# Shell Method Scenario Planning

> Navigate strategic uncertainty through multiple plausible future scenarios rather than single-point forecasting. Developed at Royal Dutch Shell, this framework replaces single-point predictions with 2-4 rich narratives representing genuinely different futures defined by critical uncertainties.

- Skill: `lev-os/shell-method-scenario-planning` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add lev-os/shell-method-scenario-planning`
- Raw SKILL.md: https://api.skillmd.com/api/skills/lev-os/shell-method-scenario-planning/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: lev-os (https://skillmd.com/u/lev-os)
- Updated: 2026-09-10
- Page: https://skillmd.com/skills/lev-os/shell-method-scenario-planning

---


# Shell Method (Scenario Planning)

## Purpose

The Shell Method is a scenario planning framework for navigating long-term strategic uncertainty. Rather than predicting "the" future, create 2-4 rich narratives representing genuinely different future states. This approach enables identification of robust strategies that work across multiple futures and contingent strategies for specific scenarios.

**Historical validation**: Successfully predicted the 1973 oil crisis, enabling Shell to rise from 7th to 2nd largest oil company while competitors operated on business-as-usual assumptions.

## When to Use This Skill

This skill should be used when:

- **Facing strategic uncertainty at 5-30 year horizons** - When long-term planning is needed and experts disagree about critical factors
- **Single forecasts prove unreliable** - When business-as-usual assumptions are being challenged by emerging uncertainties
- **High-stakes decisions with irreversible commitments** - When major investments or strategic pivots require stress-testing against multiple futures
- **Experts fundamentally disagree** - When domain experts provide conflicting predictions or data conflicts emerge
- **Industry disruption is possible** - When technological, regulatory, or market shifts could fundamentally reshape the competitive landscape
- **Need shared organizational mental models** - When coordinated response is required without central command during uncertainty

**Semantic triggers**: "scenario planning", "strategic foresight", "multiple futures", "uncertainty mapping", "long-term strategy", "what-if analysis for strategy"

## Core Approach

**Central distinction**: Separate "predetermined elements" (high-confidence trends like demographics, infrastructure, enacted laws) from "critical uncertainties" (high-impact unknowns where experts disagree).

**Key insight**: Map uncertainties on a 2×2 matrix and develop narrative scenarios for each quadrant to identify:
- **No-regret moves**: Strategies that work across all futures
- **Contingent moves**: Actions triggered by specific scenario indicators

## Process

### Step 1: Identify Focal Question

Define the strategic decision and planning horizon:

1. **Set planning horizon** - Typically 5-30 years for strategic uncertainty
2. **List critical unknowns** - Technologies, regulations, competitors, customer behavior
3. **Frame as focal question** - "What should our strategy be given [uncertainty]?"

**Questions to answer**:
- What strategic decision are we facing?
- What do we need to know but don't?
- What uncertainties could make or break our strategy?

**Example**: Shell 1971: "How should we position ourselves given potential energy supply disruptions and price volatility over the next 10 years?"

### Step 2: Research Elements and Uncertainties

Separate what is known from what is genuinely uncertain:

1. **Identify predetermined elements** - Trends with high confidence (demographics, infrastructure momentum, laws already passed)
2. **Identify critical uncertainties** - High-impact factors where experts disagree or data conflicts
3. **Interview domain experts** - Analyze data, identify divergent viewpoints
4. **Test for genuine uncertainty** - If experts agree, it's predetermined; if they disagree, it's uncertain

**Deliverables**:
- List of predetermined elements ("We know...")
- List of critical uncertainties ("We don't know...")

**Example**: Shell 1971 - Predetermined: Oil production has physical limits, demand growing. Uncertain: When will supply/demand cross? How will governments respond?

