Before you start
Backlog Health Checker
Find out how many sprints of your backlog are actually ready to work on, and what the rest needs before refinement.
Open a backlog export from Jira, Azure DevOps or a similar tool, in backlog order, with recently finished work included. It is read inside your browser; nothing is uploaded.
Backlog export
Drop a spreadsheet here
A Jira, Azure DevOps or similar export in backlog order: key, title, type, status and estimate, with the parent epic, acceptance criteria and created date if you have them. Include recently finished work, and the pace is read from it.
A long backlog is not a ready one
Two hundred items can hold less than a sprint of work a team could start tomorrow. The rest has no estimate, no acceptance criteria, belongs to no epic, is five times the size of anything the team has finished, or is titled “Fix login bug”. The gaps near the top matter most, because those are the items the next planning meeting will reach.
What is checked
Each open item against a definition of ready: an estimate, acceptance criteria in their own column or written into the description, a parent epic or feature, a size the team usually finishes, and a title that says what the work is. Pairs of titles close enough to be the same work written twice, and items old enough to question, are listed as well.
Sprints of ready work
The team’s pace is read from the finished work in the same export, sprint by sprint, or you can give it. The ready work divided by that pace is the number of sprints the backlog can feed, and the backlog is marked where each sprint would end.
What this does not do
It checks that the pieces of ready are present, not whether the criteria are good or the estimate is right. It does not reorder the backlog or split stories. For vague wording inside the criteria themselves, the Scope Language Checker reads the text.