Software Testing

Beginner path · lesson 24 of 35

BeginnerFundamentals4 min readPublished November 12, 2026

Severity vs priority: understanding bug classification

Severity is how much damage a bug does. Priority is how soon it should be fixed. They often match, but not always, and the cases where they don't are the ones that matter.

On this page

Key takeaways

  • Severity = the impact of a bug on the system or its users. It's mostly a technical judgement.
  • Priority = how urgently it should be fixed. It's a business decision.
  • A bug can be severe but not urgent, or trivial but urgent. Keeping the two apart leads to better decisions.

Two different questions

When a bug is reported, the team needs to answer two questions:

  • How bad is it? That's severity.
  • How soon do we fix it? That's priority.

They sound like the same question, and often the answers match. A crash in checkout is both severe and urgent. But the interesting cases are the ones where they differ.

Severity: how bad is the impact?

The ISTQB glossary defines severity as the degree of impact a defect has on the development or operation of a system. Testers usually set it when they report the bug.

A common scale:

SeverityMeaningExample
CriticalSystem down, data lost or corrupted, security breach, no workaround.Payments are charged twice.
MajorImportant feature broken; a workaround may exist.Invoices can't be downloaded as PDF, but can be emailed.
MinorSomething works wrongly, with little impact.Sorting by date puts today's items last.
TrivialCosmetic.Two buttons are slightly misaligned.

Priority: how soon do we fix it?

Priority is the order in which bugs get fixed. It's usually set by the product owner or during triage, a meeting where the team reviews new bugs.

Priority depends on things testers may not know: business deadlines, how many customers are affected, contracts, upcoming demos, reputation.

A common scale is P1 (fix now) to P4 (fix when there's time).

The four combinations

High priorityLow priority
High severityCheckout crashes for every customer. Fix now.An old export tool crashes, but only two internal users use it, once a year, and there's a workaround.
Low severityThe company name is misspelled on the home page. Trivial technically, embarrassing publicly.A typo in a rarely seen settings tooltip.

The two cases on the diagonal are easy. The other two are where teams argue, and where keeping the terms separate helps.

Real-world exampleFriday afternoon triage

An hour before a big product launch, the team has three new bugs and time to fix one.

  • Bug A: an admin report crashes when filtered by a date range longer than 5 years. Severity: major. Nobody needs that report until next quarter.
  • Bug B: the launch banner on the home page says "Lanch offer". Severity: trivial. It's the first thing every visitor sees today.
  • Bug C: the "Remember me" checkbox on login doesn't remember. Severity: minor.

The product owner picks Bug B as P1. Bug A becomes P2 for next week. Bug C goes to the backlog.

Lesson: the most severe bug wasn't the most urgent. The tester's job was to state the severity accurately. The business decided the priority, with full information.

Getting it right as a tester

  • Set severity honestly. Don't inflate it to get your bug fixed faster. Teams quickly stop trusting a tester for whom everything is "critical".
  • Explain the impact. "Affects all users on Safari, about 30% of our traffic" helps the product owner set priority.
  • Suggest, don't decide, priority. Give your view, then respect the business decision.
  • Re-evaluate. A minor bug can become urgent if a big customer complains, or before a demo.

Do

"Severity: major. Users can't change their delivery address after ordering. Workaround: support can change it manually. About 40 requests a week, based on support tickets."

Don't

"Severity: CRITICAL!!! This is really bad, fix immediately." No facts to support the decision, and an inflated label.

Check yourself

The app crashes when a user uploads a profile picture larger than 50 MB. The upload limit shown to users is 5 MB, and almost no one tries larger files. High or low severity? High or low priority?

Show answer

Severity: fairly high, since a crash is a serious failure. Priority: probably low, because it's rare, users are told the limit and the impact is limited to one action. The product owner might still raise it if crashes affect stability for everyone, so give them that information.