AGB 2021.10 - Atlassian (TEAM)
A Sales Efficient Software Company
Subscribe to AGB - One analysis of a good business every two weeks.
“Our mission is to unleash the potential of every team.” - 2020 Annual Report
Atlassian is a leading provider of work productivity software. The company’s main products are Jira, Jira Core, Confluence, Trello, Jira Service Desk, Opsgenie and Bitbucket. Atlassian was one of the first companies to benefit from the shift to agile software development (continuous development through short feedback loops) vs. the waterfall methodology (linear sequences of development). As other teams like HR, operations and marketing move towards an agile work environment, there are Atlassian products for each type of team.
There are many other work productivity software companies out there like Asana, Smartsheet, Basecamp, Monday.com, etc., which could imply that this segment of software isn’t as attractive as verticalized software due to the competitive dynamics, but it’s hard to ignore Atlassian’s success and unique history/operating culture.
Atlassian was founded in 2002 by two graduates of the University of New South Wales in Sydney, Australia. Shortly after graduation, Mike Cannon-Brookes and Scott Farquhar decided that founding a start-up was less risky than getting a corporate software development job. The company started off writing software tools to help developer teams find bugs in their programs. Their first successful product was Jira and it’s been updated many times to what it is today. Other products branched out from Jira’s core offering to service non-technical teams (operations, marketing, HR), service desks, DevOps, etc.
Early sales of Jira were mainly done through Internet marketing via Google AdWords, which the founders admitted were much cheaper back then, and guerilla marketing tactics like showing up to software conferences as attendees. The company still relies heavily on Internet marketing instead of a dedicated enterprise salesforce, which allows for a much leaner and efficient sales structure. 98% of transactions still happen on the company’s website.
Given that Atlassian was based in Australia and the two co-founders were young entrepreneurs, the company never took on venture capital money until some secondary sales later on. Atlassian was bootstrapped with Mike’s credit card and cash flow from the growing business. Atlassian generated $100k in revenues in its first year selling Jira for $800/year - $400 for the license, $400 for maintenance. (The company has done many prices increases since then). Revenues grew to $20M-$30M by 2006/2007 and $50M by 2009/2010. The company finally established some liquidity from two secondary funding rounds of $60M in 2010 and $150M in 2014, before going pubic in December 2015.
Most of this history is directly from the two co-founders on a podcast episode of How I Built This with Guy Raz. You can listen to the podcast episode here.
Atlassian believes that software should be bought rather than sold, which basically means the value proposition of the product on both functionality and price should be compelling enough for the buyer to seek it out. Word of mouth from users that love the product should help drive sales over time.
As mentioned before, the company has a unique go to market model, selling most of its products through online channels. Atlassian spends most of its marketing dollars online and leverages channel partners rather than employing expensive in-house enterprise sales personnel. (As the company has increased in scale and moved up-market, Atlassian now does employ a small enterprise salesforce). As a result, non-GAAP sales & marketing expenses at 15% of revenues (20%-22% GAAP) is low compared to other mid to large cap enterprise software vendors. This has also resulted in a lack of seasonality of its revenues, as typical enterprise salesforces push hard for customers to purchase product towards the end of quarters and fiscal years to meet sales quotas.
Pricing is also compelling for many customers. There are different tiers for customers to choose from depending on the number of team members and the features that they want. From free to starter to basic to premium, there is something for every company depending on its size. Atlassian rarely discounts, given that the price points/seat are fairly low (similar to other work productivity software vendors, but much lower than other enterprise software vendors like Service Now).
The savings from low sales & marketing expenses are reinvested back into the company in the form of R&D. Atlassian has one of the highest non-GAAP R&D expenses at 35% of revenues (44%-49% GAAP). This has allowed the company to develop new products and extensions of existing products (like Jira Service Desk) and develop cloud versions of all of its software products. Recently, 90% of R&D expenses go towards cloud and data center products, including the server to cloud migration tools.
Why is it a good business?
Similar to other software vendors, Atlassian benefits from its customers’ increasing switching costs. When we talk about switching costs it’s usually referring to the implicit costs related to time and effort to migrate data, train employees to use the new software and the downtime that could arise during the migration process.
