Severity vs. Priority (2024)

Community Leader

There's a historical essay to be had here, but the short version is "Severity is a bad thing to allow users to set (as is priority) so Mike and Scott dropped the concept when creating Jira, because priority was enough for them" - that's why it has no system field. Of course, those of us who need it just add a custom field.

You are right in your definitions I reckon - severity = how bad it is, priority = when you're planning work, how urgently this should be looked at.

You are also correct about the matter of organisation terms - some organisations will do 1-5 on severity, others 1-3, use different terms and so-on. And the same for priority.

Some of us get a little more automatic - we define a severity (from minor cosmetic issue through to system unusable) and an impact (ranges of numbers of people or systems affected) and calculate a priority from it.

The issue is pretty simple when tracking software development issues:

Priorities change, but the Severity of an issue never does.

A bug that causes data loss would be of the highest severity. What it's priority is depends on the phase of the development life cycle. A Priority 2 bug might become a priority 1 bug later on.

An typo in the EULA might be of the highest priority to the business, but of the lowest impact to the end user.

Severity is objective, priority is subjective.

Another way to think about the two is urgency (priority) and impact (severity). You could also use risk instead of impact.

The problem with the default Jira scheme is that someone might ask, 'why are we working on minor issues now (when there are blockers and major)?' If you have a urgency/priority, it's easy to create a pipeline.