Post by Andrés Moreno
At Vaibmu, we are working with great researchers and developers on U-Qasar, an open and flexible solution for integrating quality monitoring data that comes from developing software (Software Quality Assurance or SQA). U-Qasar has started as a European funded project (FP7), and it is the goal of the consortium that the knowledge and product created in the project lives after the project ends on September of this year. Vaibmu’s role in this project is to help the consortium decide how to make U-Qasar sustainable in a way that it brings value to companies using it. As many other companies and initiatives starting up, we are using Atlassian solutions, a Confluence wiki to manage the project and a Jira instance to track the software development process. How did Atlassian solutions become so universal? This blog post will try to explain it, and at the same time help us clarify our thoughts about possible ways forward for U-Qasar.
Atlassian is an Australian company funded by college friends Mike Cannon-Brookes and Scott Farquhar in 2002. Back then, their goal was to do something more interesting than what their college was offering, with humble financial expectations: to earn as much as a junior consultant. It took them a couple of years to even earn that much. In this period, they found out that they wanted to enter the enterprise market, and disrupt it with a different business model.
Their first product to reach the market was Jira. Originally a bug-tracking software to compete with Bugzilla, from Mozilla, in the enterprise market. From the need to document the knowledge around Jira, they developed a wiki, which they quickly sold, and still do, to customers under the name of Confluence.
They realised that they had to lower the overhead of distributing their software. Enterprise software is costly and requires salespeople to close deals. Atlassian could not hire salespeople, and instead focused on a transparent pricing model. Everything could be bought straight from the internet and no secret discounts could be found elsewhere. What now seems to be the desirable, even expected model, was very uncommon then for enterprise software.
As they could not hire salespeople to sell expensive software, they lowered the price. Thus, it all boiled down to simple facts: to make profit they had to sell more licenses. Thus, both the web and Atlassian’s open culture were paramount to their success in equal doses. Those two factors helped them to reach more and more customers. These satisfied customers could easily refer more people to Atlassian software, who could go online and buy the software with a simple clic. No phone calls or email conversations with sales managers.
They also tried to combine this word of mouth mechanism with another revenue source: OEM sales. It did not succeed. By making their software part of a bigger package, and sold under a different name, word of mouth ceased to work.
Laser focus on target customers
Later, as the company grew, they saw that only companies with 20+ employees were buying their software. How could they attract smaller companies? The solution came after a fundraising gala for Room to Read, an NGO building libraries across the world. Atlassian would support startups and small companies with a cheaper starter license ($10), which they would donate in its entirety to Room to Read. In this simple action, they signed up thousands of smaller companies to a paying service, and fulfilled their social commitment.
As promised, Atlassian has all the pricing online, and only provides clearly marked discounts to academic institutions. Also, NGOs and Open Source initiatives can get Atlassian software for free. However, the complexity of enterprise software has crept in with matrices that give you the cost of, e.g., Jira: depending on the number of users, the software it will have, and the add-ons you will need.
Common to other SaaS providers, Atlassian offers two ways of licensing their products: 1) in their cloud with monthly fees, 2) in customer’s servers with perpetual server licenses. The first option is great if you want to avoid setting up and maintaining a server to host the product. The second option requires your own server, but comes with more flexibility as Atlassian provides the source code to the product under a development license.
For example, a Jira license for a server costs $1,200 for 25 users, and $16,000 for 2,000 users. The cloud option costs $100/month for 25 users, and $1,200/month for 2,000 users. However, the server options can take more than 10,000 users, while the cloud is limited to 2,000 users.
The killer business model: transparency + ease of use
Add-ons are available in the Atlassian Marketplace for both solutions. Atlassian has committed to the marketplace by doing two important things: 1) allowing easy access to the data hosted by Atlassian products to add-ons developers, and 2) making it easy for users to install and pay for these add-ons. Currently, there are about 1,700 add-ons. Add-ons provide from simple functionality for editing text to complex solutions for software testing management.
The resulting vibrant marketplace is an important way to provide additional value to its users.
Now, Atlassian has a billionaire valuation (3,3B$ one year ago), when it had a yearly revenue of 215M$. They have achieved these impressive numbers by a constant focus on improving their products, involving their users in the process through their open approach, and making it very easy for people to buy the software. Along the way, they had to fight conventional wisdom on enterprise software, and brought common sense to the table. Their software is still expensive, but cheaper by an order of magnitude to other enterprise software, as it’s expensive to develop products that can scale well and then to provide the support that companies need.
The information for this post has been compiled from lots of sources around the web. However, the one we found more interesting is the great talk that Scott Farquhar gave at the conference Business of Software in 2010. Other interesting links:http://techcrunch.com/2014/08/22/will-developer-tools-startups-ever-find-investors/