As users engage with the software over time, knowledge of how to use the software is accumulated and habits are formed. Data that is created on the platform also adds to the stickiness and while migrating the data to another platform is feasible, it wouldn’t be easy. This is why Atlassian measures the success of its products based on the user engagement of the team members that actually use the license.
Atlassian also benefited (first mover advantage) from being the first work productivity software purchase for many of its customers. The company has mentioned that there are still many opportunities to sell to customers that are currently using Microsoft Excel or a homegrown solution.
Atlassian has expanded its product line-up around its original Jira product to increase the stickiness of its software. While Jira Service Desk can be run as stand-alone software, many firms benefit from the integration with core Jira. And if a customer wants Jira Service Management (extension of Service Desk), core Jira must be installed.
The Atlassian Marketplace also adds to the switching costs of its customers as add-ons and power-ups found on the marketplace add to the stickiness of the company’s products. Marketplace revenues has grown nicely over the past 7 years, almost keeping up with subscription revenue growth. Cumulative sales on the platform are more than $1.2B and Atlassian gets a 25% cut, implying over $400M in high margin revenues for the company.
From a combination of an efficient go to market strategy, low customer churn and high net revenue retention (121% for medium and large customers), the company has been able to efficiency grow its revenues and customer base. Atlassian’s CAC ratio is one of the lowest among software companies. Other companies with impressive CAC ratios are Veeva, Workday, Paycom, Coupa, etc.
CAC Ratio = Δ Subscription Gross Profit Dollars / New Customer Related Sales & Marketing
Returns on capital?
Over the past 7 years, Atlassian has spent 70% of its capital on R&D, 24% on acquisitions and 6% on capex. R&D has consistently ranged between 44%-49% of revenues over the past 5 years while M&A has been lumpy. Among Atlassian’s major product lines - Jira, Confluence, Jira Service Desk (with the exception of Trello), most of the product development has been done in-house.
Two of the largest acquisitions have been Trello (January 2017) and Opsgenie (October 2018). Trello was acquired for $363M in cash and $20M in stock. Trello is a virtual sticky notes and digital whiteboards software using the Kanban style of work management. It’s a creative unstructured space that is more visual than Jira or Confluence. While the financials were unknown prior to the acquisition, the company has meaningfully grown the Trello userbase since the acquisition. When Trello was acquired, its userbase had grown over 100% y/y to 19M in over 100 countries. By September of 2017, Trello had 25M users and that number doubled to 50M over the next two years. Trello is used by over 80% of the Fortune 500.
Opsgenie was acquired for $260M in cash + $36M in stock for key employees as a retention tool. Opsgenie is incident alerting and on-call schedule management software that complements Jira Service Desk. The financials for Opsgenie is also unknown and its uncertain what the return profile has been for this acquisition. However, we do know that the company has improved the offering post the acquisition, introduced a new free plan tier (up to 5 users) and reduced pricing by 35% to better align the pricing structure with the rest of Atlassian software offerings.
Atlassian started by selling on-premise software through the form of perpetual licenses + maintenance contracts. Later in the early 2010s Atlassian began to offer cloud versions of their software products and since FY13, it’s been the fastest growing part of the business.
The company has recently been pushing its customers to move to the cloud versions of its products. From the customer’s perspective, they don’t have to maintain a server or data center to host the software and there are other benefits like security and scalability. Over 95% of new customers choose the cloud version of the software already.
From the Atlassian’s perspective, moving customers over to the cloud is a win in-terms of lifetime value of the customer, but the benefits may be less than a typical on-premise to cloud transition. That’s because Atlassian’s license portion of its software is underpriced and the annual maintenance fee is much higher than other software vendors. For Jira, a small customer with under 50 users would pay $6k for the first year ($3k for the license and $3k for maintenance) and then pay $3k every year going forward. 50% of the first year fees in maintenance fees is much higher than other software vendors who typically price it around 20% of first year fees. There also isn’t the resell that other on-premise software vendors go through when a new version of the software is released.
