The pattern that makes it worse
Minute 0: Discord panic. Minute 12: a vague "we are aware" tweet. Minute 90: silence. Minute 180: reporters fill the gap with screenshots you never confirmed.
Timed sequence
0–15 min — Confirm internally
- Is it production?
- Approximate scope (funds, chains, contracts).
- Who owns the next public line?
15–30 min — Holding line (facts only) Template shape:
We are investigating a suspected incident affecting [system]. User funds status: [known / unknown]. Next update by [time + timezone].
No blame. No "hackers." No "under control" unless true.
30–60 min — Contain + correct
- What users should do (revoke, pause, wait).
- What you paused.
- When the next update lands — then hit that time.
Crypto-specific notes
- Explorers and dashboards move faster than your PR queue. Assume screenshots are public.
- Separating "bridge / vault / front-end" in language prevents wrong contagion to the whole brand.
Desk rule
Speed without a clock is chaos. Every statement needs a next update time.