Market Sizing

Market Sizing

Produces a structured TAM/SAM/SOM market sizing analysis using web research, industry data, and bottom-up calculations.

When to Activate

• User says "calculate TAM", "size the market", "market opportunity"
• User asks "how big is the market for X?"
• User needs market sizing for a pitch deck, PRD, or business case

Input

Required:

• Product/service description: What is being sized
• Target customer: Who buys this

Optional:

• Geography: Default is global
• Time horizon: Default is current year + 5-year projection
• Methodology preference: top-down, bottom-up, or both (default: both)

Execution

Context Check

Before sizing, read context/product.md and context/competitors.md if they exist. Existing product positioning and competitive landscape inform market definition and assumptions. After completing the analysis, offer to update context/competitors.md with market landscape findings.

Step 1: Define the Market

Clarify with the user if ambiguous:

• What problem is being solved?
• Who are the target customers (segment, size, industry)?
• What is the product/service category?
• What is the geographic scope?

Step 2: Research (Parallel)

Use web search tools to gather data from multiple sources:

| Data Type | Sources to Search | |:----------|:-----------------| | Industry reports | Gartner, Forrester, IDC, Statista, Grand View Research | | Company data | Public filings, earnings calls, analyst reports | | Customer data | Census data, industry associations, trade publications | | Competitive landscape | Crunchbase, PitchBook, competitor pricing pages | | Growth rates | Historical CAGR, analyst projections |

Step 3: Calculate TAM (Two Methods)

Top-Down:

1. Find total market category size from research reports
2. Document source, year, and methodology
3. Apply growth rate to current year if data is older

Bottom-Up:

1. Count total potential customers in target segment
2. Determine average annual revenue per customer (pricing research)
3. TAM = Total Customers x Annual Revenue per Customer

Step 4: Calculate SAM

Narrow TAM by applying realistic filters:

• Geographic constraints
• Product capability constraints
• Customer segment constraints
• Distribution/channel constraints

SAM = TAM x (% matching all filters)

Step 5: Calculate SOM

Estimate realistic market capture over 3-5 years:

• New entrant: 2-5% of SAM
• Established player entering adjacent market: 5-10% of SAM
• Account for competitive intensity and switching costs

Step 6: Validate

Cross-check top-down vs bottom-up (should be within 30%). Flag if they diverge significantly.

Output Format

# Market Sizing: {Product/Service}

## Market Definition
- **Problem:** {problem being solved}
- **Customer:** {target segment}
- **Category:** {market category}
- **Geography:** {scope}

## TAM: {$X.XB}
### Top-Down
- {Source}: {total market size} ({year})
- Growth rate: {X%} CAGR
- **TAM (top-down):** {$amount}

### Bottom-Up
- Total addressable customers: {count} ({source})
- Average annual revenue per customer: {$amount} ({source})
- **TAM (bottom-up):** {$amount}

### Triangulation
{How the two methods compare. Flag if >30% divergence.}

## SAM: {$X.XB}
| Filter | Reduction | Rationale |
|:-------|:----------|:----------|
| {Geographic} | {X%} | {why} |
| {Segment} | {X%} | {why} |
| {Product fit} | {X%} | {why} |
**SAM:** TAM x {combined %} = {$amount}

## SOM: {$X.XM} (Year 3-5)
- Competitive intensity: {high/medium/low}
- Realistic capture rate: {X%}
- **SOM (Year 3):** {$amount}
- **SOM (Year 5):** {$amount}

## Confidence Assessment
- **Data quality:** {high/medium/low} - {why}
- **Key assumptions:** {list the 2-3 most impactful assumptions}
- **Sensitivity:** {what changes if key assumptions shift by 20%}

## Sources
1. {Source with URL or citation}
2. {Source with URL or citation}

Rules

• Two methods minimum: Always calculate both top-down and bottom-up.
• Cite every number: No unsourced figures. If estimated, say so explicitly.
• Show your math: Every calculation should be traceable.
• Conservative by default: Use lower bounds when data is ambiguous.
• Date the data: Always note the year of each data source.
• Flag assumptions: Tag with [Assumption]. Every non-trivial assumption gets called out.
• Tag unverified data: Use [Needs data] for estimates without hard sources, [Source: X] for cited figures.
• No vanity TAMs: If the TAM seems unreasonably large, challenge the market definition.

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: mshadmanrahman
• Source: mshadmanrahman/pm-pilot
• License: MIT
• Homepage: https://github.com/mshadmanrahman/pm-pilot#readme

Install and usage instructions live in the source repository linked above.