Nobody in town could remember who built it. That was the first red flag.
The Atlassian instance at this large retail organization had been created over the course of many years, passed from admin to admin. Each one had added their own contributions, stitched on their own patch or added an arm, and then walked away. By the time we opened it up, it wasn’t a platform anymore. It was a creature. It was aliiiiive! 🌩️ đź’€
What the Health Check uncovered
548,000 issues. 531 custom fields. 94 issue types. 240 active statuses in the Jira and Confluence kingdom of this retail giant.
548k issues might mean they’re getting a lot of work done, but 240 active statuses means there’s nothing consistent about that work, so how can they measure it?
And then: the detail that made our consultant set down their coffee and stare at the screen: Aa single automation rule, “close the epic when all stories are done,” that had been copied and pasted 18 times across different projects. Not referenced. Not shared. Duplicated. Eighteen identical copies of the same logic, each one a separate point of failure, each one maintained (or more accurately, not maintained) independently.
In another corner of the instance, we found 73% of every single issue in the system was set to “Major” priority. ⚠️ Not because everything was major. Because when everything is major, nothing is. Priority had become meaningless, a vestigial organ on a creature that had grown far beyond what anyone originally designed.
The permission schemes were even worse. 40 out of 83 weren’t attached to any project at all. They just… existed. Floating in the configuration like ghosts in the walls, how spooky! đź‘»
The Solution
The teams using this system had learned to work around the 240 statuses and they stopped trusting priority fields years ago. They’d accepted that “this is just how Jira works” the way someone living in a haunted house eventually stops noticing creak in the attic.
Our health check didn’t just document the creature. It gave the team a surgical plan: which fields to consolidate, which schemes to remove, which automations to centralize, and which statuses could be retired without breaking active workflows. Then we worked on the action plan together, and prioritized which life and death procedures we could focus on first.
The monster didn’t need to be killed. It just needed to be understood. đź’š
What’s lurking in your instance? A health check will show you.