---
name: "opportunity-sizing-and-business-case"
description: "Use when the user asks how big an opportunity is, names TAM, SAM, or SOM, or needs an investment case for a GTM bet: the sized opportunity with stated math and the case with returns and risks."
triggers: ["TAM", "SAM SOM", "market size", "how big is this", "business case", "investment case", "payback"]
version: "1"
---
# Opportunity sizing and business case
Use this when the user asks how big an opportunity is, names TAM, SAM, or
SOM, or needs an investment case for a GTM bet. Start from the segment or
market in scope, the pricing and conversion assumptions, and the
investment asked for.
## Bound the market in three bands
TAM (everyone with the problem), SAM (those you can reach), SOM (those you
can win in the horizon). Each band gets its method, sources, and date —
top-down, bottom-up, or value-theory, said plainly. When Apollo is
connected, the bottom-up count of accounts matching the ICP firmographics
is a real number, not a guess; use it and cite the filter.
## Triangulate before you trust
At least two methods, and name where they disagree. The gap between
methods is the honest range; a case built on one method says so. Flag a
bottom-up vs top-down delta over 3x, and treat SOM capture above 5% in
year one as needing extraordinary proof — 1–5% is the usual range for a
new entrant.
## Build the business case
Investment, expected return, payback, and the net revenue retention,
LTV/CAC, or pipeline math that carries it. State the denominator behind
every rate and the benchmark you judge it against.
## Stress it three ways
Base, Adverse, Opportunity. Each grade gets the driver that breaks it and
the leading indicator that warns you first. Never silently rebaseline — a
reforecast says what changed and why.
## What you hand back
The verdict and the risks: fund, fund-with-conditions, or pass, each with
the evidence grade behind it. You build the case; finance validates the
economics before money moves. Save under sizing, one folder per case with
the model — a sheet when Sheets is connected — the assumptions log, and
the verdict.
## Fallbacks
No hard data means a triangulated range with the method named, never a
single false-precise number. No conversion baseline means you size the
ceiling and flag the case as direction-only. Never invent a market size,
a growth rate, or a customer count. Label each line FACT, INFERENCE, or
UNKNOWN.