### Step 3: Select Two Critical Uncertainties

Choose two independent uncertainties for a 2×2 matrix:

1. **Rank uncertainties** - By strategic impact and unpredictability
2. **Select two independent factors** - Not correlated with each other
3. **Define clear endpoints** - For each axis (e.g., "Loose regulation" vs "Tight regulation")
4. **Validate quadrants** - Ensure four distinct, meaningful future states

**Quality checks**:
- Are these uncertainties genuinely unpredictable (not just unknown)?
- Are the two axes independent of each other?
- Do the four quadrants represent meaningfully different futures?

**Output**: 2×2 matrix with four quadrants representing distinct future states

**Example**: Shell - Axis 1: Oil prices (low vs high), Axis 2: Government intervention (minimal vs heavy)

### Step 4: Develop Rich Narratives

Create detailed stories for how each quadrant future unfolds:

1. **Write 3-5 page narrative for each quadrant**
2. **Include causal chains** - Triggering events, stakeholder responses, second-order effects
3. **Give memorable names** - e.g., "Wild West", "Regulated Utility", "Climate Crisis"
4. **Use storytelling techniques** - Make vivid, coherent, plausible
5. **Ensure internal consistency** - Each scenario must pass logical scrutiny

**Story elements**:
- What event could trigger this future?
- How would key stakeholders respond?
- What second-order effects cascade from initial changes?
- What does daily life/business look like in this future?

**Output**: 2-4 detailed scenario narratives (3-5 pages each) with memorable names

**Example**: Shell 1973 Type A scenario: Technical extraction limits → supply shortage → price spike → economic shock → geopolitical realignment

### Step 5: Test Strategy Against Scenarios

Evaluate current and alternative strategies across all scenarios:

1. **Identify current strategy assumptions** - What future is the current plan betting on?
2. **Stress-test across scenarios** - How does current strategy perform in each quadrant?
3. **Identify vulnerabilities** - Which scenarios break the current strategy?
4. **Design strategy variants** - Create alternatives optimized for different scenarios

**Analysis questions**:
- Does our current strategy only work in one scenario?
- Which scenarios would cause strategic failure?
- What early indicators would signal which scenario is unfolding?

**Output**: Strategy vulnerability analysis across scenarios

### Step 6: Develop Robust and Contingent Strategies

Design strategies that account for multiple futures:

1. **Identify no-regret moves** - Actions that improve outcomes across ALL scenarios
2. **Design contingent strategies** - Moves triggered when specific scenarios unfold
3. **Define leading indicators** - Metrics that signal which scenario is materializing
4. **Establish decision triggers** - Conditions that activate contingent strategies
5. **Monitor for scenario shifts** - Ongoing surveillance for early warning signals

**Strategy types**:
- **Robust/no-regret**: Invest in adaptive capacity, diversification, scenario monitoring systems
- **Contingent**: Pre-plan responses but delay execution until scenario clarity emerges

**Deliverables**:
- List of no-regret moves (execute now)
- List of contingent strategies (execute if scenario X unfolds)
- Dashboard of leading indicators
- Decision triggers for strategy pivots

**Example**: Shell post-1973 - No-regret: Build scenario planning capability. Contingent: If supply constraints emerge, pre-position for scarcity market.

## Practical Techniques

### Technique 1: Type A vs Type B Scenarios

**Purpose**: Distinguish between psychological reframing and genuine uncertainty exploration.

**Process**:
1. **Type A scenarios** - Uncomfortable but plausible futures that challenge business-as-usual mental models
2. **Type B scenarios** - Variations on current trends (less transformative)
3. **Focus on Type A** - These provide the most strategic value by breaking groupthink

**Example**: Shell's Type A (supply crisis) vs Type B (steady growth continuation)

### Technique 2: Predetermined Elements Filter

**Purpose**: Avoid treating known trends as uncertainties.

**Process**:
1. List all factors affecting the focal question
2. Test each: "Do credible experts disagree fundamentally?"
3. If NO → Predetermined element (incorporate into all scenarios)
4. If YES → Critical uncertainty (becomes scenario differentiator)

