Why Leadership Decisions Are Hard
Leadership decisions differ from individual decisions in ways that make them systematically harder: they’re made under uncertainty (the relevant information isn’t fully available), they affect others who have stakes in the outcome (which adds complexity and accountability), they’re made under time pressure (which compresses the reflection that good decisions benefit from), and their consequences may not be visible for months or years (which makes it difficult to learn from the decision’s outcome in time to inform the next similar decision).
The cognitive traps that affect leadership decisions specifically: confirmation bias (seeking information that confirms the preferred conclusion rather than information that might challenge it), authority bias (giving too much weight to the opinions of senior people or confident voices), sunk cost fallacy (continuing to invest in a direction because of what’s already been spent rather than because of future expected value), and overconfidence (underestimating uncertainty and the probability that the preferred approach is wrong). These traps don’t disappear with leadership experience; if anything, they can become more entrenched as leaders receive more deference that reduces the challenging information they encounter.
The Decision Framework: Type Before Process
Not every decision deserves the same decision process, and applying a thorough deliberative process to every decision produces both decision overload (so many decisions requiring extensive analysis that the volume becomes unmanageable) and decision underload (insufficient deliberation for the decisions that genuinely require it). The first leadership decision skill is accurately categorising the decision to match the right process to it.
Jeff Bezos’s widely cited framework distinguishes between Type 1 decisions (irreversible, high-consequence decisions that require deliberation and multiple perspectives) and Type 2 decisions (reversible, lower-consequence decisions that should be made quickly and corrected if wrong rather than deliberated extensively). Most leadership decisions are Type 2, and the common mistake is treating Type 2 decisions as Type 1 — slowing down, over-analysing, and creating committee review processes for decisions that should be made quickly by an informed person and reversed if the outcome is wrong.
Gathering Information Without Analysis Paralysis
The information gathering discipline that produces better decisions without producing analysis paralysis: define the minimum information needed to make a good decision (not the maximum information that could theoretically be gathered), gather that information from the most credible sources, and make the decision rather than seeking additional confirmation that the preferred conclusion is correct. The additional information gathered beyond the minimum threshold typically provides marginal improvement to decision quality while consuming significant time.
The information sources that most improve leadership decision quality: the people closest to the relevant reality (the frontline employee who works with the process being changed, the customer who uses the product being considered, the expert who has specific knowledge the leader doesn’t), rather than the most articulate people or the most senior people. Leaders who build the habit of seeking information from the most relevant rather than most accessible source make better decisions than those who rely on the same internal team regardless of their proximity to the relevant domain.
Involving Others: When to Decide Alone and When to Decide Together
The decision about how to make a decision is itself a leadership decision: some decisions benefit from the input and buy-in of others; others should be made quickly by the person with the best relevant information. The leader who involves the whole team in every decision trades decision quality for inclusion; the one who makes all decisions alone trades commitment and buy-in for speed. The relevant factors: how much does the decision quality improve with input from others, and how much does the implementation of the decision depend on others’ commitment to it?
The Vroom-Yetton-Jago decision model provides a structured framework for this meta-decision: it guides leaders through questions about decision quality requirements, the leader’s information sufficiency, the acceptance requirement, and the likelihood of subordinate acceptance to recommend the appropriate level of involvement. Most experienced leaders develop intuition for this calibration through experience; the framework is useful for making that intuition explicit and for examining cases where the intuitive choice may be wrong.
Learning From Decisions: The Practice Most Leaders Skip
Leadership decision quality improves when leaders create a feedback loop between decisions and their outcomes — a practice that requires both discipline (deliberately revisiting decisions to compare expected versus actual outcomes) and psychological safety (willingness to examine decisions that turned out badly without defending them). The leader who reviews outcomes only for decisions that worked out well is optimising for self-esteem rather than for learning; the one who examines both successes and failures with equal curiosity is building the calibration that improves future decisions.
The decision review practice that produces the most learning: for significant decisions, write down the decision made, the reasoning behind it, the alternatives considered and why they were rejected, and the specific outcomes expected. Six to twelve months later, compare actual outcomes to expected outcomes. Where outcomes matched expectations, understand why the reasoning was right; where they diverged, understand what was missing from the original analysis. This retrospective discipline — the business equivalent of the pre-mortem and post-mortem that surgical teams use — converts decision experience into decision wisdom at a rate that experience alone doesn’t produce.