of the most influential and beloved "tech companies" globally; Atlassian. With legendary stories of business and product success, everyone in the "startup" world seems to use at least one of Atlassian’s products. Quite frankly, if you aren't yet, you probably should be.

Their story begins back in 2002, almost two decades ago, when the internet was a place of remixing and decidedly poor UX. Yes, those were the Dark Ages of having to connect to the internet using a dial-up modem.

In this archaic internet period, Mike Cannon-Brookes and Scott Farquhar created a new product for issue management and project tracking of software code that would revolutionize the world. All it took was $10,000 in credit card debt to get things going and let word of mouth become their primary growth weapon.

According to Cannon-Brookes, their previous experience using enterprise software "was just a horrible experience." As with most digital products, the frustration with the existing options in the category of team collaboration tools for developing digital products led them to create something different. 

And just like that, Atlassian was born, and its iconic product, Jira, an issue tracker system for software development, was released to the public. 

But how did a bootstrapped company manage to grow so much (billions) with a freemium model? To answer that question, we must first look at what they got right from the start, systems thinking, and first principles.

Systems Thinking: Beyond the obvious problem to the process.

As with most problems in life, you have two options. Fix whatever is not working as soon as possible, i.e. the symptom, or try to understand why that problem exists in the first place to make sure it does not occur again. In school, we learn to solve problems quickly, follow instructions, and function within a particular system. That’s the easy path, your classic response of slapping a band-aid on a problem. In life, and particularly in business, a simple problem is seldom that simple, but rather a thread of moments, decisions, or events that trigger such an issue.

In many ways, what Cannon-Brookes and Farquhar started doing by creating an issue tracker system for software development (Jira) was inventing a new and better method of coding. At the time, it was challenging to communicate how code changed from a particular developer to the next, leading to people fixing the same mistake repeatedly or misunderstanding the intent of a fellow programmer’s code. Developing software was slower and much more confusing than it is today.

Atlassian founders, both CS majors, could have quickly tried to create a platform that would allow teams to code faster or to simplify code, among many options. Instead, they realized the problem was not coding itself; it was communication. And so, tracking changes seemed like an intuitive feature that solved a very relevant pain-point for software development. 

What they did, in other words, was not simply improve how coders communicated within their team, but simplify the very essence of coding, from a waterfall-like methodology to an agile-like process (Agile became mainstream in the early 2000s, but Jira was, in many ways, an essential piece that allowed for such work framework to emerge).

In a nutshell, Jira pushed the coding process from civil and mechanical engineering frameworks (waterfall) and early agile methodologies to a predictable and scalable software bug tracking framework that was actually agile-friendly. 

Atlassian did not invent Agile. But Jira was one of the first tools that made Agile software development easier to implement, for both small and large teams.

First-principles thinking and Aristotle.

First-principles thinking, a rather antique way of thinking about the world around us, was discussed and developed by the ancient Greeks (Aristotle, in particular) thousands of years ago. TLDR:  It's a simple way of understanding the world by breaking down problems into essential elements to find new and non-analogic solutions.

The most famous example of such a framework being useful might be Elon Musk's approach to entrepreneurship. As James Clear extensively covers on his blog, several great thinkers such as Johannes Gutenberg (the guy responsible for the Printing Press) use this framework, "but no one embodies the philosophy of first principles thinking more effectively than entrepreneur Elon Musk."

To better understand what this means, think about how SpaceX came to become a successful company in a fraction of the time it took NASA. When Musk decided to embark on his space odyssey, one of the main challenges was finding rockets at a price point he could afford. To solve this problem, he had a couple of options:

  1. Figure out who sold rockets, haggle, and buy them from them cheaper.

  2. Figure out combustion and create a system to do it efficiently and more affordable.

Most people would probably go for option number one. I mean, building a rocket sounds like an impossible task for someone who knows nothing about it. But, for Musk, it made a lot more sense to understand the fundamental building blocks of combustion to develop his company from scratch.

Atlassian, in many ways, did something similar. Even though they were not the first in the software category to think about a bug-tracking solution, as Jay Simmons mentions in this podcast, they found a way to tackle the fundamental communication problems that programmers had in a more friendly manner. Instead of figuring out how to create a way to build software better, they found a way for developers to collaborate and work well together. That meant:

  • Better communication: A platform that tracked issues, but more importantly, that informed teams on the overall product.

