Sensitivity
Analyze variable sensitivity for "$ARGUMENTS". Determine which inputs have the greatest impact on key financial outputs to prioritize risk management and decision-making.
Prerequisites
None — utility skill, run anytime. Works best when financial data exists in .metapowers/finance/$ARGUMENTS/.
Process
Gather context:
- If prior phase artifacts exist, read them for baseline assumptions
- Identify the output metrics to test (revenue, EBITDA, cash runway, valuation)
- Determine the base case values for all input variables
Select variables to test:
- Revenue drivers: growth rate, churn rate, ARPU, conversion rate
- Cost drivers: headcount, cost per hire, vendor costs, COGS
- Capital drivers: payment terms, collection period, funding timing
- Choose 6-10 variables with the most uncertainty
Run one-at-a-time analysis:
- For each variable, calculate the output when the variable moves +/- 10%, 20%, 30%
- Hold all other variables constant at base case
- Rank variables by impact magnitude (tornado chart data)
Run two-variable data tables:
- Pick the top 2-3 most impactful variables
- Create data tables showing output for combinations of values
- Identify "danger zones" where combinations produce unacceptable outcomes
Identify breakeven thresholds:
- For each key variable, find the value at which the business breaks even
- Calculate the margin of safety (how far current value is from breakeven)
Write the artifact to
.metapowers/finance/$ARGUMENTS/sensitivity.mdwith sections:- Variables Tested — list of inputs, base values, and ranges
- Tornado Chart Data — ranked impact of each variable on key outputs
- Data Tables — two-variable sensitivity matrices
- Breakeven Thresholds — critical values for each variable
- Risk Prioritization — which variables to monitor most closely
Output
The sensitivity analysis written to .metapowers/finance/$ARGUMENTS/sensitivity.md. Present a summary highlighting:
- Top 3 most impactful variables
- Breakeven thresholds for critical inputs
- Highest-risk combinations