Decision Rules and Checklists: Deciding Before the Crunch

Pressure is a poor time to think through a decision from scratch. Information is incomplete, people are tired, and the most urgent task crowds out the most important one. Organizations that handle volatile moments well have usually done much of their thinking in advance.

Two simple tools carry that thinking forward: decision rules, which say what you will do when a specific situation arises, and checklists, which make sure critical steps happen even when attention is stretched.

This guide draws on military planning doctrine, research on if-then planning and NASA-funded human factors research on aviation checklists to show how to build both, and where they stop working.

Why decide in advance

US Army doctrine on planning, ADP 5-0 (July 31, 2019), states the case directly: "By anticipating potential events beforehand, planning promotes flexibility and rapid decision making during execution."

The doctrine names three tools for this:

• A decision point: "a point in space and time when the commander or staff anticipates making a key decision concerning a specific course of action."
• A branch: a contingency option built into the base plan, for changing direction in response to anticipated events, opportunities or disruptions.
• A sequel: the next operation or phase, based on how the current one turns out, including success, stalemate or defeat.

Army planners record these in a decision support matrix that, in the doctrine's words, "describes where and when a decision must be made if a specific action is to take place." A business version can be a single page.

Write if-then decision rules

Psychologists call a plan in the form "if this happens, then I will do that" an implementation intention. A 2016 review in Academic Medicine by Saddawi-Konefka and colleagues, including psychologist Peter Gollwitzer, describes these as "if-then" plans: "If I encounter situation X, then I will respond with action Y." The review contrasts them with simple goals and says they have shown substantial effectiveness over goals alone in increasing goal attainment.

The research mainly concerns individuals, but the same structure makes organizational rules clearer. Some hypothetical examples:

• If cash on hand falls below eight weeks of payroll, then we pause discretionary hiring and the owner approves every purchase over $2,000.
• If our largest customer's orders fall 25 percent for two straight months, then we start our customer diversification plan.
• If a key system is down for more than two hours during business hours, then the operations lead switches to manual order taking and notifies customers.

Good decision rules share four features: a trigger anyone can observe, a named owner, a specific first action and a date to review the rule.

Write exit rules too. A field guide from the European Commission's Joint Research Centre and the Cynefin Centre advises creating response strategies early in a crisis, "including what the triggers will be for a phase shift back to normal operation when the crisis is over." Without that, temporary crisis measures tend to become permanent.

What aviation learned about checklists

Aviation has studied checklists longer than most fields. In a 1990 NASA contractor report, Asaf Degani and Earl Wiener noted that the improper use, or non-use, of the normal checklist by flight crews is often cited as a probable cause or contributing factor in aircraft accidents.

They described two main methods:

• Challenge-response, which they suggested could be more accurately termed "challenge-verification-response": crews set up the aircraft from memory, then use the checklist to verify that critical items were done.
• Do-list, or "call-do-response": the checklist leads the crew through each step in a cookbook approach. The authors noted that without the double-check, a mistake can easily pass unnoticed once the sequence is interrupted.

They also observed that when checklists were long, crews tended to perform items while reading them, losing the built-in check. Length is not a neutral design choice.

For a business, the distinction is practical. Use a verify-style checklist for experienced people doing routine work, and a step-by-step list for rare, high-stakes tasks people do not remember well, such as restoring from a backup.

Design guidelines worth borrowing

Degani and Wiener closed their report with design guidelines for flight-deck checklists. Several translate well beyond the cockpit:

• Responses should state the desired status or value, not just "checked" or "set."
• Break a long checklist into smaller task checklists.
• Put the most critical items as close to the beginning as possible, so they get done before interruptions.
• Make the completion call the last item, so everyone knows the list is finished.
• Recognize that the procedure is highly susceptible to production pressure, which can push people to skip it to save time.
• Allow no compromise on the critical "killer" items.

Their conclusion is a useful reminder for anyone writing procedures: checklists must be "human-centered," because if a person chooses not to use one, no one can force them to.

Build a small crisis kit

You do not need a thick binder. Start with a few situations that are both plausible and costly for your organization, such as losing a key person, a major system outage, a supplier failure or a cash squeeze. For each, write:

1. The trigger that tells you the situation has started.
2. The owner who decides, and a backup.
3. A short first-hour checklist, with the most critical items first.
4. One or two decision points with branches: if this, then that.
5. The trigger for returning to normal.

Walk through each one with the people involved while things are calm, and revise anything that confuses them.

Know the limits

Rules and checklists work best for situations you can anticipate. In the terms of our guide The Cynefin Framework: Matching Your Approach to the Situation, they fit clear and complicated problems far better than novel, complex ones. Marine Corps doctrine in Warfighting (1997) warns that "We should base our decisions on awareness rather than on mechanical habit."

Treat rules as defaults, not handcuffs. If people keep overriding a rule for good reasons, rewrite it. And after any real use, ask what the rule or checklist missed.

Key takeaways

• Decide in advance where you can: identify decision points, branches and triggers before pressure hits.
• Write decision rules as if-then statements with an observable trigger, an owner, a first action and a review date.
• Include exit rules so crisis measures end when the crisis does.
• Keep checklists short, put critical items first and state the desired result for each item.
• Use rules as defaults and revise them when they stop fitting reality.

Sources

• Headquarters, Department of the Army, ADP 5-0 The Operations Process (July 31, 2019)
• Saddawi-Konefka and colleagues, Changing Physician Behavior With Implementation Intentions: Closing the Gap Between Intentions and Actions, Academic Medicine (September 2016)
• Asaf Degani and Earl L. Wiener, Human Factors of Flight-Deck Checklists: The Normal Checklist, NASA Contractor Report 177549 (May 1990)
• Cynefin.io, Field Guide to Managing Complexity (and Chaos) in Times of Crisis
• US Marine Corps, MCDP 1 Warfighting (June 20, 1997)