Why Decisions Are Hard
Most important decisions share three characteristics that make them difficult: incomplete information (you never have all the facts), uncertainty (the future is inherently unpredictable), and emotional interference (fear, excitement, and ego cloud judgement). Frameworks don't eliminate these challenges — they give you a structured way to navigate them.
The Real Goal
The goal of better decision-making isn't to be right every time — that's impossible. The goal is to improve your batting average over hundreds of decisions, and to make the cost of being wrong survivable. Good decision processes produce good outcomes on average, even when individual decisions sometimes fail.
The Two Types of Decisions
Jeff Bezos frames decisions as either Type 1 (irreversible) or Type 2 (reversible). This distinction should change everything about how you approach them:
| Dimension | Type 1 (Irreversible) | Type 2 (Reversible) |
|---|---|---|
| Nature | One-way door — can't easily undo | Two-way door — can reverse course |
| Examples | Selling a company, having a child, major pivot | Launching a feature, hiring (with probation), pricing changes |
| Approach | Slow, deliberate, seek maximum info | Fast, bias toward action, iterate |
| Common mistake | Rushing through | Over-analysing (treating as Type 1) |
| Decision speed | Take all the time you need | Decide with ~70% of information |
The most common organisational failure is treating Type 2 decisions as Type 1 — subjecting every reversible decision to extensive analysis, committee review, and consensus-building. This creates paralysis. Most decisions should be made quickly and reversed if wrong.
Core Frameworks
The Pre-Mortem
What: Before executing a decision, imagine it's 12 months later and the decision has failed spectacularly. Now work backward: what went wrong?
How:
- Gather your team (or sit alone with a notebook).
- Say: "It's one year from now. We went ahead with [decision]. It was a disaster. Why?"
- Each person writes down every reason they can think of for the failure — independently, without discussion.
- Share and compile the reasons.
- For each risk identified, decide: can we mitigate this? Is the risk acceptable? Does this change our decision?
Why it works: It gives people psychological permission to voice doubts they'd normally suppress. It converts vague anxiety into specific, addressable risks. It's far cheaper to identify problems in advance than to discover them in production.
The Decision Journal
What: A written record of important decisions — captured at the time of the decision, not after the outcome is known.
Record for each decision:
- Date and context
- The decision and alternatives considered
- Your reasoning — why this option over the others
- What you expect to happen (specific predictions)
- How you feel about the decision (confident? uncertain? pressured?)
- What would change your mind (pre-committed tripwires)
Why it works: It defeats hindsight bias — the tendency to rewrite history so that past decisions look obviously right or obviously wrong. By recording your reasoning at the time, you can honestly evaluate whether your thinking process was good, independent of whether the outcome happened to be good.
10/10/10 Analysis
What: For any decision, ask: How will I feel about this in 10 minutes? 10 months? 10 years?
Why it works: It separates short-term emotion from long-term consequence. Many decisions that feel terrifying in the moment (leaving a job, having a difficult conversation, launching a product) feel obviously correct at the 10-month and 10-year horizon. Conversely, decisions that feel immediately gratifying (avoiding conflict, taking the safe option) often look like missed opportunities in hindsight.
The Regret Minimisation Framework
What: Jeff Bezos's personal framework for big life decisions. Project yourself to age 80 and ask: "Will I regret not having done this?"
Application: Bezos used this to decide to leave his hedge fund job and start Amazon. At 80, would he regret trying and failing? No — he'd regret never trying. This framework is most useful for one-way-door decisions where the default is safety and the alternative is risk with asymmetric upside.
Decision Hygiene
Good frameworks need a clean operating environment. These practices reduce noise and bias:
Separate the Decision from the Outcome
A good decision can produce a bad outcome (you made the right call with the information available, but got unlucky). A bad decision can produce a good outcome (you took an unjustifiable risk and got lucky). Judge decisions by the quality of the process, not the result. This is called "resulting" — and it's one of the most common errors in evaluating past decisions.
Seek Disconfirming Evidence
Before committing to a decision, actively seek reasons it's wrong. Ask people who disagree with you. Read the opposing argument. If your decision survives genuine challenge, you can have more confidence in it. If it doesn't, better to know now.
Assign Probabilities
Replace "I think this will work" with "I assign a 65% probability of success." This forces precision, creates accountability (you can track your calibration over time), and makes it easier to adjust when new information arrives.
Set Tripwires
Before executing a decision, pre-commit to conditions that would cause you to reverse course. "If we haven't reached 1,000 users by month 3, we'll pivot." Without tripwires, escalation of commitment (the sunk cost fallacy) will keep you invested in failing decisions long past the point of reason.
The 70% Rule
For reversible decisions, decide when you have approximately 70% of the information you wish you had. Waiting for 90% means you're almost certainly too slow — by the time you have that much data, the opportunity may have passed. The cost of moving fast and correcting course is almost always lower than the cost of waiting for certainty that never arrives.
Common Decision Traps
| Trap | What It Is | How to Counter It |
|---|---|---|
| Analysis Paralysis | Endless research and deliberation instead of deciding | Set a deadline. Use the reversibility test. Decide with 70% info. |
| Sunk Cost Fallacy | Continuing because of past investment rather than future value | Ask: "If I were starting fresh today, would I choose this?" If no, stop. |
| Anchoring | Over-weighting the first piece of information received | Deliberately generate multiple reference points before deciding. |
| Status Quo Bias | Preferring the current state simply because it's familiar | Evaluate the status quo as if it were a new option. Would you choose it? |
| Groupthink | Conforming to team consensus without critical evaluation | Assign a devil's advocate. Have people write opinions before discussion. |
| Overconfidence | Being more certain than the evidence warrants | Track your predictions. You're almost certainly less calibrated than you think. |
The Meta-Decision
Before you decide, decide how to decide. Is this a Type 1 or Type 2 decision? Who needs to be involved? What information do I need? What's my deadline? What framework will I use? Spending 5 minutes on the meta-decision saves hours of unfocused deliberation.