Module 1 of 10

Think Like a Strategist

Learn to work with AI. Then learn to think. Both are non-negotiable.

1.0The AI Mental Modelupdated Sep 18, 2026
I want to understand how to work with you more effectively.
My situation: [WHO I AM - role, industry, company stage]
What I typically use AI for: [YOUR MAIN USE CASES]
Based on this, tell me:
1.Where am I most likely leaving quality on the table?
2.What context should I always give you that I probably forget?
3.What is one thing I can do differently from today that would
       improve every output I get from you?
Be direct. Assume I am smart but new to working with AI well.
How to improve AI output?
Define Role
Specifies the AI's perspective, leading to
more tailored responses.
Provide Context
Informs the AI about the situation, ensuring
relevance.
Set Expectations
Guides the AI towards the desired output
format and content.
PRO TIP
The most common mistake is treating a bad output as a model failure. It is
almost never the model. It is almost always a missing Role, thin Context, or no
stated Expectation. Diagnose before you retry.
1.1C.A.R.E Promptingupdated Sep 18, 2026
Role:               Act as a [EXPERT ROLE  e.g. B2B sales coach].
Context:        I am [YOUR SITUATION  company, stage, industry].
Action:           [WHAT YOU WANT e.g. Give me 5 positioning ideas
                        to differentiate in the [MARKET / REGION]].
Expectation: Format as a numbered list. Each point 2–3 lines.
Use plain business language. No jargon.
CRAFTING EFFECTIVE AI PROMPTS
4 State Expectation
Clearly outline the desired format,
length, and tone of the output.
3 Start with Role
Begin the prompt by defining the
AI's persona.
2 Fill C.A.R.E Layers
Provide context, action, role, and
expectation for the prompt.
Identify Need
1
Determine the specific task you
want the AI to perform.
PRO TIP
If your output feels off, the problem is almost always a vague Role or missing
Expectation. Fix those two before rewriting the whole prompt.
1.2The Context Stackupdated Sep 18, 2026
Situation           :     [What is happening - current challenge or goal]
Background      :     [Industry, team size, stage, audience, product, region -   
                                  whatever is relevant]
Constraints       :     [Time, budget, tone, audience level, things to avoid]
Output format  :     [List / paragraph / table / email draft]
Now                    :     [YOUR ACTUAL QUESTION OR REQUEST]
AI Output Feels Shallow: Missing Context
Recurring Tasks
No built-in template
Reusing Context
functionality
Stack
Requires manual pasting
Non-Obvious each time
Constraints
AI misses important
limitations Imprecise Context
Stack
Leads to imprecise AI
output
No AI Memory
Every prompt starts from
zero
PRO TIP
Save your most-used context stacks as text snippets. Paste them at the start of any AI
session. This single habit cuts prompt iteration time by half.
1.3Role Prompting - The Expert Chairupdated Sep 18, 2026
Act as a [SPECIFIC EXPERT  e.g. senior B2B SaaS sales strategist with 15 years
selling to enterprise CTOs].
You care most about:  [THEIR PRIORITY - e.g. shortening sales cycles and
increasing deal value, not just volume].
From your expert perspective: [YOUR QUESTION OR SCENARIO - be specific]
Give your honest assessment. Flag risks I may not have considered.
Define AI Roles
High Importance
Broad Roles Expert Roles
General professional Highly specialized
perspectives professional viewpoints
Low Specificity High Specificity
Generic AI Niche Roles
Standard, unassigned AI Highly specific but less
output critical perspectives
Low Importance
PRO TIP
The most underused version: assign a role that would disagree with your plan. 'Act as a
sceptical investor' or 'Act as the customer who would not buy this' reveals blind spots no
internal meeting will find.
1.4The Refinement Loopupdated Sep 18, 2026
Step 1    —   Draft: [Paste your original prompt and generate first output]
Step 2   —   Critique: Review the response you just gave.
                     What are the 3 weakest parts? Where is it too vague or generic?
Step 3   —    Improve: Rewrite the original response with those 3 weaknesses fixed.
                     Keep what was strong. Improve what was weak.
Step 4   —    Lock:  Anything that would make this significantly better?
                     If yes, apply it. If no, confirm this is the final version.
