The Five Places Your Organization Loses Time in a Crisis

When a crisis goes badly, the post-mortem blames the decision. Wrong call, wrong person, wrong moment. But across 15,000+ incidents, the decision is rarely the problem — the time it took to get there is. Time doesn't bleed out in one place. It bleeds out in five. Deepwater Horizon lost it in all five — and had almost six hours of warning.

1. Recognition Lag — the signal arrives, no one acts on it

The threat almost always announces itself first: the alarm, the anomaly, the pressure test that "could have an innocent explanation." It dies in alarm fatigue, normalized deviance ("we see this all the time, it's never a problem"), and weak signals hidden in noise. On the rig, the first anomaly to serious concern took over an hour. Target: under 5 minutes. Close it with automated detection, trained pattern recognition, and a just culture that makes it safe to say "something seems off."

2. Assessment Lag — you see it, but can't agree what it means

Detection isn't understanding. Teams burn time on ambiguity, divided expertise, and the fatal urge to seek certainty — "let's gather more data." On Deepwater Horizon, assessment ran five-plus hours with no decisive conclusion. Target: under 2 minutes for known patterns. Close it with pre-defined severity criteria (green / yellow / red), time-boxed assessment, and the 80% Rule — act on 80% confidence, not 100%.

3. Decision Lag — the call waits on authority

This is where hierarchy becomes latency. The person who sees the problem often can't act on it, and the person who can act isn't in the room. Add risk aversion and no standing protocol, and the decision doesn't get made — it gets forced by the event. The single highest-leverage fix in the entire system: Pre-Authorized Response Protocols. Decide in advance, under calm conditions: "If pressure rises above X in Y minutes, stop operations — no further approval needed." Target: under 1 minute for pre-authorized scenarios.

4. Execution Lag — the decision doesn't travel or doesn't move

A correct decision that reaches the field late is a wrong decision. Time dies in communication failures, coordination handoffs, and resources that aren't staged. Close it with Closed-Loop Communication (command → acknowledge → confirm, the fireground model), pre-positioned resources, parallel execution, and training to automaticity. Target: under 2 minutes for simple actions.

5. Adaptation Lag — Plan A fails and you cling to it

Your first response won't always work. How long until you admit it and pivot? Commitment bias and the absence of a Plan B turn minutes into days — Deepwater Horizon took 87 days from blowout to control. Close it with pre-defined failure criteria ("if we don't see X by Y, Plan A is dead"), a parallel Plan B, and a culture where pivoting is smart, not weak. Target: under 30 minutes.

The point

Add the five up and you have your total latency — the real gap between "something's wrong" and "it's handled." Most organizations have never measured one of them. That's not a discipline problem. It's a measurement problem, and it's fixable. Had Deepwater Horizon halved each lag, they would have shut down by 5:00 PM. The blowout would never have happened. Eleven men would still be alive.