Simulation Patterns Reference — Magi
Purpose: Apply horizon-specific simulation models without mixing cadence, assumptions, or output grain.
Contents
- Core simulation rules
- Short-term patterns
- Mid-term patterns
- Long-term patterns
- Quality checklist
Core Rules
- Always disclose assumptions.
- Always generate
Baseline / Optimistic / Pessimistic. - Always include sensitivity analysis on major drivers.
- Keep cadences separate:
SHORT = monthly or quarterlyMID = annualLONG = 3/5/10-year directional blocks
SHORT-TERM SIMULATION (0-1 Year)
Pattern ST-1: Monthly KPI Forecast
Use for MRR, ARR, customer-count, and churn-driven growth tracking.
Formula:
MRR_next = MRR_current × (1 + growth_rate) - churned_MRR + new_MRR + expansion_MRR
growth_rate = (new_customers × ARPU + expansion_revenue) / MRR_current - churn_rateRequired inputs:
- current MRR
- new customers per month
- ARPU
- churn rate
- expansion rate
- CAC
Default scenario shift:
Optimistic = +30%Pessimistic = -25%
Pattern ST-2: Cash Flow and Runway
Use for runway, burn, and fundraising timing.
Formula:
Monthly Burn = Fixed Costs + Variable Costs - Revenue
Runway = Cash Position / Monthly Burn
Break-even = Fixed Costs / Gross MarginRequired inputs:
- cash balance
- fixed costs
- variable costs
- revenue
- gross margin
Pattern ST-3: Crisis Response
Use when a discrete shock changes assumptions fast.
| Shock | Immediate response (0-30 days) |
Medium response (30-90 days) |
|---|---|---|
top customer churns (>20% of revenue) |
replace pipeline, cut spend | diversify concentration risk |
| major competitor enters | sharpen differentiation, lock in customers | invest in product separation |
| recession signal | protect retention, compress CAC | rework cost base |
| key employee leaves | retention action, temporary coverage | hiring and knowledge transfer |
| security incident | incident response and customer communication | trust recovery and prevention |
Pattern ST-4: Budget vs Actual
Use for monthly operating review and forecast correction.
Track:
- budget
- actual
- variance
- variance %
- cause
- recommended correction
MID-TERM SIMULATION (1-3 Years)
Pattern MT-1: Market Expansion
Use Ansoff-based strategy comparison:
- market penetration
- market development
- product development
- diversification
Minimum outputs:
- initial investment
- 3-year revenue uplift
- ROI
- risk
- recommendation
Pattern MT-2: Portfolio Plan
Use BCG-based resource allocation.
Track:
- current quadrant
- 1-year expected quadrant
- 3-year expected quadrant
- investment policy
Pattern MT-3: Organization and Hiring Roadmap
Track:
- hiring by function
- personnel cost
- personnel cost ratio
- purpose of each hiring wave
Pattern MT-4: P&L Forecast
Required parameters:
- revenue growth by scenario
- gross margin movement
- fixed-cost growth
- variable-cost ratio
Minimum outputs:
- revenue
- gross profit and margin
- SG&A
- R&D
- operating income and margin
- EBITDA
LONG-TERM SIMULATION (3-10 Years)
Pattern LT-1: Industry Transformation
Use for 5-10 year structure shifts.
Required outputs:
- three named scenarios
- explicit probabilities
- industry impact
- company opportunities and threats
- strategy valid across all scenarios
Pattern LT-2: M&A Scenario
Track:
- target
- valuation
- strategic rationale
- integration difficulty
- ROI / payback
- major risk
Pattern LT-3: Exit Strategy
Required inputs:
- current valuation method
- current value
- target exit timeline
Compare:
- IPO
- Strategic M&A
- PE
- continue independently
Pattern LT-4: Long-Range Financial Model
This is directional, not precision forecasting.
Required inputs:
- TAM and growth
- attainable share
- business model durability
- external shocks
Required outputs:
- current vs
3 / 5 / 10year metrics - optimistic downside and disruption case
- disclaimer that long-term numbers confirm direction, not exact results
Quality Checklist
- assumptions explicitly listed
- 3 scenarios generated
- sensitivity analysis at
±20% - time grain matches horizon
- benchmark comparison included when relevant
- roadmap is decomposable by Sherpa
- risk matrix included
- long-term output carries directional-not-precise disclaimer