1
Define the scope and change categories
Start by listing every type of change your organization needs to govern β IT systems, operational processes, organizational structure, product configuration. Then group them into tiers (standard, normal, emergency) with clear risk-score thresholds for each.
π‘ Limit the initial scope to your highest-risk change domains. You can expand scope in a future revision once the procedure is embedded in daily operations.
2
Name the roles by specific title, not function
Replace generic labels like 'Change Owner' or 'Approver' with actual job titles or named individuals. Assign the CAB membership to specific roles with named backups for each seat.
π‘ An unsigned or unassigned responsibility matrix is the single most common reason change procedures fail at the first real incident.
3
Build the change request form fields
List every field required on a change request β description, justification, impact assessment, affected systems, proposed window, rollback plan. Flag which fields are mandatory versus optional so requestors know what will cause a rejection.
π‘ Pilot the form on three real changes before publishing. You will discover missing fields and unnecessarily complex ones in the first real submissions.
4
Set review and approval timelines as SLAs
Assign a maximum calendar-day response time to each step: initial screening, CAB review, final approval, and scheduling. Record these in the procedure and monitor them monthly.
π‘ Emergency changes need a separate, accelerated SLA β typically 4 hours for initial approval β with retroactive documentation requirements stated explicitly.
5
Document the implementation and rollback steps
For each normal and emergency change type, outline the sequence of implementation actions and the specific trigger conditions that initiate a rollback. Confirm that rollback steps can be executed by someone other than the original implementer.
π‘ A rollback plan that only the original engineer can execute is not a rollback plan β it is a dependency risk.
6
Specify the post-implementation review requirements
State who completes the PIR, within how many business days of closure, and what fields must be filled in. Include a mandatory section for lessons learned even when the change succeeded.
π‘ Link PIR findings to your standard-change register β a change type with three consecutive successful PIRs is a strong candidate for reclassification as standard.
7
Establish the change log structure and access rules
Define which system holds the authoritative change log, who can view and edit records, the minimum retention period, and the reporting cadence for leadership or compliance review.
π‘ Export the change log to a read-only format before any audit β auditors should never have direct write access to live records.
8
Obtain sign-off and distribute to all affected teams
Have the procedure reviewed by the CAB chair, IT leadership, and compliance before publishing. Distribute it to all stakeholders named in the roles section and store the signed copy in a version-controlled repository.
π‘ Schedule a mandatory re-review every 12 months β or immediately after any major change-related incident β and update the version number and effective date each time.