run-scenario-planning

Run Scenario Planning

Develop 2–4 coherent future scenarios that enable strategic decisions robust to uncertainty rather than anchored to a single forecast.

Why This Is Best Practice

Adopted by: Shell Oil (since 1970s), World Economic Forum, McKinsey & Company, RAND Corporation, and national defense planning agencies Impact: Shell's scenario planning team predicted the 1973 oil crisis and the fall of the Soviet Union years in advance, allowing Shell to outperform all major oil company competitors through those disruptions. Schoemaker (1995) documented that companies using scenario planning make better long-horizon capital allocation decisions and respond 40% faster to unexpected market shifts. Why best: Single-point forecasting fails because the future is not a linear extrapolation of the present. Scenario planning replaces a false sense of certainty with structured exploration of the possibility space, building organizational flexibility.

Sources: Schwartz "The Art of the Long View" (1991); Schoemaker "Scenario Planning: A Tool for Strategic Thinking" Sloan Management Review (1995); De Geus "Planning as Learning" HBR (1988)

Steps

1. Define the focal question — articulate the specific decision or time horizon the scenarios will inform: "What should our infrastructure strategy be through 2030?" Scenarios built without a focal question produce interesting stories that inform no decisions.

2. Identify key driving forces — brainstorm 15–25 external forces that will shape the future: technological, economic, sociopolitical, environmental, regulatory, and demographic. Use STEEP (Social, Technological, Economic, Environmental, Political) as a checklist.

3. Cluster into predetermined vs. uncertain factors — predetermined elements (aging population, climate change physics) are near-certain and belong in all scenarios. Uncertain factors (AI regulation, interest rate trajectories, geopolitical alliances) become the axes.

4. Select two critical uncertainties — from the uncertain factors, identify the two dimensions with: (1) highest potential impact on the focal question, and (2) highest genuine uncertainty. These two axes define a 2×2 scenario matrix.

5. Name the four scenario quadrants — each quadrant is the intersection of one axis state × one other axis state. Name each quadrant with a memorable label (not "Scenario A"). Names like "The Great Unraveling" or "Steady State" help teams reason with them as narratives.

6. Write scenario narratives — for each of the four quadrants, write a 1–2 page story set in the future year. Include: what happened, what the world looks like, how the industry changed, who won, who lost, and what your organization's situation is. Make them plausible, not preferred.

7. Identify implications for each scenario — for each scenario, answer: What decisions must we make today? What capabilities would we need? What would be different about our strategy? What would be the same?

8. Find robust strategies — identify strategic moves that are positive across all four scenarios (or at least three of four). These are the high-confidence investments. Identify moves that are only positive in one scenario — these require hedging or optionality.

9. Define early warning indicators — for each scenario, identify 3–5 leading indicators that would signal which scenario is materializing. Assign monitoring responsibility and cadence. This converts scenarios from a planning exercise into a living strategic radar.

10. Revisit annually — update the scenarios when driving forces shift materially. Run a 2-hour "scenario update" annually to recalibrate the probability distribution and adjust strategic posture.

Rules

• Scenarios must include at least one deeply uncomfortable future — if all scenarios look manageable, the exercise is too optimistic.
• Never assign probabilities to scenarios — this reintroduces the point-forecast fallacy that scenario planning exists to escape.
• Each scenario must be internally consistent — no scenario should contain contradictory elements.
• Do not build scenarios around preferred outcomes — this produces propaganda, not planning tools.
• The 2×2 matrix must use genuinely independent axes — correlated axes collapse into two scenarios, not four.

Common Mistakes

• Best/base/worst case framing — this is sensitivity analysis, not scenario planning; it assumes only one variable changes at a time.
• Too many scenarios — more than four scenarios overwhelm strategic decision-making; teams cannot hold more than four distinct futures in working memory.
• No action implications — scenarios that don't produce specific strategic decisions are academic exercises that waste executive time.
• One-time exercise — scenarios built in 2021 that haven't been updated through 2024 macro shifts are worse than useless — they create false confidence.

When NOT to Use

• Decisions with a horizon under 18 months (use financial modeling and sensitivity analysis instead)
• Decisions with no meaningful uncertainty (engineering specifications, legal compliance requirements)
• Teams without senior leadership buy-in (scenarios inform no decisions without executive sponsorship)