ITIL Problem Management: Root Cause Analysis, Known Errors and Workarounds

Incident management gets users working again; problem management stops the same incident returning. It reduces the likelihood and impact of incidents by identifying actual and potential causes and managing workarounds and known errors. Without it, a support team spends its days fixing the same faults again and again.

Key Terms

TermMeaning
ProblemA cause, or potential cause, of one or more incidents
WorkaroundA solution that reduces or eliminates the impact of an incident or problem for which a full resolution is not yet available
Known errorA problem that has been analysed but not resolved, with its cause and a workaround documented
Root causeThe underlying reason that, if removed, prevents recurrence

Incident vs Problem

The difference is easiest to see with an example. Fifty users cannot log in because an authentication server ran out of disk space. The incident is “users cannot log in”; clearing the disk resolves it. The problem is “why did the disk fill up?”, perhaps because logs are never rotated. Fixing the problem means changing the log settings so it cannot happen again.

The Three Phases in ITIL 4

  1. Problem identification: spot problems from incident trends, repeated incidents, supplier reports, monitoring data and post-incident reviews. Regularly review incident data for patterns.
  2. Problem control: analyse the problem, find the cause where possible and document a workaround. Once a workaround is documented the problem becomes a known error.
  3. Error control: manage known errors, decide whether to fix them and, when approved, raise a change to remove the cause. Known errors are shared with the service desk so agents can use the workaround.

Root Cause Analysis Techniques

  • 5 Whys: ask “why?” repeatedly until you reach a cause you can act on. Quick and effective for simple chains.
  • Ishikawa (fishbone) diagram: group possible causes under headings such as people, process, technology and suppliers.
  • Pareto analysis: find the few causes behind most incidents; roughly 80% of effects often come from 20% of causes.
  • Timeline analysis: list events and changes before the incident to spot what changed.
  • Kepner-Tregoe: a structured method that compares what is and is not affected to narrow down causes.

Worked Example: 5 Whys

QuestionAnswer
Why did the payment page fail?The application server stopped responding
Why did it stop responding?It ran out of memory
Why did it run out of memory?A new release leaked memory
Why was the leak not found?Load testing did not run for the release
Why did it not run?The test step was skipped to meet the deadline and the release checklist did not make it mandatory

The fix is not “be more careful” but a change to the release checklist and the pipeline. Good problem management ends with a concrete change, not a lecture.

Reactive and Proactive Problem Management

Reactive work responds to incidents that have already happened. Proactive work looks for risks before they cause incidents: trend analysis, review of monitoring alerts, supplier bulletins and lessons from other organisations. Mature teams spend an increasing share of their effort on proactive work.

Practical Tips

  • Do not open a problem for every incident; focus on repeated, high-impact or unexplained ones.
  • Give every problem an owner and a review date.
  • Make known errors visible to agents, ideally linked from the knowledge base (see Knowledge Management).
  • Measure the number of repeat incidents and the time known errors stay open.
  • Keep it blameless: the aim is to understand the system, not to find someone to punish.

Common Mistakes

  • Stopping at the workaround and never fixing the cause.
  • Treating problem management as a delay to incident closure.
  • Producing long analysis reports without a change to act on.
  • Blaming individuals, which makes people hide information.
Key takeaways
  • A problem is the cause of one or more incidents; problem management reduces recurrence.
  • A known error is an analysed problem with a documented workaround.
  • Use techniques such as 5 Whys, Ishikawa and Pareto to find root causes.
  • Good analysis ends in a change, and the process works best when it is blameless.

Related Lessons