The most underused Cowork superpower. Claude helps you structure complex decisions, stress-test your reasoning, apply the right frameworks, explore scenarios, and identify risks you haven't considered — before the stakes are real.
Human decision-making is systematically flawed by cognitive biases: confirmation bias (we seek evidence that confirms our view), anchoring (our first estimate sticks too long), loss aversion (we over-weight potential losses), and narrative fallacy (we construct plausible-sounding stories from incomplete data).
Claude serves as a powerful decision support tool precisely because it:
🪞
Is Bias-Free (Differently)
Claude has its own biases from training, but they're different from yours. It doesn't share your organizational politics, career stakes, or ego investment in a particular outcome.
📚
Knows the Frameworks
Claude has internalized hundreds of decision frameworks — from Economics (expected utility) to strategy (Porter's 5 Forces) to psychology (Kahneman's System 1/2). It applies them on demand.
🔄
Steelmans Alternatives
Claude can genuinely argue for options you're inclined to dismiss — producing the strongest case for each option rather than just validating your preference.
⚡
Is Available at Decision Speed
Boards and advisors aren't available at 11pm before a morning deadline. Claude is. For time-pressured decisions, having an always-on thinking partner changes the quality of decision-making.
🔑 Critical Rule
Claude informs decisions. You make them. Never outsource final judgment, moral responsibility, or accountability to Claude. It is a thinking tool, not a decision-maker. The line matters — especially in professional, legal, and personal contexts with real consequences.
Structured Decision Analysis
Master Decision Analysis Prompt
Help me think through a significant decision. I want structured analysis, not just validation of my instinct.
THE DECISION: [State the decision clearly — e.g., "Whether to accept a job offer at [Company] / whether to enter [market] / whether to fire/promote [role]"]
MY CURRENT LEAN: [What are you currently inclined to do and why?]
THE OPTIONS I'M CONSIDERING:
Option A: [description]
Option B: [description]
Option C: [any other options, including "do nothing"]
CONSTRAINTS:
- Non-negotiable factors: [things that MUST be true for any chosen option]
- Time pressure: [do I need to decide by when?]
- Who else is affected: [stakeholders who will be impacted]
WHAT I KNOW: [Relevant facts, data, context]
WHAT I DON'T KNOW: [Key uncertainties I'm working with]
Analyze:
1. CRITERIA: What are the 5 most important criteria for this decision? (suggest based on context, then I'll confirm)
2. OPTION SCORING: For each option, score it on each criterion (1-5) with brief justification
3. THE CASE FOR EACH OPTION: What's the strongest argument for each option?
4. RISKS OF EACH OPTION: What's the worst-case scenario for each?
5. WHAT I MIGHT BE MISSING: What am I not considering that I should be?
6. MY BLIND SPOTS: Based on my stated lean, what biases might be distorting my reasoning?
7. RECOMMENDATION: Given everything, what would you do and why? (Be direct.)
Decision Framework Application
Claude can apply established decision frameworks on demand. Tell it which framework to use, or ask it to recommend the right one:
Framework Application Prompts
# SWOT Analysis
"Apply a rigorous SWOT analysis to [decision/project/business]. Don't give me generic categories — be specific to our situation and identify the MOST significant factor in each quadrant."
# First Principles Thinking
"Break this problem down to first principles. Strip away all assumptions and industry conventions. Starting from first principles, what would the optimal solution to [problem] look like?"
# Second-Order Thinking
"For the decision to [do X], what are the second and third-order consequences? Not just what happens directly, but what happens because of what happens?"
# Expected Value Analysis
"Help me structure an expected value analysis for [decision]. For each option, estimate the probability and magnitude of the best and worst cases, then calculate expected value. Be explicit about your probability assumptions."
# 10/10/10 Rule (Suzy Welch)
"Apply the 10/10/10 rule to this decision: How will I feel about this choice 10 minutes from now? 10 months from now? 10 years from now?"
# Regret Minimization (Bezos Framework)
"Apply the Regret Minimization Framework: imagine myself at 80 years old looking back. Which of these options would I regret NOT choosing? Which would I regret choosing? Use this to identify the path I'd be most at peace with."
Devil's Advocate & Steelmanning
The most powerful decision support technique: ask Claude to argue the opposite of your current lean as persuasively as possible. This isn't about finding the right answer — it's about stress-testing your reasoning.
Devil's Advocate Prompt
I've decided to [YOUR DECISION — e.g., "hire Candidate A over Candidate B / accept this offer / pursue this strategy"].
My reasoning is: [explain your rationale]
Play aggressive devil's advocate. Give me the strongest possible case for why I'm WRONG to make this choice. Don't hold back — I want to stress-test my reasoning, not feel validated.
