Silence can become a second service failure
A government service goes offline near an application deadline. The technical team is investigating, but users do not know whether submissions were received or whether they should try again. Communication needs its own operational timeline alongside restoration.
Distinguish confirmed facts from unknowns
An initial notice should explain what is known, what is affected and what remains under investigation. Do not invent a restoration time or imply full containment without evidence.
Create communication triggers
Significant changes in service impact, safety, user action or available alternatives should trigger review of public messaging. Use authorised incident procedures and applicable obligations.
Keep a timestamped decision record
Document when impact was detected, when information was verified, who approved communications and when messages were released. This supports later review without pretending every delay is avoidable.
Coordinate channels
Websites, call centres, frontline teams and other authorised channels should provide consistent practical instructions. Conflicting notices can multiply harm.
Protect affected users
Explain available alternatives and what users should do about pending transactions or deadlines where an authorised answer exists. Privacy and security remain important.
Update and correct openly
When facts change, explain the correction and identify which earlier advice is superseded. Acknowledging uncertainty is better than allowing an inaccurate notice to persist.
Break-test: the service returns but records are incomplete
Can the agency distinguish restored access from fully reconciled applications and tell affected users what action is required?
Connected knowledge
See Data Incident Notification, Service Recovery and Public Service Division.