Scenario Planning

Scenario Planning & Sensitivity Analysis

Build rigorous base/bull/bear scenarios with probability-weighted expected values, trigger identification, financial impact quantification, and strategic response planning per scenario.

Required Inputs

Execution Steps

Step 1: Identify Key Variables

Determine the 5-8 variables that most influence outcomes:

1. Revenue drivers: Market size, market share, pricing, volume, win rate, churn
2. Cost drivers: COGS, headcount, CAC, variable costs, CapEx
3. Market factors: Growth rate, competitive intensity, regulatory changes
4. Operational factors: Capacity, efficiency, time-to-market
5. External factors: Macroeconomic conditions, FX rates, commodity prices

Variable prioritization matrix:

Rule: Include variables scoring >=12 on Impact x Uncertainty. Typically 4-6 variables drive 80%+ of outcome variance.

Step 2: Define Scenario Framework

Build three core scenarios plus optional stress test:

Probability constraint: All scenario probabilities must sum to 100%.

Scenario construction rules:

1. Each scenario must be internally consistent (a world where all assumptions fit together)
2. Scenarios should differ on the KEY drivers, not every variable
3. Bear case is not "everything goes wrong" — it is the most likely bad outcome
4. Bull case is not "fantasy" — it is the most likely good outcome
5. Base case is the median expectation, not the optimistic plan relabeled

Step 3: Build Scenario Assumptions

For each key variable, define the value under each scenario:

Calibration check: Are the ranges supported by historical data, industry benchmarks, or analogues? If bear case has never happened in the industry's history, it may be too extreme (or not extreme enough if tail risks are real).

Step 4: Financial Impact Quantification

Build the P&L (or relevant financial model) for each scenario:

Multi-year projection (repeat for each year of time horizon):

Step 5: Probability-Weighted Expected Value

Calculate the expected value across scenarios:

Key insight: How does the expected value compare to the base case? If expected value is significantly below base case, the risk profile is skewed to the downside.

Step 6: Trigger Identification

For each scenario, identify what would cause it to materialize:

Monitoring cadence: [Weekly/Monthly/Quarterly] review of leading indicators against trigger thresholds.

Step 7: Scenario Tree (Decision Mapping)

Map key decision points and branching outcomes:

                            [Initial Decision]
                           /         |         \
                    [Path A]     [Path B]     [Path C]
                    /    \        /    \        /    \
              [Bull]  [Bear]  [Bull]  [Bear]  [Bull]  [Bear]
              p=[X]%  p=[X]%  p=[X]%  p=[X]%  p=[X]%  p=[X]%
              EV=$X   EV=$X   EV=$X   EV=$X   EV=$X   EV=$X

Decision rule: Choose the path with the highest expected value, subject to:

• Acceptable downside (bear case is survivable)
• Acceptable regret (if bull case materializes on unchosen path)
• Strategic optionality (path preserves future flexibility)

Step 8: Monte Carlo Considerations

For key variables with continuous distributions, consider Monte Carlo simulation:

1. Define probability distributions for each key variable:

   • Normal: For variables with symmetric uncertainty (e.g., market growth)
   • Log-normal: For variables that are bounded at zero (e.g., revenue)
   • Triangular: When you know min, most likely, and max
   • Uniform: When all values in a range are equally likely

2. Correlation matrix: Identify which variables move together (e.g., market growth and pricing power are often correlated)

3. Simulation outputs (if running Monte Carlo):

   • Mean and median outcome
   • Standard deviation
   • 10th percentile (downside) and 90th percentile (upside)
   • Probability of achieving target (e.g., P(revenue > $X) = Y%)
   • Value at Risk (VaR): What is the worst outcome at 95% confidence?

4. Simplified distribution table (when full Monte Carlo is not feasible):

   Outcome Metric	P10 (Downside)	P25	P50 (Median)	P75	P90 (Upside)
   Revenue	$[X]	$[X]	$[X]	$[X]	$[X]
   EBITDA	$[X]	$[X]	$[X]	$[X]	$[X]
   Cash flow	$[X]	$[X]	$[X]	$[X]	$[X]

Step 9: Strategic Response per Scenario

Define what actions to take under each scenario:

Pre-committed actions: For each scenario, define 2-3 actions that are pre-approved and can be executed immediately when triggers are hit, without additional deliberation.

Output Template

Scenario Analysis: [Business/Project] — [Decision Context]

Date: [Date] | Prepared for: [Client/Project] | Time Horizon: [X] years

1. Key Variables & Ranges

2. Scenario Narratives

Bull case ([X]% probability): [2-3 sentence narrative of what this world looks like]

Base case ([X]% probability): [2-3 sentence narrative]

Bear case ([X]% probability): [2-3 sentence narrative]

Stress test ([X]% probability): [2-3 sentence narrative]

3. Financial Impact Summary

(Include tables from Steps 4 and 5)

4. Probability-Weighted Expected Value

(Include table from Step 5)

5. Sensitivity Tornado

6. Trigger Dashboard

(Include table from Step 6)

7. Strategic Response Plan

(Include table from Step 9)

8. Decision Recommendation

Recommended path: [Decision recommendation] Expected value: $[X] Key risk: [Primary risk with mitigation] Decision reversibility: [Reversible / Partially reversible / Irreversible]

Quality Checks

• All scenario probabilities sum to exactly 100%
• Variable ranges are calibrated against historical data or industry benchmarks
• Each scenario tells a coherent, internally consistent narrative (not random variable combinations)
• Bear case is genuinely unfavorable, not just "slightly below base case"
• Financial impact is quantified in dollar terms, not just directional
• Probability-weighted expected value is calculated and compared to base case
• Triggers are specific, measurable, and time-bound (not vague)
• Leading indicators are identified for each trigger with monitoring cadence
• Strategic response for each scenario includes specific pre-committed actions
• Sensitivity tornado ranks variables by actual impact magnitude
• Stress test addresses existential risk (can the business survive?)
• Decision recommendation addresses reversibility and optionality
• Monte Carlo considerations address variable correlations, not just independent ranges