Achieving AI Output Excellence
Critique
Ask AI to evaluate its
work
Draft Improve
Generate initial AI Ask AI to fix
output weaknesses
PRO TIP
Use Step 2 as a standalone habit. After any important output-  email, proposal, plan and
ask AI 'What is the weakest part of this?' before you send it. Takes 10 seconds.
1.5Six Thinking Hatsupdated Sep 18, 2026
Act as a strategic facilitator using De Bono's Six Thinking Hats.
Topic           :    [YOUR DECISION OR IDEA - be specific]
White Hat   :    Facts and data we have or need?
Red Hat       :    Gut feelings and emotional reactions?
Black Hat    :    Risks, flaws, what could go wrong?
Yellow Hat  :    Benefits and best-case outcomes?
Green Hat   :    Creative alternatives and new directions?
Blue Hat      :    Recommended next step or process?
Give 3 specific points per hat. Be concrete, not generic.
Six Thinking Hats Process
White Hat
1
Focus on facts and data
Red Hat
2
Consider emotions and gut reactions
Black Hat
3
Identify risks and problems
Yellow Hat
4
Explore benefits and upside
Green Hat
5
Generate creative alternatives
Blue Hat
6
Define process and next steps
PRO TIP
Run this twice  once for your plan, once asking AI to argue the opposite position. The
gap between the two outputs is where the most useful insight is hiding.
1.6First Principles Thinkingupdated Sep 18, 2026
Act as a first principles thinking coach.
Current approach or problem: [DESCRIBE YOUR SITUATION SPECIFICALLY]
Step 1  :  List every assumption in this approach.
Step 2 :  For each - is this actually true, or just inherited convention?
Step 3 :  Which are fixed constraints vs. just habits?
Step 4 :  Reconstruct a fresh approach using only what is genuinely true and necessary.
Flag the 3 most dangerous assumptions first.
FIRST PRINCIPLES THINKING: UNVEILING HIDDEN ASSUMPTIONS
Stuck Solution
Obvious problem
 with no breakthrough
Inherited Approach
Unquestioned way 
of doing things
Embedded Assumptions
Unverified beliefs underlying 
Fundamental Truths the approach
Basic, undeniable 
facts about the problem
PRO TIP
The most valuable output isn't the new idea, it's the list of assumptions you'd never
questioned. That list alone is worth running this prompt every quarter.
1.7Second Order Thinkingupdated Sep 18, 2026
Act as a strategic advisor specialising in consequence mapping.
Proposed action     :   [YOUR DECISION OR ACTION]
Context                    :   [INDUSTRY, STAGE, STAKEHOLDERS AFFECTED]
1st Order                  :   Immediate, direct effects?
2nd Order                :   What does each 1st-order effect then cause?
3rd Order                 :   What does each 2nd-order effect lead to?
Flag any 2nd or 3rd order effects that could undermine the original goal
even if 1st order looks positive.
Third-Order Thinking
Examines long-term
implications
Second-Order Thinking Negative Consequences
Considers subsequent Identifies potential
consequences downsides
First-Order Thinking Decision Adjustment
Focuses on immediate Modifies plans based on
outcomes analysis
The Power of
Second-Order
Thinking
PRO TIP
Don't just apply this to risky decisions. Good decisions often have harmful second-order
effects that nobody maps because everyone is celebrating the first-order win.
1.8Inversion Thinkingupdated Sep 18, 2026
Act as a strategic risk advisor using inversion thinking.
Goal        :   [WHAT I AM TRYING TO ACHIEVE  be specific]
Context  :   [SITUATION, INDUSTRY, STAGE, TEAM]
Step 1      :   List everything that would guarantee I fail. Think broadly - execution,  
                     market, team,  timing, assumptions, behaviour.
Step 2     :   Rank these failure causes by probability.
Step 3     :   For the top 5, give one specific prevention action I can take 
                    before moving forward.
STRESS-TEST PLANS WITH INVERSION THINKING
Review and Mitigate List Failure Causes
Cross off already Ask AI for failure
addressed risks guarantees
Build Prevention Inversion Thinking
Define Goal
Actions
Be specific about
Create actions to
achievement
avoid failure
Unrevealed Plan Problems Stress-Tested Plan
Forward thinking misses risks Risks are identified and avoided
PRO TIP
Share the failure list with your team before launch. Ask each person to add one item you
missed. The combined list will contain at least one thing that would have actually stopped
you  and now it won't.