Specifically argue:
1. Why the option I'm choosing is weaker than I think
2. Why the option I'm rejecting is more attractive than I'm giving it credit for
3. What I'm over-weighting in my analysis
4. What I'm under-weighting
5. What scenario would make me regret this choice in 12 months?
After the devil's advocate case, tell me honestly: do you still think I'm making the right call? Or did your own argument convince you I should reconsider?
💡 The Steelman Technique
A steelman is the opposite of a strawman — it's the strongest possible version of an argument you disagree with. Ask Claude to "steelman" any option or position you're inclined to dismiss. If Claude's steelman argument changes your thinking, your original reasoning was weaker than you thought. If it doesn't move you at all, you probably have the right answer.
Scenario Planning
Scenario Planning Prompt
Help me do a structured scenario analysis for [DECISION/STRATEGY].
Context: [what you're deciding or planning for]
Time horizon: [how far ahead to plan — e.g., "12 months", "3 years"]
Key uncertainties: [what factors could go very differently than expected]
Build 3 scenarios:
SCENARIO A — BEST CASE: What does the world look like if conditions are favorable?
SCENARIO B — BASE CASE: What's the most likely realistic outcome?
SCENARIO C — WORST CASE: What does the world look like if key risks materialize?
For each scenario:
1. What conditions would make this scenario occur? (probability drivers)
2. What are the specific business/life implications for me?
3. What decisions would I make differently in this scenario?
4. What early warning signals would tell me this scenario is emerging?
Then: What is a ROBUST strategy that works reasonably well across all 3 scenarios? (The goal isn't to bet on one scenario but to find decisions that are resilient regardless of which materializes.)
Pre-Mortem Analysis
The pre-mortem (developed by psychologist Gary Klein) is one of the most effective decision tools: imagine your decision has already failed, then work backward to understand why. Claude is exceptional at generating pre-mortems because it has no ego investment in your plan succeeding.
Pre-Mortem Prompt
Run a pre-mortem on my plan. Imagine it is [TIME HORIZON — e.g., "12 months"] from now, and [PLAN/DECISION] has failed badly.
My plan: [describe it]
What "failure" would look like: [define the failure condition specifically]
You are a future version of my team, looking back at why this plan failed.
Generate:
1. THE AUTOPSY: What are the 5 most likely causes of failure? (rank by likelihood)
2. THE UNEXPECTED: What failure mode did we NOT see coming that seemed obvious in hindsight?
3. HUMAN FACTORS: What behaviors, decisions, or politics caused or contributed to failure?
4. EXTERNAL FACTORS: What market/timing/competitive factors worked against us?
5. WHAT WE IGNORED: What warning signs were visible before we started that we discounted?
Then, as your present-day self:
6. MITIGATION PLAN: For each failure mode, what can I do NOW to reduce the probability or impact?
7. KILL SWITCHES: What early indicators would tell me this plan is failing with time to change course?
Personal Life Decisions
Decision support isn't just for business. Claude is uniquely valuable for the major personal decisions that don't come with easy frameworks or external advisors:
🏡
Major Life Purchases
"Help me think through whether to buy vs. rent, given these specifics. I don't want generic advice — I want analysis calibrated to my actual situation: [details]"
💼
Career Decisions
"I'm evaluating two job offers. Here are the full details of each. Apply multiple frameworks and give me your honest recommendation."
🎓
Education Choices
"Should I do the MBA now or in 3 years? Help me map out the financial, career, and personal tradeoffs with specific scenarios for each path."
🌍
Life Direction
"I'm at a crossroads. Help me think through a major values/direction question using structured thinking rather than just my gut." — Claude excels at helping structure complex personal reflection.
Decision Ethics & Limits
⚠️ What Claude Cannot Do
Predict the future: Claude can model scenarios but cannot know what will happen. Its probability estimates are educated reasoning, not data.
Replace domain experts: For legal, medical, financial, or regulatory decisions with significant stakes — consult qualified professionals. Use Claude to become a better-informed client, not to avoid professional advice.
Know your full context: Claude only knows what you've told it. Critical context you haven't shared could completely change the optimal decision.
Own the decision: Claude can recommend but cannot be held accountable. The responsibility, risk, and consequence belong to you.
🏆 The Thinking Partner Philosophy
The best way to use Claude for decisions: treat it like an exceptionally smart, well-read friend with relevant knowledge who has no personal stake in the outcome. Ask it to push back. Ask it to worry on your behalf. Ask it to be honest when your reasoning is weak. The quality of your decisions will improve not because Claude decides for you — but because the act of externalizing and articulating your reasoning to Claude reveals the flaws in that reasoning before you commit to action.