  • Fewer meetings: Developers could focus on developing software. 

  • Team-focused software: Instead of selling to large enterprises, they made their product easy to try out (the freemium model). Once the teams noticed it worked, they sold Jira internally to the company. Or so, it seems.

And it is precisely here, in the "Freemium Model" concept instead of the "Enterprise Model," that things become much more interesting to understand Atlassian and its growth model.

The Freemium Dream / Nightmare.

Wouldn't it be fantastic if you could release a limited version of your product that people loved so much they would be willing to pay for a premium version of it?

That is, in essence, the Freemium model. While it isn't a model that works for all organizations, it became the strategy that the early Atlassian team adopted for growing their product.


It makes sense, as mentioned quite specifically in the First Round Podcast interview with Jay Simmons, they were starting their business in a category with similar open-source products, some of them free. If your competition’s product is either free or is priced low, it could be challenging to develop traction for a business by charging enterprise fees. Let's also remember that Atlassian had a particular set of conditions that almost drove it to become almost experimental. First, they were creating a software product in Australia back in 2002 (something quite rare), and second, they didn’t have access to an influx of capital from outside venture capitalists which would help them develop and grow products. These two particular limitations led the founders to broaden access to their product to show the world that it changed the nature of software development in hopes that their target audience would notice.

And they did. As Farquhar mentions in this video, their product:

  • It had to be cheap

  • It had to sell itself

  • It had to sell a lot (Global)

  • It had to be easy to use and try (Freemium)

  • It had to be easy to buy (open, price available, no quotes, no sales).

It was a unique paradigm, perhaps a stroke of fortune, and the insight to create a software product that solved a cultural and communication issue for programmers. It was never positioned as software that tracked issues, but a better way to connect with teammates to create excellent software products.

But, as the world has seen with other "tech companies" (arguably, every company is a tech company, but you get the point), such as Salesforce, word of mouth coupled with a great product are just prerequisites for growth, not amplifiers.

One of its most iconic employees in the sales department, Aaron Ross, said it best:

"It's popular now to believe you can grow a business through values like purpose, happy employees and customers, vision and fulfillment. They work, as proven by companies like Zappos!

But - "purpose" isn't enough, and if your sales are sucking wind, it's really hard to be happy." 

Long story short, Ross joined Salesforce back in the early 2000s and, together with his manager, Shelly Davenport, faced a steep challenge: "to create an outbound sales process that would succeed without any money or marketing support and at a company that was pretty much unknown in the Fortune 2000 market we were growing into at the time."

To solve that impossible task, Ross and his team created a sales process that was predictable and scalable by getting primarily three things right:

  1. Setting in place a predictable Lead Generation process

  2. Training a Sales Development Team between marketing and sales.

  3. Developing consistent sales systems.

And, it also worked quite well. It worked so well that Aaron is now a world-famous consultant in the sales field, and if you are in sales or business development, you’ve either heard about him or read his book.1

In a nutshell, it was a predictable early-stage demand-generation process that allowed Salesforce to become the titan it is today. 

Countless other examples demonstrate that freemium is a great way to acquire users, but not so much to grow your revenue. Take Bidsketch, for example, a site that focuses on helping its users create perfect proposals in no time. As Rob Walling's essay extensively covers, it was precisely when Ruben Gamez, founder of Bidsketch, removed the site's free plan that revenue spiked 800%. Or take Pandora, Evernote, and Mailchimp in this in-depth analysis of their revenue streams and how freemium did not work very well for them either. 

Unfortunately, the inability of the freemium model to covert free to paid customers, seems to repeat itself over time.  Company X wants to accelerate growth and impress investors with their positive metrics, and so they launch a free demo version in the hope that they can cross-sell/upsell to their already existing free customer base. 

How many people do you know that pay for Slack? It is no coincidence that Salesforce acquired slack for 27.7 billion and that prominent figures such as Brent Leary, principal analyst at CRM Essentials, see this move as a strategic step towards taking Slack to enterprises.  I mean, imagine the robust infrastructure and powerful systems of Salesforce together with the video game-like nature of Slack together. It is a brilliant move. No wonder Andrew Wilkinson regrets not taking equity, as he mentioned on this popular Twitter thread last year.

