Problem identification criteria
When to raise a problem versus keep work on the incident — recurring cases, significant single events, suspected causes, known error candidates, and proactive trends.
Identification criteriaProblem Management
Problem management identifies root causes behind recurring incidents, reducing repeat disruption and improving long-term service stability.
Don Francis
Problem Manager
As a Problem Manager, I focus on stopping repeat incidents by understanding what really went wrong, not just closing tickets. I prioritise investigations, coordinate technical and business stakeholders, and make sure root causes are addressed so the same failure does not keep coming back. I also maintain known errors and workarounds so service can stay stable while lasting fixes are delivered.
Problem management in ITIL 4 is the practice of reducing the likelihood and impact of incidents by identifying root causes, managing workarounds, and initiating actions to remove those causes over time.
The goal of problem management is to:
- eliminate recurring incidents and reduce incident volume
- improve service reliability and customer experience
- make root cause and workaround information visible and usable
- feed learning into change and improvement activity
Guidance and examples
Problem investigation and known errors
Structured problem records, clear identification criteria, and proportionate root cause analysis reduce repeat outages and shorten recovery when issues do recur. Use the guidance below for process decisions, techniques, and worked examples that connect incident data, change, and service improvement.
Root cause techniques
When to use 5 Whys, Ishikawa (fishbone), Pareto, chronological analysis, and situation appraisal — plus what good looks like and common pitfalls.
Root cause techniquesKnown error example
A concrete known error record: intermittent SSO failures, documented cause, service desk workaround, and the linked change for a lasting fix.
Known error exampleInvestigation example
End-to-end problem investigation for recurring application timeouts — from trend identification through analysis to known error and change.
Investigation example
