Summary of "5 AI Prompts to Think Like a Strategy Consultant"

Core thesis

AI’s highest business value is as an intellectual sparring partner that helps you think structurally and critically — not merely as a faster task-executor. Use prompts and playbooks to make AI apply consulting-grade frameworks to real problems.

Frameworks, processes, and playbooks

Driver tree (3 levels, mathematical) Break a top-line metric into additive/multiplicative sub-drivers. Example:
- mortgage originations = applications × average loan size
- applications = leads × conversion rate
Charting decision tree (Slide Science) A rule-based flow to select the optimal visualization (example: waterfall for year-over-year revenue drivers).
Cost–benefit / feasibility shortlist List approaches per driver, do a rough cost/benefit and feasibility estimate, and prioritize the best interventions.
Assumption-testing checklist Enumerate implicit assumptions, then design tests (prefer quantitative) to validate or invalidate them.
Counterargument generation Force AI to produce alternative strategies with reasons, hidden risks, and second-order effects.
Trade-off table (A vs B) Compare two options across dimensions, score each dimension, and consider hybrid solutions.
Rapid experiment suggestions Prioritize historical-data analysis and controlled market tests to validate assumptions quickly.

Key metrics, KPIs, and relationships called out

Top-line metric: mortgage originations (in dollars).
Explicit multiplicative relationship: number of applications = number of leads × conversion rate.
Example scenario metric: production cost lowered by 10% (used to evaluate pricing moves).
Recommended visualization for Year 1 → Year 2 driver contributions: waterfall chart.
Business outcomes to consider: market share, revenue, margins, customer perception, support costs (human vs chatbot).
Implied KPIs to track:
- leads
- conversion rate
- average loan size
- price elasticity / market share response
- margin impact
- customer satisfaction / perception
- support cost per ticket
- escalation rate

Concrete examples, case studies, and actionable recommendations

Mortgage portfolio example
- Prompt AI to build a 3-level mathematical driver tree for total mortgage originations (explicitly request math relationships).
- Use the driver tree to list interventions per driver, then run a rough cost–benefit and feasibility shortlisting.
Chart selection example
- Provide AI with your charting decision tree and the dataset (CSV). AI maps data to the recommended chart (e.g., waterfall for revenue-driver breakdown).
Pricing / market-share scenario
- Scenario: production cost down 10%; company considers lowering price to win share.
- Use AI to identify implicit assumptions (market price-sensitivity, competitor non-response, unchanged customer quality perception).
- For unconsidered assumptions, prompt AI for quantitative tests (recommended: historical-response analysis, controlled market tests).
M&A vs partnership (fintech example)
- Original plan: acquire a fast-growing fintech to modernize digital banking and attract younger customers.
- AI-generated alternative: partner with multiple fintechs — benefits include lower risk exposure, capital preservation, scalability, and brand/positioning flexibility.
Support automation example
- Compare AI chatbots vs human agents using a trade-off table across dimensions (cost, quality, resolution time, customer satisfaction, scalability).
- Recommendation: consider a hybrid model — chatbots for basic queries; humans for escalations and complex issues. Score dimensions to guide the decision.

Actionable prompt templates (copy / adapt)

Driver tree (mortgage):

I work at a bank and I want to understand what levers I can pull to improve total mortgage originations in dollars. Create me a driver tree that's three levels deep and shows the main drivers. Visualize using bullet points where each level indicates the level of the driver tree. Make it purely mathematical and show mathematical relationships (addition, subtraction, multiplication, division).

Charting decision tree:

Here is the slide science charting decision tree. Whenever I give you data and ask you to suggest a chart, use this as your guide. If you're not sure, ask me questions to clarify the insight I want to show.
[Then upload CSV]

Implicit assumptions:

Identify any implicit assumptions in this decision that may be incorrect. List them as bullet points of one to three sentences. Only use bold at the beginning of sentences.

How to test an assumption:

I never considered whether competitors would respond. How do I test this assumption to determine whether it's true? Give preference to quantitative tests and keep your description short.

Counterargument / alternative:

Generate a compelling argument for an alternative that achieves the same or similar outcome with five reasons why this option is preferred. Consider implicit assumptions, hidden risks, second-order consequences, and strategic alternatives. Present reasons as short bullet points; only use bold at the beginning of sentences.

Trade-off table (A vs B):

List the trade-offs that we are making. Use a table where column A is the trade-off, column B is option one, column C is option two. Keep responses short and only use bold at the beginning of sentences.

Tactical tips for using AI as a strategist

Be explicit in prompts: specify format (driver tree, table), depth (e.g., 3 levels), and whether relationships must be mathematical.
Provide your frameworks (decision trees, cheat sheets) so AI applies your preferred method to the data.
Ask AI to identify assumptions and propose short, preferably quantitative, tests before executing strategy.
Use AI to produce counterarguments and trade-off tables to avoid tunnel vision; then score dimensions numerically if useful.
Prefer visualization guidance (e.g., recommend waterfall for additive YoY drivers) rather than relying on generic chart suggestions.

Presenters / sources

Presenter: unnamed video creator referencing a personal “strategy system” and the “Slide Science charting decision tree.”
External reference: consulting practices (e.g., firms like McKinsey) mentioned as conceptual background.