Content Design for Atlassian’s AI Rule Creation Feature
⏱️
Context
This feature was created for Atlassian Automation. Automation is available in Jira and Confluence, two of Atlassian’s most popular products. For context, Automation helps users to focus on what matters. It removes the need for users to perform repetitive manual work, by allowing them to create rules which will automatically perform the tasks on their behalf. Automation rules can carry out tasks such as sending emails, commenting on pages, transitioning work and more.
🧩
Problem Space
At the beginning of 2023, many users were starting the rule creation process in Automation, but not all of these rules were successfully enabled and used. This was due to issues with the UX and UI of the Automation rulebuilder, which caused a steep learning curve for new users, and contributed to a lack of guardrails when creating rules. To solve these issues, we believed we could use AI could take on the heavy lifting of rule creation. With AI, users could simply input a natural language prompt (NLP) and a rule will be created on their behalf - removing the need for users to always rely on manual rule creation.
💭
What were my tasks?
As the lead content designer and UX writer for this project, I was tasked with the following:
Write UX copy - including writing descriptions, naming features, adding example prompts and finalising success/error messaging.
Write and publish support documentation - creating help content to be published on Atlassian’s support site, release notes and community posts.
Contribute to Atlassian’s AI content guidelines - refine them and shape them in their early days of development.
UX Copywriting
When writing the copy for the AI rule creation feature, which you can see the gallery of images, I ensured the feature:
Contained simple and concise instructions that promote the use of AI for anyone. I tried to stay away from confusing, technical jargon unless absolutely necessary.
Provided support and assistance, every step of the way - both within the prompt modal on the landing page, and in the Automation rulebuilder, accessible by selecting the info or question mark icon. The user should always have access to help if they need it.
Contained the necessary labels, disclaimers and warnings required - to let users know the feature was in its Beta stage (which was the case when it was first released in 2024), and that the content generated by AI may vary in its output quality.
In-Product Support Content
When writing short-form support content that would appear within the UI itself, I made sure to:
Include info that was most relevant to the rule creation process, saving more menial or long-form content for the external documentation.
Write with clear hierarchy and use dot points for succinctness and clarity.
Get the point across as clearly as I could.
External Support Documentation
In my documentation for AI rule creation in Automation, which I wrote to ensure users had more detailed support if they needed it, I ensured I had:
Included the required disclaimer aligning with Atlassian’s AI content guidelines.
Provided clear instructions on how to use the AI feature, with step-by-step overviews.
Placed the article in a topic set with high visibility for easier access for users.
✅
Results & Impact
Once AI rule creation for Jira and Confluence was publicly released in 2024, we observed the following results, showing the positive impact of this project:
An 18% increase in successful rule creation in the following months after its release (meaning rules were created, AND successfully enabled for use)
Use of the AI rule creation feature contributed to AI adoption company-wide.
AI rule creation was highlighted at Atlassian’s largest conference, TEAM, in the 2024 AI announcement by company’s CEO and founder.