L1, L2 and L3 in brief

LevelTypical roleExamples
L1Intake, standard checks and communicationLocked account, guidance, context collection
L2Technical analysis and known remediationLogs, configuration, data and integration
L3Product or architecture changeCode, databases, APIs and performance

In a small team, one person may cover several levels, while workflow and required information remain distinct.

What a useful incident contains

Priority is not set by the number of exclamation marks. It follows impact, affected users, workaround availability and risk to data or operations.

  • time and environment where it occurred;
  • user, role and reproduction steps;
  • expected and observed result;
  • screenshots or identifiers without exposing sensitive data;
  • impact and available workaround;
  • recent application or infrastructure changes.

Incidents, problems and change requests

An incident restores service. A problem investigates a recurring cause. A change request adds or modifies behavior. When all use one unclassified queue, urgent work interrupts development and causes remain untreated.

A shared backlog can remain transparent when every item has a type, impact, owner, acceptance criterion and priority decision.

What to prepare for a support supplier

  • application and dependency inventory;
  • controlled access and emergency procedure;
  • environments, deployment, backup and restore;
  • existing monitoring, logs and alerts;
  • business owners who can confirm impact;
  • third-party suppliers and responsibility boundaries;
  • incident history and known risks.

How to measure support quality

Response time alone does not show whether an issue was understood or resolved. Track time to restore, recurrence, incidents prevented through maintenance, failed changes, monitoring coverage and communication quality. For a small application, a few consistent indicators are more useful than a crowded dashboard.

Do you need to define or take over application support?

We can audit the current state, separate intervention levels and propose a workflow proportional to application impact.

Discuss your project

All articles