Network shutdowns are rarely a surprise in hindsight. They cluster around elections, verdicts, protests and exam periods, and they follow a pattern: platform throttling first, then mobile data, then fixed lines, with international transit last to go.
Before
- Decide now which single channel is authoritative during an outage, and tell your audience where it is.
- Mirror critical content somewhere outside the jurisdiction, and test that the mirror actually serves.
- Distribute an offline copy of the contact tree. Phone numbers on paper are not a joke.
- Pre-write the holding statement. You will not want to draft it under pressure.
- Make sure at least two people can publish. Single-key access is a continuity risk, not a security control.
During
Assume the connection you do have is monitored and unreliable. Prioritise ruthlessly: one accurate update reaching people matters more than five channels half-updated.
- Publish short. Low-bandwidth conditions punish heavy pages badly.
- Use SMS and radio partners you arranged in advance; new arrangements will not come together now.
- Log what happened and when, because timestamps become evidence later.
- Do not rotate every credential at once over a degraded link. You will lock yourselves out.
The organisations that come through a shutdown well are not the ones with the best tools. They are the ones who agreed on the plan while the internet still worked.
After
Restoration is the window where intrusions get missed, because everyone is busy catching up. Check what changed while you were dark: new accounts, altered DNS, content edits, unfamiliar sessions. Then write down what you wish you had prepared, and prepare it.