Chapter 25 Exercises: Decision Support, Analysis, and Strategic Thinking

These exercises build practical AI-assisted decision analysis skills. Work through them in sequence — each one builds conceptual understanding and prompt-crafting ability that the next exercises require.

Foundation Exercises (Exercises 1-5)

Exercise 1: Decision Matrix Construction

Objective: Build a decision matrix for a real or realistic decision using AI.

Setup: Identify a decision you're facing or have recently faced. It should have at least 3 options and multiple competing considerations.

Task: 1. Write a 3-5 sentence description of the decision, including context, constraints, and what success looks like. 2. Run the decision matrix prompt from Section 25.2 with this context. 3. Review the AI-suggested criteria. For each criterion: - Accept, modify, or reject it (with one sentence explaining why) - Adjust the weighting if you disagree with AI's suggested weight 4. Review the AI-generated scores. For each option/criterion cell where you strongly disagree with the score, note what the AI doesn't know about your specific situation that explains the discrepancy. 5. Re-run the matrix with your corrections and note how the ranking changes.

Reflection prompt: How many of the criteria did AI identify that you wouldn't have included? How many of the criteria did you reject as inapplicable? What does the ratio tell you about AI's usefulness for structuring vs. populating a decision matrix?

Exercise 2: The Devil's Advocate Challenge

Objective: Experience the value of structured adversarial analysis.

Task: 1. Identify a decision you've already made — one you felt confident about at the time. 2. Run the devil's advocate prompt from Section 25.3 against this decision. 3. Read the AI output with genuine openness. For each argument it makes: - Did you consider this when making the original decision? - Is it a strong argument, a weak argument, or irrelevant? - If you're making the decision again today, does this argument change anything? 4. Run the reverse devil's advocate on the option you rejected (or the option you nearly chose but didn't). 5. Write a paragraph reflecting on whether you still believe your decision was right — and if your confidence level has changed.

Reflection prompt: This exercise asks you to challenge a decision you've already made. Did you find it easy or hard to engage seriously with arguments against your choice? What does that tell you about how to use this technique for future decisions?

Exercise 3: Assumption Surfacing

Objective: Identify and validate the critical assumptions underlying a decision.

Task: 1. Choose a decision you're currently considering or have recently made. 2. Before running any AI prompt, write down the assumptions you know your reasoning depends on. Try to get at least 5. 3. Run the assumption surfacing prompt from Section 25.4. 4. Compare your manual list to the AI's list. Which assumptions did you list that AI didn't? Which did AI list that you didn't? 5. For each assumption AI identified that you hadn't: rate it as "obvious in retrospect / genuinely missed / not applicable." 6. Identify your single most critical assumption (the one where, if it's false, the decision collapses). Write a sentence about how you could test this assumption before committing.

Deliverable: An annotated assumption map with your critical assumption identified and a testing plan.

Exercise 4: Scenario Analysis

Objective: Apply scenario analysis to a decision with significant external uncertainty.

Task: 1. Identify a professional decision where the right answer depends heavily on how an uncertain external situation evolves (market conditions, technology adoption, regulatory environment, competitive landscape, etc.). 2. Run the scenario analysis prompt from Section 25.2 with this decision. 3. Review the four scenarios generated. For each: - Is this scenario plausible? Rate 1-5. - Are there scenarios you think are more plausible that aren't represented? 4. Identify which option is most "robust" (performs acceptably across the widest range of scenarios) vs. which option "bets" on a specific scenario. 5. Write a recommendation for your decision that accounts for the scenario analysis.

Reflection prompt: Does scenario analysis change which option you prefer? Or does it just give you a more structured way to explain a choice you were already making? What's the value of the analysis even if it confirms your original intuition?

Exercise 5: The Socratic Session

Objective: Use AI as a Socratic partner to clarify your own thinking.

Task: 1. Identify a decision you find genuinely difficult — one where you're stuck or conflicted. 2. Run the Socratic questioning prompt from Section 25.6. Provide a detailed description of the decision and your current thinking. 3. Receive the questions from AI. Do not ask for answers — just questions. 4. Write out your answers to each question as honestly as you can. 5. Send your answers back to AI and ask for a second round of follow-up questions. 6. After the second round, write a 2-3 sentence statement of what you now believe about the decision and why — specifically noting what changed from your initial framing.