**Benefit**: Reduces scenario complexity by removing false uncertainties

### Technique 3: Early Warning Indicator System

**Purpose**: Detect which scenario is materializing as the future unfolds.

**Process**:
1. For each scenario, identify unique leading indicators
2. Monitor these indicators monthly/quarterly
3. When multiple indicators align, activate contingent strategy
4. Update scenarios annually as predetermined elements increase

**Example**: Monitoring oil inventory levels, OPEC statements, alternative energy investment as signals

## Common Pitfalls

Avoid these anti-patterns:

1. **Probability assignment** - Do NOT assign probabilities to scenarios; maintains openness to surprises
2. **Best/worst/baseline framing** - Creates confirmation bias toward baseline; use neutral names
3. **Too many scenarios** - 2-4 is optimal; more creates analysis paralysis
4. **Single-variable scenarios** - Need at least two independent uncertainties for richness
5. **Insufficient narrative detail** - Thin scenarios don't change mental models; aim for 3-5 pages each
6. **No strategy implications** - Scenarios without strategic testing are just stories; must drive decisions

## Integration

**Complements**:
- Pre-mortem analysis (test strategy failures in each scenario)
- OODA loop (rapid adaptation when scenario signals emerge)
- Red team/blue team exercises (stress-test scenario narratives)
- Strategic roadmapping (timeline no-regret vs contingent moves)

**Conflicts with**:
- Single-point forecasting methodologies
- Deterministic planning approaches
- Short-term optimization frameworks

**Leads to**:
- Dynamic strategy portfolios
- Adaptive capacity building
- Resilience-oriented decision-making

## Time Estimates

**Full process**: 3-6 months for comprehensive organizational scenario planning
**Rapid sprint**: 1-2 weeks for simplified scenario exploration
**Annual refresh**: 1-2 weeks to update scenarios and indicators

**Complexity**: High - Requires cross-functional input, expert interviews, and organizational buy-in

## Reference

**Authors**: Pierre Wack, Ted Newland, Peter Schwartz (Royal Dutch Shell, 1960s-1990s)

**Category**: Strategic foresight, decision-making under uncertainty

**Historical context**: Developed at Royal Dutch Shell to navigate oil industry volatility, gained legendary status after predicting 1973 oil crisis.
## Technique Map

- **Predetermined vs critical uncertainties** — Separate high-confidence trends from high-impact unknowns where experts disagree; because scenario value comes from exploring genuine uncertainty, not known trends.
- **2×2 matrix** — Two independent uncertainties as axes; four quadrants = four distinct futures; because 2-4 scenarios is optimal; more creates analysis paralysis.
- **Rich narratives (3-5 pages each)** — Include causal chains, memorable names, storytelling; because thin scenarios don't change mental models.
- **Type A over Type B** — Focus on uncomfortable-but-plausible futures; because Type B (variations on current trends) provides less strategic value.
- **No-regret + contingent strategies** — Identify moves that work across all futures and moves triggered by specific scenario indicators; because robust planning requires both.

## Technique Notes

Never assign probabilities to scenarios. Use neutral names, not best/worst/baseline. Test strategy against all scenarios. Define leading indicators for each. Update annually as predetermined elements increase. Full process: 3-6 months; rapid sprint: 1-2 weeks.

---

## Prompt Architect Overlay

**Role Definition:** Shell Method scenario planning facilitator. Navigates strategic uncertainty through 2-4 rich narratives representing genuinely different futures. Replaces single-point forecasting with contingent strategy design.

**Input Contract:** Accepts 5-30 year strategic horizon, high-stakes decisions, expert disagreement, industry disruption possibility, or "what-if analysis for strategy." Focal question and planning horizon required.

**Output Contract:** 2×2 matrix with quadrants, 2-4 detailed scenario narratives (3-5 pages each), strategy vulnerability analysis, no-regret moves list, contingent strategies with triggers, leading indicator dashboard.

**Edge Cases & Fallbacks:** If experts agree on factor→treat as predetermined, not uncertainty. If too many uncertainties→rank by impact and unpredictability; select two independent. If user wants probabilities→refuse; maintains openness to surprises.