Market Sizing Engine

Bottoms-up market sizing in Python. Model TAM, SAM, and SOM from real data instead of pulling numbers from analyst reports.

Why this exists

Every product strategy deck has a market sizing slide. Most of them are top-down guesses pulled from Gartner or Statista. This tool forces bottoms-up rigor: start with units, multiply by price, layer in assumptions, and stress-test them.

I built this because I got tired of seeing "$50B TAM" slides with no math behind them.

How it works

from market_sizer import MarketModel

model = MarketModel("US Telehealth — Behavioral Health")

# Define the funnel
model.set_population(
    total_addressable=330_000_000,  # US population
    filters=[
        ("Adults 18+", 0.78),
        ("Has mental health need (per NIMH)", 0.227),
        ("Has insurance or ability to pay", 0.87),
        ("Willing to use telehealth", 0.64),
    ]
)

# Revenue assumptions
model.set_revenue(
    avg_sessions_per_year=12,
    avg_revenue_per_session=150,
)

# Run
model.calculate()
model.summary()

Output:

┌─────────────────────────────────────────────────────┐
│  US Telehealth — Behavioral Health                  │
├─────────────────────────────────────────────────────┤
│  TAM    $11.38B   330.0M × 22.7% × $150 × 12      │
│  SAM    $6.36B    TAM × 87.0% insured × 64.0% willing │
│  SOM    $63.6M    SAM × 1.0% capture rate           │
├─────────────────────────────────────────────────────┤
│  Sensitivity: ±18% at 90% CI                        │
└─────────────────────────────────────────────────────┘

Features

• Bottoms-up modeling — Start with population, apply filters, multiply by revenue
• Sensitivity analysis — Monte Carlo simulation on every assumption
• Scenario comparison — Model bull/base/bear cases side by side
• Export — Markdown tables, JSON, or CSV for your deck

Install

pip install -r requirements.txt

Usage

CLI

# Interactive mode
python market_sizer.py

# From a config file
python market_sizer.py --config examples/telehealth.yaml

# Compare scenarios
python market_sizer.py --config examples/telehealth.yaml --scenarios bull,base,bear

As a library

from market_sizer import MarketModel, Scenario

base = Scenario("Base", capture_rate=0.01)
bull = Scenario("Bull", capture_rate=0.03)
bear = Scenario("Bear", capture_rate=0.005)

model = MarketModel("My Market")
# ... configure ...
model.compare([base, bull, bear])

Project structure

market-sizing-engine/
├── market_sizer.py       # Core engine
├── models/
│   ├── __init__.py
│   ├── population.py     # Population funnel modeling
│   ├── revenue.py        # Revenue calculation
│   └── sensitivity.py    # Monte Carlo sensitivity analysis
├── examples/
│   └── telehealth.yaml   # Example config
├── tests/
│   └── test_market_sizer.py
├── requirements.txt
└── README.md

The math

TAM = Total addressable population × Revenue per unit per year

SAM = TAM × Serviceable filters (geography, willingness, access)

SOM = SAM × Realistic capture rate

Sensitivity = Monte Carlo simulation (10,000 runs) varying each assumption ±20% with uniform distribution. Reports 90% confidence interval.

License

MIT