Reflection prompt: Did the Socratic process change your framing of the decision, your preferred option, or your confidence level? Which question was most productive for your thinking?

Intermediate Exercises (Exercises 6-10)

Exercise 6: Strategic Analysis with Frameworks

Objective: Apply multiple strategic frameworks to a real situation and compare their insights.

Setup: Choose a strategic situation: a market entry decision, a competitive response, a product launch, an organizational change, or a partnership evaluation.

Task: 1. Run the SWOT/PESTLE/Porter's Five Forces prompt from Section 25.2 with full context. 2. Review each framework's output separately. For each one: - What is the most important insight from this framework? - What does this framework highlight that the others miss? 3. Write a synthesis paragraph: given all three frameworks, what is your strategic recommendation and why? 4. Identify one specific insight from the analysis that you would not have reached without AI's systematic prompting.

Deliverable: A 1-2 page strategic analysis document with three frameworks and a synthesis recommendation.

Exercise 7: The Competitive Analysis Stress Test

Objective: Use AI competitive analysis to identify gaps in your competitive understanding.

Task: 1. Choose a market or competitive situation you know reasonably well from direct experience. 2. Run the competitive analysis prompt from Section 25.5 with your known competitors. 3. Compare the AI's competitive analysis to your own knowledge: - What did AI get right that aligns with your experience? - What did AI get wrong or overstate? - What did AI miss entirely that you know from direct experience? 4. Write one "blind spot" question that AI's analysis caused you to investigate further. 5. Research that blind spot (using actual sources, not just AI). What did you find?

Reflection prompt: For competitive analysis, what's the right balance between AI-generated structure and primary research? When is AI competitive analysis useful and when is it potentially misleading?

Exercise 8: Decision Documentation Practice

Objective: Create a decision record for a significant past decision.

Task: 1. Choose a significant decision from the past 12-18 months — preferably one whose consequences are now partially visible. 2. Run the decision record prompt from Section 25.7 for this decision. 3. Review the AI-drafted record. Revise it to accurately reflect: - The actual reasoning at the time (not the post-hoc rationalization) - The actual options that were considered (not just the final framing) - The assumptions that turned out to be true vs. false 4. Add a "what actually happened" section (2-3 sentences). 5. Add a "what I'd do differently" section (2-3 sentences).

Reflection prompt: Reading the decision record now, what assumption was most wrong? Did you know it was an assumption at the time, or did you treat it as fact?

Exercise 9: The "Question You're Not Asking" Challenge

Objective: Use AI to reframe a persistent problem by changing the question.

Task: 1. Describe a problem you've been working on or worrying about for a while — something that hasn't resolved despite your attention. 2. Describe how you've been framing this problem — what question you've been trying to answer. 3. Run the "question you're not asking" prompt from Section 25.6. 4. For each question AI suggests you should be asking, write: - Why this question might be more important than your current framing - What you'd need to do to answer it - Whether reframing the problem this way changes what options are available 5. Select the question you find most generative. Spend 30 minutes exploring it — using AI to help you develop the analysis.

Deliverable: A one-page problem reframing document that starts with the original framing and ends with a different question and what it implies.

Exercise 10: Build-Buy-Partner Analysis

Objective: Apply structured AI decision support to a classic business architecture decision.

Setup: Use a real or realistic "build vs. buy vs. partner" decision. This could be a technology decision (build software, buy a tool, use a platform), a capability decision (build the skill internally, hire, or outsource), or a product decision.

Task: 1. Run a decision matrix analysis comparing build/buy/partner options. 2. Run the devil's advocate prompt for whichever option you're currently leaning toward. 3. Run the assumption surfacing prompt on your current leaning. 4. Run a scenario analysis considering 2-3 ways the external environment might evolve. 5. Write a one-page recommendation that integrates all four analyses, noting explicitly where the analyses agreed, where they conflicted, and how you resolved conflicts.

Deliverable: A one-page decision recommendation with analysis summary.

Advanced Exercises (Exercises 11-15)

Exercise 11: The "AI Was Wrong" Analysis

Objective: Understand AI's failure modes in decision support by analyzing a case where AI analysis would mislead you.

