content
Rajeev Rajan, CTO at Atlassian, has a simple theory of developer performance: joy unlocks productivity.
While "sparking developer joy" may sound like a frivolous goal for one of the largest tech companies in the world, data shows that this approach makes developers significantly more productive.
By contrast, "motivating" tactics like quotas cause stress and pressure developers to take shortcuts — undermining team morale and the quality of your codebase.
Today, we'll cover how Atlassian helped developers surpass company goals by investing in joy.
What is "developer joy"?
The 2022 Software Developer Happiness Report identified the two biggest causes of dissatisfaction among developers:
Work/life balance
Lack of "quality work"
Developer joy means solving complex problems with great code in a healthy environment. If developers are creatively fulfilled, they are more likely to meet company goals and stick with you long-term.
Of course, not all aspects of traditional software development are inspiring. But inefficient workflows can dominate the workday, wasting time that could be spent experimenting and building.
Atlassian set out to eliminate inefficiencies and repurpose that time for creativity.
Defining metrics
As with any new initiative, it's crucial to establish baseline metrics and track changes over time. To ensure metrics aligned with developer experience, engineering asked teams: what prevents you from staying in creative flow states longer?
From there, engineering defined baseline metrics in the following areas:
Overall developer satisfaction
Time between commit and deploy
Pull request cycle time
Self-serve documentation
Self-serve dependency maintenance
As the initiative rolled out company-wide, engineering made strategic changes to tools, workflows, and culture — and measured their impact on baseline metrics.
Let's explore 3 ways that engineering successfully boosted developer joy and its natural side effect: productivity.
3 ways to spark developer joy
1. Tooling
It's no secret that long cycle times for deployment and pull requests decrease developer joy.
Lags in the development process make it difficult for developers to maintain creative momentum and excitement about projects. This creates a vicious cycle of low motivation, low productivity, and high turnover.
We recently explored this phenomenon at Meta. Engineering saw warning bells in their developer satisfaction metrics, so they targeted a top source of friction: inefficient review cycles. Meta introduced a Machine Learning feature that cut time between reviews, dramatically increasing productivity and developer satisfaction.
So what did Atlassian do to reduce friction in their tooling?
Introduce new automations in pull reviews
Migrate to new test libraries
Mock external services
As a result of these efforts, Atlassian is on track to reduce deployment cycles for shared services by 90%.
2. Documentation
At the same time, engineering realized that improving tools was only half the battle. Developers also needed accurate, accessible documentation to help them build features efficiently.
Atlassian products use hundreds of shared microservices and libraries. In an ideal workflow, developers would be able to locate the right docs and get up to speed on the relevant components quickly. In practice, developers could waste hours or days finding the information they needed — especially as Atlassian moved towards a distributed workforce.
Scattered, incomplete documentation is the enemy of developer joy. It disrupts creative flow as developers search for answers. Gaps in understanding can also weaken code through shortcuts and bugs, causing expensive delays in development.
Atlassian reckoned with these consequences by investing in a central knowledge base. Having one home for docs makes it easy to find information like:
Owners of software components
Changelogs
Health-check metrics
Quick access to the right docs has helped developers build faster while making stronger technical decisions. It has also freed them up to spend less time searching for answers, and more time innovating.
3. Autonomy
At the start of the developer joy initiative, Atlassian asked developers: what prevents you from staying in creative flow states longer?
One of the biggest takeaways from these conversations was that every team faces a unique set of challenges. There are common sources of friction across the org, such as inefficiencies in tooling and documentation. However, macro improvements can't address team-specific obstacles to better, more satisfying work.
Engineering leadership recognized that developers are the experts on what's holding them back. So, Atlassian asked developers to allocate 10% of their time for reducing barriers to happier, more productive workdays.
Since Atlassian gave teams more autonomy over their roadmaps, there have been measurable improvements to developer productivity. For example:
The Confluence team worked on restructuring repositories, improving test efficiency, and automating the deployment pipeline. Now, 31% of pull requests reach production within 48 hours (up from 12%).
The Trello team eliminated 24,000 lines of dead code and made adjustments to their branching model. This led to a significant reduction in code freezes.
A central services team automated approvals for low-risk changes, increasing deployment frequency by 300%.
Atlassian's takeaways for your team
Atlassian's multi-pronged approach to developer joy shows that there is no silver bullet for happier, more productive teams. The most effective productivity initiatives are rooted in a deep understanding of what works for your developers.
Sparking developer joy is about:
Understanding where your team experiences friction
Making strategic adjustments to remove friction
Iterating based on clear metrics
Not all meaningful changes will be technical. Atlassian found success with a wide range of strategic adjustments, ranging from cultural (increased team autonomy over the roadmap) to technical (automations in the pull request process).
However, Atlassian's results show the power of technical solutions to boost developer joy — when paired with improvements to culture and process. In particular, developers can leverage Machine Learning and AI to introduce new automations, reducing friction in deployment and pull request cycles.
Not every developer needs to be an AI expert. But if your team lacks foundational skills, you'll miss opportunities to improve workflows and fuel creativity.
When you equip developers with AI skills — and give them the autonomy to optimize processes — joy and productivity grow in tandem.
Free Resources
