When testers find defects, someone must decide what to fix first. Two separate ratings help:
- Severity: how bad the impact is when it happens. Critical: a core function fails, data is lost or wrong, no workaround. High: a major function fails, with a hard workaround. Medium: a function partly fails, with a reasonable workaround. Low: cosmetic.
- Priority: how soon it must be fixed, given the business situation.
They usually move together, but not always. A spelling mistake in the company name on the home page is low severity, high priority: nothing breaks, but it embarrasses the brand on day one. A crash in a yearly report run in December is high severity, low priority in March.
In triage, the BA, the business owner, the test lead and a developer go through new defects together. The BA's job is to explain the business impact, so priority reflects it.