The Decision Stack: The Framework for Faster, Better Decisions at Every Level.

The Problem With Your Decisions

You are making decisions without a framework. Small decisions get escalated because nobody knows who should make them. Medium decisions take weeks because every stakeholder has a different criteria for evaluating them. Large decisions get made by the loudest person instead of the person with the best information. The result is slow, inconsistent decisions that the team does not trust.


Layer One: The Decision Map

Before any decision gets made, name the decision. Not the project. The specific decision. What are you actually choosing between? Who is the decision owner — the one person with authority to make it? Who needs to be consulted before it is made? Who needs to be informed after it is made? The decision map prevents the most common failure: the wrong people making the decision for the wrong reasons.


Layer Two: The Criteria

Every medium and large decision needs criteria before it is discussed. Not criteria that justify a decision already made. Criteria that define what a good decision looks like. The criteria should be written before the options are discussed. Priorities among the criteria should be explicit. This prevents the post-hoc rationalization problem: finding reasons to like the option you already preferred.


Layer Three: The Options

Generate three real options, not three variations of the same option. The worst decision processes produce two options: the preferred choice and the sacrifice. The best processes produce three genuinely different paths. If you cannot think of three real options, that is a signal that the decision space is narrower than you thought.


Layer Four: The Commitment

Every decision needs a commitment, not a recommendation. The decision owner commits to a specific action with a specific timeline. The commitment is recorded and shared with everyone who needs to know. The commitment includes what will be true in six months if the decision was right, and what will be true if it was wrong.