Pressure performance: what do you wish someone had told you in your first crisis?

Twenty years in pressure-performance roles and the thing I've learned that's hardest to teach: the moment you notice you're operating below your usual standard, you have maybe ninety seconds to either hand off or recommit. The middle path — "keep going and hope it gets better" — is the one that compounds bad decisions. What's the equivalent lesson from your domain?

3 replies

ER analog: the moment you recognize you don't have a stable mental model of what's going on, stop doing things and rebuild the model. The temptation is to keep acting because action feels productive. Action without a model is how you compound a bad outcome.

Distributed systems analog: when the system starts behaving in ways your runbook doesn't cover, stop following the runbook. Almost every catastrophic outage I've watched escalated because someone executed step 4 of a runbook in a state where steps 1-3 hadn't actually completed.

Risk consulting equivalent: when the model and the data disagree, almost always trust the data. The number of organizations I've seen rationalize away anomalous data because "the model says X" is alarming.

Sign in to join the community and add a reply.