Is freemium a dream? Or is it a nightmare? It might depend on your business and your customers. But, what we do know is that even though in concept it sounds lovely, in practice, up-selling is much more challenging than it appears.

So, what was it about the Atlassian freemium model that worked so well? We already mentioned their particular context; was there something else?

Yes. Partners.

Atlassian Partners.

Finding an analytical account of the importance of Atlassian partners in terms of revenue is almost impossible. The 42 Agency team reached out to 11 Atlassian partners worldwide, and so far, none of them has expressed a desire to talk more about their partnership with Atlassian. Perhaps our emails were not convincing enough; perhaps, there is no intention of sharing how the partnership works. However, we know that, thanks to their partnership programs, Atlassian has managed to expand its sales and support operations without heavily investing in internationalization.

Now, we are not trying to say there is something fishy about Atlassian's partners. There is plenty of information regarding the process of becoming a partner. And it looks pretty sweet. But, as mentioned by Thad West, Co-Founder, and CEO of Isostech (Premium Partner), in this video, it is evident that Atlassian heavily relies on its partners and the ecosystem to deliver professional services to clients. 

In other words, even though Atlassian does not have a dedicated sales team, it does have a team of associated service experts that helps them sell and onboard new clients. In fact, Atlassian tries so hard to create an internal culture that is innovative and engaging that potentially, all employees, through service, should be able to sell the product. Cameron Deatsch, Chief Revenue Officer of Atlassian, claims that Atlassian brought in 17,000 new customers without direct or indirect (partners) contact. That is quite a feat. However, we wonder, how many customers do they acquire with the support of their partners and their service department?

According to Seibert Media, one of the largest Atlassian sales partners, buying Atlassian licenses directly from them has several advantages:

  • Localization: They are located in Germany, and buying from them if you are in Germany simplifies the tax/paying process.

  • They have access to support from both Atlassian and Seibert Media (English and German).

  • Professional, free consultations for Atlassian products as well as a variety of third-party apps.

  • Personalized support (not just email).

We don’t know for sure, but it's likely that their internal customer success department, as well as their partners, have a lot to do with their unparalleled growth. 

This partner phenomenon is not unique to Atlassian. It seems that several SaaS heavily rely on partners, including quiet giants such as Avalara. Likewise, iconic companies that radically change the internet, such as Hubspot, rely extensively on partners, as we can see in this piece where 52 digital agencies share several of the benefits of working as Hubspot partners. According to their partners, working with Hubspot symbolizes trust, knowledge, and even power; why wouldn't you want to be their partner in the first place?

Partners are critical for growth. And, even if it isn't intuitive, it is quite possibly a determinant factor for Atlassian's development.

Network Economies. 

Let's step back a bit and focus on the bigger picture. We noticed a pattern and a fundamental shift of value in macroeconomic landscapes that allowed Atlassian to become an actual titan; the internet networks. 

Before the internet, most people bought a product or service, and profit was calculated as the final price minus costs. If software costs were 1X, and one sold it at 1.1X, companies had a 10% margin that determined their growth model. It was a relatively easy model to understand, and all one needed to do was make sure that your profits matched your expectations. 

And then, and going back to the dial-up modem we mentioned at the start, the internet completely changed the way economies work. 

Network economies are "a new business model that uses technology to connect people, organizations, and resources in an interactive ecosystem in which amazing amounts of value can be created and exchanged."

Platform Revolution, one of the most fundamental readings anyone must have in mind to understand economics today, develops the argument that networks changed how value is created and captured through network effects. What matters for this story and to frame Atlassian's success is that we are at a historical period in which "rather than flowing in a straight line from producers to consumers, value may be created, changed, exchanged, and consumed in a variety of ways and places." It is the connections the platform facilitates that allow entrepreneurs and startups to capture value. 

And with this conceptualization of value, the world radically shifted. 

We see the rise of Uber and its competitors, DoorDash and all its competitors, and an endless series of categories that have completely changed thanks to platforms and networks. If you happened to catch an Uber in NYC a couple of years back, you probably remember how cheap the rides were at certain hours in comparison with taxis.  The same thing happened with almost all categories of human consumption, from movies (Netflix) to banking (Nubank), to booking unique locations worldwide (Airbnb)--it was great.

