Before: Panic-Fueled Crisis Response. After: Cool-Headed Rational Contingency Planning. Here’s How.

Years ago, I sat in a client’s conference room at 3 p.m. It felt like the eye of a hurricane. Nobody answered emails. Nobody rallied the troops. Everybody was waiting on the CISO, a sharp man who had not slept in 36 hours and was running on anger, adrenaline and terrible coffee.

Then the phone rang. Another alarm, another escalation. He walked in red-eyed, told the CEO that the incident scope had doubled and asked for strategic guidance. The CEO did what most leaders do in that moment. He turned to the room and asked a question I’ve heard a hundred times: “So. What do we do now?”

There’s no elegant answer to that question once the storm is already overhead. You don’t rise to the occasion. You sink to the level of your training. What you do in a crisis is limited by what you bothered to decide, in advance, during the calm.

That’s the difference between a panic-fueled crisis response and one grounded in rational contingency planning. Let’s unpack what that actually means.

The “Before”: What Most Crisis Management Really Looks Like

Most companies have a document called a “crisis management plan.” It usually lives on a shared drive nobody remembers the password to. It was probably written by an overwhelmed intern three jobs ago, approved by a vice president who skimmed it, and never touched again.

That document gives you a page and a half about communication protocols, a list of phone numbers, and a section on “public statements” that reads like generic corporate fluff. It identifies four or five tidy scenarios: fire, data breach, workplace injury, product recall. But the crisis that actually hits you—the weird, folded, multi-system failure—will not look like any of those neat boxes. A real crisis is a system failure, an unplanned interaction between people, technology, processes and external pressure.

So leaders improvise. They jump on a call, look at the same ambiguous information, and push for immediate action because action feels better than uncertainty. That is exactly the wrong instinct.

Brace yourself, because here’s the kicker: instinct is the enemy of protocol. In a real emergency, your instincts will tell you to do the first thing that comes to mind. They will encourage you to reassure everyone, to defend the company’s reputation, to hide your uncertainty. Those impulses feel protective, but they also steamroll rational analysis. The result is response without assessment, action without authorization, and ultimately leadership by whoever can shout the loudest on the bridge.

I’ve watched leaders make costly decisions in that state—decisions they later had to walk back. They open investigations too early, admit fault too broadly, or shut down legitimate communications because they fear disclosure. What would have helped? Not a binder. A pre-committed framework for how decisions get made under stress.

What Rational Contingency Planning Actually Is

Let’s get one thing straight: rational contingency planning is not predicting the future. It is not assuming the worst and hoarding canned goods. It is a disciplined method for making decisions before you need them, so that when the pressure arrives you aren’t starting from scratch.

Think of it like a prenuptial agreement. Nobody enjoys drafting one, but the conversations required to write it expose what each side actually expects. The document itself is less valuable than the thinking you did to produce it.

The same principle applies to crisis management protocols. You gather the right people in a room, you face the uncomfortable scenarios nobody wants to discuss, and you reason through them now, at low stakes. You decide what you would do, who would approve it, and what criteria would trigger a shift in strategy. When the day comes—and it will come—you don’t retrace that entire reasoning path. You reach for the path you already paved.

This approach comes from a family of ideas used by military planners, emergency managers and high-reliability organizations. They don’t ask “what will happen?” They ask “what could happen, and what is our best response to that possibility, given what we know today?” That is a fundamentally different discipline than prediction. It’s preparation.

Step 1: Get Honest About What Could Hurt You

The first step is uncomfortable. You need to list your biggest risks honestly, in writing, and rank them. That includes risks you don’t want to admit, like the single point of failure sitting in your engineering team, or the supplier who could bankrupt you with a two-week delay.

Here’s the interesting part. Scenario planning isn’t about crafting elaborate narratives. It’s about separating remote possibilities from high-impact vulnerabilities without letting emotion distort the analysis. For each scenario, you need to estimate two things: probability and impact. Sounds simple. It isn’t, because human brains are terrible at probability. We overestimate dramatic risks and underestimate slow-moving ones. To counter that, you need more than gut feel.

Rather than relying on the loudest voice in the room, build your risk matrix from evidence—industry incident data, your own audit findings, near-miss reports from your own teams. A near-miss is a gift. It means your organization has already shown you the shape of its next problem. Did you pay attention, or did everyone shrug and move on?

If you want to structure this systematically, there are good public frameworks and resources available. FEMA’s emergency management planning guidance is one of the most accessible starting points for organizations of any size. You don’t need to become a government agency to borrow from their playbook.

Once you have your list, rank it. Focus first on the risks that combine moderate probability with high impact, because those are where rational contingency planning buys the most breathing room. The once-a-decade pandemic analogy always makes people roll their eyes until they live through one. Prepare for the plausible, not just the probable.

Questions worth asking in your first planning session

• What event would force us to stop operating for more than a week?
• Which single person could walk out tomorrow and break our ability to respond?
• Which of our current protocols would fail if the company doubled in size overnight?
• What would we never put in the news release, and why are we hiding it?

Those four questions create a different kind of conversation. They expose dependencies. They reveal assumptions. And they usually catch one or two glaring holes that keep everyone quiet for a moment.

Step 2: Pre-write Your Decision Triggers

A crisis management protocol without triggers is just a list of suggestions. Suggestions are useless at 2 a.m., when nobody can agree on whether the situation is serious enough to wake the CEO.

Here’s what I mean. A trigger is a predetermined, observable condition that tells your team which playbook to open. It removes the existential debate that kills response speed. For example, a hospital doesn’t debate whether chest pain is a “bad sign.” They use a triage scale. It’s not fun, but it works.

So map your own thresholds. What has to be true for you to classify an incident as Level One, Level Two, or Level Three? Perhaps Level One is an incident contained within one team with no customer impact. Level Two is broader, but you still have 24 hours to respond without serious damage. Level Three is the one where lives, massive financial loss, or public trust are on the line. The rational contingency plan assigns the right protocol to each level before you know which level you’ll face.

Here is where the phrase “crisis management protocols based on rational contingency planning” stops being jargon and becomes practical. Pre-set thresholds remove the emotional debate. When something happens, you don’t ask “Is this a big deal?” You ask “Which criteria have we already defined?” That tiny difference in phrasing changes everything. It moves you from interpretation to identification.

And the response to a trigger should be automatic—who gets notified, who convenes, what information must be collected in the first hour. You are allowed to choose