1.9Mental Models Libraryupdated Sep 18, 2026
Act as a mental models coach.
My situation           :    [DESCRIBE THE CHALLENGE OR DECISION]
Apply these models and give one specific insight each:
Pareto 80/20         :    Where is the highest leverage?
Parkinson's Law    :    Where is time being wasted?
Dunning-Kruger   :    Where might I be overconfident?
First Principles.    :    What assumption am I not questioning?
Inversion                :    What would guarantee this fails?
Give one concrete action per model. Be direct.
MENTAL MODELS RANGE FROM SIMPLE TO COMPLEX UNDERSTANDING.
Dunning-Kruger Parkinson's Law
Maps confidence vs competence Explains work expansion over time
Iceberg Model Pareto's 80/20
Shows surface vs hidden drivers Identifies high-impact areas
Simple Complex
PRO TIP
After every quarter, add one new mental model you discovered and note a situation
where it would have helped. In three years you'll have a thinking toolkit no one can buy.
1.10OODA Loopupdated Sep 18, 2026
Act as a strategic decision advisor using John Boyd's OODA Loop.
Situation                         :    [DESCRIBE WHAT IS HAPPENING RIGHT NOW]
Available information   :    [LIST WHAT YOU KNOW , even partial data]
Constraints                     :    [TIME AVAILABLE, RESOURCES, KEY RISKS]
Observe                           :    What are the most critical signals in this data?
Orient                              :    What patterns, threats, or opportunities do these reveal?
Decide                             :    Give me 3 decision options ranked by speed and impact.
Act                                    :    For the top option, what is the immediate first move?
Prioritise speed. Flag what I should monitor after acting.
OODA LOOP — SPEED UNDER UNCERTAINTY
Observe
Act
Gather information about
Implement the decision the environment
Decide Orient
Choose the best course of Analyze and understand
action the situation
PRO TIP
The Orient phase is where most leaders get stuck,  they have the data but can't see the
pattern fast enough. That's exactly what AI compresses. Use it there, not as a
replacement for your judgment on the Decide and Act stages.
1.11SCAMPERupdated Sep 18, 2026
Act as a creative innovation facilitator using SCAMPER.
I want to innovate on   :  [YOUR PRODUCT, SERVICE, OR PROCESS]
Context                           :   [INDUSTRY, TARGET CUSTOMER, CURRENT LIMITATIONS]
Substitute                       :  What component or step could be replaced?
Combine                         :   What could be merged for more value?
Adapt                               :   What works elsewhere that I could borrow?
Modify                             :   What could be changed, enlarged, or reduced?
Put to other uses           :   What other jobs could this do?
Eliminate                         :   What could be removed entirely?
Reverse                            :   What if the order or direction was flipped?
Give one specific, actionable idea per lens.
SCAMPER: Structured Creativity for Innovation
Stagnant Ideas Innovative Solutions
Apply SCAMPER
Variations of the same New directions for the
Lenses
thing idea
Clearly state the Use seven creative Merge ideas from Develop the most
innovation target lenses different lenses promising ideas
PRO TIP
Run SCAMPER on your competitors' products, not just your own. Ask 'what would
happen if a competitor applied SCAMPER to our product?' The answers will tell you
your vulnerabilities before they do.
1.12MECE Frameworkupdated Sep 18, 2026
Act as a McKinsey-trained strategy consultant.
Problem or question to structure: [YOUR PROBLEM, DECISION, OR ANALYSIS TOPIC]
Step 1       :    Break this into MECE categories. Each category must be 
                       distinct,  no overlap. Together they must cover everything, no gaps.
Step 2      :    Check your structure. Flag any overlaps or missing areas.
Step 3      :    For each category, list the 2–3 key questions I need to answer to 
                       address it completely.
Format as a structured outline I can use directly.
MECE Analysis Framework
Complete Analysis
Outcome of MECE
analysis
Problem
Issue to be
analyzed
Collectively Exhaustive
Covers all
possibilities
Mutually Exclusive
No overlap between
categories
MECE
Core concept of
comprehensive analysis
PRO TIP
MECE is not just for analysis  use it to structure any important communication. Before a
board presentation or investor meeting, ask AI to check if your argument is MECE.
Incomplete structure is the most common reason smart people fail to persuade.