scenario-planning
Scenario Planning
Core principle: The future is uncertain. Scenario planning doesn't try to predict which future will happen — it maps a set of plausible futures and tests strategies against each. A robust strategy performs reasonably well across multiple scenarios. A fragile strategy only works if one specific future occurs.
The Core Process
Step 1: Define the Decision or Strategy
What are we planning for?
- What time horizon? (6 months / 2 years / 5 years)
- What decision needs to be made now?
- What would "good" look like across different futures?
Step 2: Identify Key Uncertainties
What are the 2–3 most important factors that are both:
- Highly uncertain (we genuinely don't know how they'll unfold)
- Highly impactful (they would significantly change what the right strategy is)
These are the scenario axes — the dimensions along which futures differ most.
Avoid certainties (things that will definitely happen) and minor factors. Focus on the pivotal unknowns.
Step 3: Build the Scenarios
From the key uncertainties, construct 3–4 distinct, internally consistent futures. Each scenario should be:
- Plausible: Not science fiction — could actually happen
- Distinct: Each scenario is meaningfully different from the others
- Challenging: At least one scenario is uncomfortable for the current plan
Common scenario structures:
Two-axis matrix (for two key uncertainties):
High market adoption
|
Slow tech ──┼── Fast tech
change | change
|
Low market adoption
This generates 4 quadrant scenarios.
Three narrative scenarios:
- Expected: The future most people are implicitly planning for
- Optimistic: Key uncertainties resolve favorably
- Pessimistic: Key uncertainties resolve adversarially
- Wild card: A low-probability, high-impact disruption
Step 4: Stress-Test the Strategy
For each scenario:
- Does the current strategy work well / adequately / poorly?
- What would need to change in the strategy for it to work in this scenario?
- What's the cost of being wrong about which scenario occurs?
Step 5: Identify Robust Actions
Find actions that perform well across multiple scenarios — these are the highest-confidence moves regardless of which future occurs.
Also find hedging options — small bets that preserve flexibility and cost little in the expected scenario but pay off in unlikely ones.
Output Format
🌐 Key Uncertainties
| Uncertainty | Why it matters | Range of outcomes |
|---|---|---|
| [Factor 1] | [Impact on strategy] | [From X to Y] |
| [Factor 2] | [Impact on strategy] | [From A to B] |
📖 The Scenarios
For each scenario (3–4 total):
Scenario Name (give it a memorable name)
- Description: 2–3 sentences painting the picture of this future
- Key conditions: What's true in this world?
- Probability estimate: Rough likelihood (should sum to ~100% across scenarios)
- Key signals: What early indicators would tell us we're in this scenario?
🧪 Strategy Stress Test
| Scenario | Current strategy performs... | Why | What must change |
|---|---|---|---|
| Scenario A | Well | [Reason] | Nothing |
| Scenario B | Adequately | [Reason] | [Adjustment] |
| Scenario C | Poorly | [Reason] | [Major pivot] |
| Scenario D | Catastrophically | [Reason] | [Fundamental rethink] |
🏆 Robust Actions
Actions that make sense across most or all scenarios:
- [Action 1] — works because [reason across scenarios]
- [Action 2] — works because [reason across scenarios]
🎯 Scenario-Specific Actions
If a particular scenario becomes likely, what additional actions should be taken?
| Scenario | Trigger signal | Response action |
|---|---|---|
| Scenario B | [Leading indicator] | [Action to take] |
| Scenario C | [Leading indicator] | [Action to take] |
🚨 Fragility Assessment
- Which scenario would the current plan handle worst?
- What's the single assumption the plan most depends on?
- What's the cheapest hedge against the worst scenario?
Scenario Planning Pitfalls
- Too many scenarios: 3–4 is ideal. More becomes unmanageable.
- Scenarios too similar: Each must be meaningfully distinct — different enough to require different strategies.
- Anchoring on the expected scenario: Give real attention to uncomfortable scenarios, not just the comfortable one.
- No early warning signals: Every scenario needs leading indicators — how do we know we're in it before it's too late to adapt?
- Planning for the scenario, not the strategy: The goal isn't to predict which scenario occurs. It's to build a strategy resilient enough to handle whichever one does.
Thinking Triggers
- "What are we implicitly assuming about the future in this plan?"
- "What's the world where this strategy fails completely — and how likely is it?"
- "What would we do differently if we knew we were in Scenario C?"
- "What are the early signals that would tell us which future we're heading into?"
- "What's the cheapest move that protects us in the bad scenarios without sacrificing the good ones?"
Example Applications
- Product roadmap: What if AI commoditizes our core feature? What if regulation changes the market? What if a major platform shifts behavior?
- Architecture decision: What if traffic grows 10x? What if the third-party API sunsets? What if the team doubles?
- Business strategy: What if the funding environment tightens? What if the key competitor undercuts pricing? What if adoption is 5x faster than projected?
- Agent pipeline design: What if LLM costs drop 90%? What if context windows become unlimited? What if a single agent becomes capable enough to replace the pipeline?