While the company is pushing for all customers to move to the cloud version, Atlassian should benefit much more from the large enterprise customers. The example the company gave at its Analyst Day in 2020 was for a customer with 2,200 users. The annual maintenance for this customer would be $25.2k but the cloud version would ramp up to $92.8k in year 3 after getting discounts for the first two years. The selling point to these enterprise customers is that the savings from not needing support staff to service the software and the server hosting costs outweigh the increase in the subscription fees, resulting in lower total cost of ownership. Atlassian is well aware that enterprise customers tend to move slowly in these migrations, so the company is offering discounts for moving earlier (55% reduction in FY21, 40% in FY22 and 20% in FY23).
We estimate that Atlassian has generated returns on incremental capital invested between 40%-60% over the past 5 years. This is inline with other high returning software companies like Veeva and Workday. While it would be interesting to see what the returns have been for the acquisitions, the returns on R&D are likely high enough to make-up for any lower returning acquisitions.
Atlassian has consistently grown its customer base with small teams (Fortune 500k) but recently has had more success penetrating larger enterprises. Over the past 5 years, annual growth of customers with over $50k and $500k in billings have outpaced the company average by 17% and 34%, respectively.
The outperformance with larger customers has to do with Atlassian’s investment into enterprise sales/product advocates and channel partners. While the company still doesn’t like to admit to having a traditional enterprise salesforce, the growth in these sales associates has helped the company move up market.
Atlassian estimates that its TAM is 1.1B knowledge workers, of which 45M are software team users and 100M technical team users. This compares to the company’s monthly active users of 15M (roughly half of which are technical), which implies that penetration into technical teams is roughly 7.5%. In terms of dollars, Atlassian views $13B is addressable for technical teams and $11B for non-technical teams, totaling a $24B opportunity for the company.
When we talk about SaaS companies, a second product or platform is usually what’s needed for the company to exceed expectations (and for the stock to keep working). However, for Atlassian, this really hasn’t been the case. Core Jira + Confluence have been growing nicely and is still remain the largest contributors to revenue for the company. At the time of Atlassian’s IPO, Jira + Confluence made up 75% of revenues. As of the last analyst day, that number has gone down to 60%.
The add-on modules for Jira and Confluence from Atlassian Marketplace and the extensions like Service Management add to the growth of the core products. Atlassian recently launched a $50M venture fund to support applications that are being built to complement Jira and its other products on the marketplace.
With a reinvestment rate between 35%-50% and a return on incremental capital between 40%-60%, we estimate that Atlassian’s value has increased between 20%-25% over the past 5 years.
What else is important?
Price increases pulls forward revenues
As mentioned previously, Atlassian tends to underprice its products for the amount of functionality the customer receives. This is especially true for the starter and free tiers. The company recognizes that this is a great way for customers to try their products and increases the number of opportunities Atlassian has to pitch their higher tier packages as these customers grow.
For the larger customers, there are relatively material annual prices increases that usually occur near the end of company’s fiscal first quarter (September). These prices increases average 15% but they’re usually not across all product lines. The company continuously tries to optimize pricing while still leaving some customer surplus. Atlassian’s view is that the added functionality that is provided each year more than warrant the price increases. This was also the view from the company when the original pricing scheme for Jira was introduced, with maintenance being such a large % of first year fees.
With these annual price increases, there are some pull forwards from customers that try to re-up their software agreements prior to the increase. There will usually be talk of pull forward revenues in the fiscal quarter prior to the price increase.
The company could do more work on developing partnerships with other large software vendors as another distribution channel for its software. The medium to large enterprise market is still a great opportunity for Atlassian but the company may have to invest more in an in-house salesforce, which could put pressure on margins.
If you made it this far, I hope you received some value from reading our analysis. Please subscribe to the free newsletter and share with anyone that would find it valuable. Thank you for your support!
Hi YHB, great article, thanks!
The most telling chart in this whole article, for me, was the Large Customer Growth charts.
What, if any, are the implications of this? At first I thought that it will increase their operating leverage, but unless this cohort is more retentive, or they're able to cross/up-sell more to it, it should stay the same.
In the first order, I think it means that the rate of revenue increase will go up, what do you think are the second order effects of $TEAM trying to capture more of the up-market/big-ticket segment?