Problem Management

What is Problem Management?

Problem management is the IT service management practice of finding and removing the underlying causes of incidents and managing workarounds, so that incidents stop recurring.

Updated November 2026 · 3 min read


Problem management is the IT service management practice of finding and removing the underlying causes of incidents, so that they stop recurring or happen less often, and of managing workarounds until a permanent fix is in place. Where incident management aims to restore service as quickly as possible, problem management asks why the incident happened and how to prevent it. It turns a cycle of firefighting into steady improvement in reliability.

Problem management is one of the core practices in ITIL and in the ISO/IEC 20000 standard for service management. Its ideas overlap with root cause analysis in quality management and with blameless postmortems in site reliability engineering.

Key Terms

TermMeaning
IncidentAn unplanned interruption or reduction in the quality of a service
ProblemThe cause, or potential cause, of one or more incidents
WorkaroundA way to reduce or eliminate the impact of a problem until it is fixed, such as restarting a service when memory runs low
Known errorA problem whose cause has been analyzed but which has not been resolved
Known error database (KEDB)A record of known errors and their workarounds, used by the service desk to resolve incidents faster

Reactive and Proactive Problem Management

  • Reactive problem management starts from incidents that have already happened, such as a major outage or a pattern of similar tickets.
  • Proactive problem management looks for problems before they cause incidents, using monitoring trends, vendor advisories, capacity data and lessons from other teams.

The Problem Management Process

ITIL 4 describes three phases:

  1. Problem identification. Spotting problems through trend analysis of incidents, major incident reviews, recurring tickets, monitoring and supplier information.
  2. Problem control. Analyzing the problem, prioritizing it by risk, finding the root cause and documenting workarounds. Common techniques include the five whys, fishbone diagrams, timeline analysis, fault tree analysis and Kepner-Tregoe problem analysis.
  3. Error control. Managing known errors, assessing possible permanent fixes and raising changes to implement them, usually through the organization's change process.

Not every problem is worth fixing. Some are cheaper to live with through a workaround, and problem management includes deciding which to fix based on cost, risk and impact.

Why It Matters

  • Fewer repeat incidents and less time spent fixing the same thing again.
  • Faster incident resolution, because the service desk can apply documented workarounds.
  • Better use of engineering time, by focusing fixes on the causes with the most impact.
  • Improved reliability over time, rather than a plateau of constant firefighting.

An Example

Consider a hypothetical online travel company that logs 46 incidents in a quarter where the booking site slows down in the early evening. Each is resolved by restarting application servers, and each is closed. A problem analyst reviewing incident trends groups them into one problem. The timeline shows the slowdowns coincide with a nightly data export that started running two hours earlier after a time zone change. The workaround, restarting servers, goes into the known error database so the service desk can apply it in minutes. The permanent fix moves the export to a read replica of the database. Evening slowdown incidents fall to two the following quarter.

Common Pitfalls

  • Closing incidents without ever opening problems, so the same issues recur.
  • Stopping at the first cause found, such as "human error," instead of asking why the error was possible.
  • Problems that are opened but never prioritized or resolved.
  • Blame, which makes people hide information and makes causes harder to find.

Problem management is part of IT service management and works closely with incident management and blameless postmortems. Its effect shows up in measures such as MTTR and incident volume.