The Millennial Lifestyle Subsidy phenomenon was probably one of the best and worst things that has happened to us. Companies everywhere were willing to offer subsidies to consumers, to show traction, growth numbers, and move from pre-seed to further rounds of funding and so on. It worked. We bought into it, and if you reflect a bit, it is pretty likely that you recently did a transaction in some platform-like model of a startup that no longer exists because of a greedy and futile customer acquisition model.

It was the golden age of profitless startups that aimed to gain traction and customers to, later on, figure out a profitable model. 

And this paradigm, this new value creation system, just so happens to be the perfect petri dish in which profitless freemium models thrive. 

"This is how we have grown so far, and we'd like to continue this forever, which is — people really like it and so they tell other people about it, and then other people start using it," what a powerful and symbolic quote by Stewart Butterfield (Slack) back in 2016. But, as we have seen and as history teaches us, regardless of your intentions of not having a sales team because the product should sell itself, it does not always work out. In fact, during the pandemic and the rise of remote work, Slack had a unique opportunity to capture market share. And, as we have noticed, not only were they unable to capitalize on the massive enterprise opportunity, but they also ended up selling to Salesforce (a giant that could take them there). 

It’s one thing is to have a perfect product and dream of a world where every worker and team can choose their tools. But, as Aaron Levie, Box's CEO, mentions: 

"The reality with the enterprise is that you can have the best product, but that's not good enough... You need distribution. And what Salesforce has — they have the procurement officers, they have the finance people. They have all of the apparatus you need to interact with to sell software, and they have it for the top 100,000 corporations around the world".

Microsoft did crush Slack even though few people enjoy using the Teams software. The secret? They focused on selling the product & bundled it with their existing suite

In the end, Slack had to sell to enterprises, and since they had proudly ignored that process from the start, Salesforce was an intuitive brand that would help them get there. 

Sales matter. A lot.

In other words, bottom-up customer acquisition exercises are great for building TOFU momentum, but, in the end, you still need sales. This need occurs because teams at large corporations are not independent islands that decide what tools to use but rather a massive interconnected set of gatekeepers, stakeholders, decision-makers, and restraints. If you already use the Microsoft suite (Word, Excel, PowerPoint, etc.) for work, doesn't it seem more likely to go with Microsoft Teams than with Slack?

Well, of course, it does. And, for companies that don’t have the good luck of Atlassian's timing back in the remixing era, it is much harder today to become an enterprise brand without a sales team.

And that is why the networked economy trap is equally powerful and dangerous for startups. Sure, you may not be profitable today; you may not even care about that because your product has product-market-fit and you've acquired incremental amounts of customers month-to-month. With the power of the network and a strong pivot or revenue model, you will be profitable soon. I mean, in a networked economy based on Metcalfe's law with a dash of Moore's Law, assuming that customer acquisition at an accelerated rate can become a value-capturing machine is a fair assumption.

But, it does not always work. Atlassian is a rare find. And, as we have already discussed, their impact wasn't exclusively technological but cultural.

Why shall we focus on revenue-first models?

As we have seen in recent news, and as Kevin Roose puts it, "the fact that some high-end services are no longer easily affordable by the merely semi-affluent may seem like a worrying development, but maybe it's a sign of progress." And, in a way, understanding that freemium models fall within this framework of subsidies and only seem to work to companies like Atlassian could also be something to worry about, but in reality, it may be a sign of sanity in the SaaS space.

Growing a business is hard work. Demand Generation and revenue-first models are almost always essential, and the illusion of traction that Freemium models offer is tempting and dangerous at the same time. 

Atlassian's context, the fundamental understanding of the category, and their systemic approach to product development (and the nature of work and communication) was a huge breakthrough. But, that does not mean that a freemium model can be successful for all SaaS companies or platforms. The maxim, correlation does not equal causation, is far more relevant in life (and business development) than we imagine. 

Maybe for most companies, it would be wiser to not listen to the song of the sirens and instead, listen to Cannon-Brookes' advice and not try to copy their model.

Did you like this essay? Be sure to follow us, and sign up for our newsletter. Do not miss out on our next piece.