Change Management Approach
Change Management Approach refers to how changes to systems — including code, infrastructure, and configurations — are assessed, approved, and deployed.
In high-performing DevOps organisations, change management is lightweight, decentralised, and integrated into team workflows to enable speed and safety without bureaucratic friction.
Level 1 – Initial (Ad Hoc)
Change management is reactive, manual, and often a bottleneck.
Changes may lack documentation, formal approval, or traceability.
- Teams operate without structured change processes
- Approvals are informal or based on individual judgment
- Risk assessment is inconsistent or absent
- Change failures are frequent and poorly managed
- Change freezes or emergency patches are common
Level 2 – Managed (Emerging Practice)
Formal processes emerge, often centralised through change advisory boards (CABs).
These processes add consistency but also introduce delays and overhead.
- Changes must go through a review board or central process
- Documentation and scheduling are required for every change
- Approvals are based on role or seniority rather than system knowledge
- Teams follow the process but may see it as a blocker
- Audit and compliance improve, but speed suffers
Level 3 – Defined (Standardised)
Change management is decentralised and embedded in delivery pipelines.
Low-risk changes can be deployed automatically with appropriate controls and accountability.
- Peer reviews, automated testing, and CI/CD pipelines serve as approval mechanisms
- High-risk changes are routed through appropriate escalation paths
- Change risk is assessed based on impact, not policy alone
- All changes are traceable, version-controlled, and auditable
- Standard operating procedures exist for emergency and rollback handling
Level 4 – Quantitatively Managed (Measured & Controlled)
Change performance is measured and optimised based on risk, speed, and quality.
Policies are dynamic and informed by system data and delivery metrics.
- Metrics include change failure rate, time to deploy, and approval latency
- Low-risk changes are auto-approved and deployed frequently
- High-risk changes are reviewed with data-informed controls
- CABs (if they exist) focus on systemic risk, not routine changes
- Continuous improvement cycles reduce risk over time
Level 5 – Optimising (Continuous Improvement)
Change management is a strategic enabler of speed and resilience.
The system self-regulates based on real-time signals, and learning from change outcomes drives systemic improvements.
- Change policies adapt dynamically based on historical performance and system health
- Teams use predictive risk models or machine learning to assess change impact
- Post-deployment signals (e.g. error spikes, alert noise) feed back into risk models
- Change management is invisible to teams unless intervention is needed
- Rapid, safe change is a competitive advantage — not a governance burden