Task: 1. Choose a decision in your domain where the "obvious" or "average" answer is wrong for your specific context. (Think of a case where conventional wisdom doesn't apply to your situation.) 2. Describe the decision to AI using only the information someone from outside your organization would know. 3. Note the AI's recommendation. 4. Now add the context that AI didn't have — the organizational history, the specific relationship, the unusual market dynamic, the personal constraint. 5. Run the prompt again with the full context. Note how the recommendation changes. 6. Write a paragraph describing what type of context AI most systematically lacks for decisions in your domain.

Reflection prompt: What does this exercise reveal about the minimum context you must always provide to AI for decision analysis in your domain to be reliable?

Exercise 12: Cross-Framework Decision Analysis

Objective: Synthesize insights from five different decision frameworks applied to the same decision.

Task: Apply all five of the following to the same significant decision: 1. Decision matrix (Section 25.2) 2. Pros/cons with asymmetric analysis (Section 25.2) 3. Scenario analysis (Section 25.2) 4. Devil's advocate (Section 25.3) 5. Assumption surfacing (Section 25.4)

After running all five: - Where do the frameworks agree in their conclusions? - Where do they conflict? How do you resolve the conflict? - Which framework was most useful for this particular decision? Why? - Is there a decision type where this framework would be the first one to reach for?

Deliverable: A decision analysis report synthesizing all five frameworks with a final recommendation and confidence rating.

Exercise 13: The Ethics Check

Objective: Practice identifying decisions where AI decision support is appropriate vs. where human judgment must dominate.

Task: 1. Describe the following types of decisions and assess: how much weight should AI analysis carry in each? - Which candidate to advance to the next interview round - Whether to approve a loan application - Which project to cut when budget is reduced - Whether to escalate a customer complaint to senior leadership - Whether to pursue a strategic acquisition - How to handle a performance issue with a team member

2. For each decision, write: - What role AI can legitimately play (analysis, information structuring, documentation) - What must remain exclusively human (the judgment call itself) - What risks arise from over-relying on AI in this specific decision type

3. Develop a "decision ethics test" — a set of 3-5 questions you'll ask before using AI in a decision to determine what role AI should play.

Exercise 14: Strategic Scenario Workshop

Objective: Run a complete scenario planning session for a strategic challenge in your domain.

Task: 1. Identify a strategic question with a 3-5 year horizon in your professional domain. 2. Identify the two highest-uncertainty, highest-impact external factors. 3. Run the scenario planning prompt from Section 25.5. 4. For each of the four scenarios generated: - Develop the scenario further with additional context you know from your domain expertise - Identify 3 specific strategic actions that would be effective in this scenario 5. Identify the "no-regrets moves" — actions that are worth taking across all four scenarios. 6. Identify the "big bets" — actions that are highly valuable in some scenarios and risky in others. What would need to be true for you to make these bets?

Deliverable: A scenario planning document suitable for sharing with senior leadership.

Exercise 15: The Decision Audit

Objective: Apply retrospective analysis to a portfolio of past decisions to identify your personal decision-making patterns and blind spots.

Task: 1. List 10 significant decisions you've made over the past 2-3 years. Include decisions that turned out well and ones that didn't. 2. For each decision, use AI to help you analyze: - What type of decision it was (strategic/operational/personnel/financial) - What assumptions it rested on (in retrospect) - Which assumptions proved correct and which proved false - What you knew vs. what you assumed 3. Ask AI to help you identify patterns: "Based on these 10 decisions, what appears to be my systematic bias or blind spot as a decision-maker?" 4. Write a "decision-making user manual" for yourself: 5-7 specific guidance points based on your identified patterns.

Reflection prompt: What is your single most consistent decision-making blind spot? What specific prompt or process would help you counteract it?

Capstone Exercise

Exercise 16: The Full Decision Package

Objective: Produce a comprehensive decision analysis package for a real, current, consequential decision.

Constraint: The decision should be one you actually need to make — stakes should be high enough that the analysis matters.

Deliverable: A complete decision analysis package including: - Decision statement and context (1 paragraph) - Options analysis (decision matrix) - Scenario analysis (3-4 scenarios) - Devil's advocate analysis for your preferred option - Critical assumption identification and testing plan - Decision record (what you decided, why, what would trigger review)

Rules: - AI must support, not make, the decision - Every AI output must include your annotation of what you agreed with and what you changed - The final recommendation must be yours, and you must be able to explain it without referring to what AI said

Reflection: What was the most valuable contribution AI made to this decision? At what point did you need to stop relying on AI and exercise your own judgment?