The 5-Step Decision System

Why structure matters

Most people make significant decisions the same way they make minor ones: by thinking about the situation until a preferred option emerges, then looking for reasons to support it. This process works reasonably well for low-stakes, reversible decisions where feedback is fast. It is unreliable for decisions where the stakes are high, the outcome will not be visible for months or years, and the biases operating are invisible to the decision-maker.

The problem is not intelligence. Intelligent people make systematic decision errors because intelligence, without structure, primarily makes you better at building elaborate justifications for positions you already hold. Research by Philip Tetlock, whose superforecaster studies tracked thousands of predictions, found that structured, deliberate reasoning processes produced significantly better predictions than unstructured intuition, even when the structured reasoners were less expert in the subject matter.

The 5-step decision system is not a replacement for judgement. It is a structure that forces judgement to operate on the right material, in the right sequence, before a commitment is made. Each step addresses a specific and well-documented failure mode in human decision-making.

The five steps

Step 1: Capture and Clarify. Write the decision in one sentence. Most decisions are fuzzy because they have never been stated precisely. "I need to figure out my career" is not a decision. "Should I leave my current role and join Company X as a senior manager by 1 September?" is a decision. The discipline of writing one sentence forces you to identify what is actually being decided and what timeline applies. If you cannot write the decision in one sentence, you are not yet ready to decide: you are still in the information-gathering or problem-definition phase.

Step 2: Frame with a Mental Model. Choose one model that fits the type of decision you are making. Life pivots with long time horizons respond to the Regret Minimization Framework. Financial bets with estimable probabilities respond to expected value thinking. Execution risks in complex projects respond to the pre-mortem. First-principles thinking helps when the standard approach in your field is producing consistently poor results. The model does not decide for you; it provides the lens through which to structure your analysis. Choosing the wrong model is worse than using no model, so take 10 minutes to identify which decision type you are facing before selecting.

Step 3: Run the Bias Scan. Identify the two or three biases most likely to be distorting your thinking on this specific decision. Not every bias is active in every decision. For a career decision, status quo bias, loss aversion, and social proof are typically the most active. For a financial investment, overconfidence, anchoring to the purchase price, and the sunk cost fallacy are the most common distortions. Naming the specific biases that apply gives you a checklist to work through rather than a vague instruction to "watch out for bias." For each identified bias, ask: in what specific way might this bias be pushing me toward my current preferred option?

Step 4: The Pre-Mortem. Imagine it is 18 months from now and the decision has been executed. It has failed. Not partially failed; it has failed clearly and significantly. Write down, in as much detail as you can, what caused the failure. The pre-mortem, developed by psychologist Gary Klein, works because it gives people explicit permission to think about failure before committing, which is a social role that is otherwise difficult to play. The pre-mortem reliably surfaces risks that were present in the decision-maker's mind but suppressed by optimism, group pressure, or the desire to appear decisive. The goal is not to produce a reason to abandon the decision but to surface the most plausible failure modes so they can be addressed in the execution plan.

Step 5: Decide, Record, and Track. Make the decision. Write it down along with: the reasoning that led to it, the alternatives you considered and why you rejected them, the biases you identified in Step 3 and how you accounted for them, the key risks identified in Step 4 and how you plan to mitigate them, and the expected outcome with a timeline for review. Set a calendar reminder for 90 days. At the 90-day mark, review what actually happened against what you predicted. This review is the feedback loop that, over time, improves the quality of your decision-making by calibrating your judgements to reality.

When to apply the full system

Not every decision warrants all five steps. Reserve the full process for decisions that are hard or impossible to reverse, where the consequences will compound over more than 12 months, where significant resources or relationships are at stake, or where you feel strong emotional pull toward a particular option. These conditions indicate that the decision is consequential enough for structured analysis and that the biases are likely to be most active.

For reversible, low-stakes decisions with a short feedback loop, the overhead of the full system outweighs its benefits. A hiring decision for a senior role warrants all five steps. A decision about which content to publish next week does not.

How to get started

Identify one significant decision you are currently avoiding or circling around. Apply Step 1 right now: write it in one sentence. If the sentence does not come easily, spend 10 minutes free-writing about what the decision actually involves, then distil it. The act of writing the decision clearly is often the highest-value single action in the entire process, because it forces the fuzzy tension you have been carrying to become a specific question with specific options.

Once Step 1 is complete, run through Steps 2 through 4 before your next conversation about the decision with anyone else. Having completed Steps 2-4 independently means you bring structured analysis to the conversation rather than unexamined instinct. The conversation then becomes a test of your analysis rather than the source of your views.

Use DecisionsMatter.ai to run the guided version of the system for your first attempt. The tool walks you through each step with prompts, surfacing the relevant mental models and biases for your specific decision context and generating a structured record of the full analysis.

---