market-sizing-tam-sam-som
Sizes TAM, SAM, and SOM using independent top-down and bottom-up methods, narrowing each step with explicitly named constraints (geography, segment, go-to-market capacity) rather than an arbitrary per
SKILL.mdopportunity-recognition/skills/market-sizing-tam-sam-som/SKILL.md
GitHubMarket Sizing (TAM/SAM/SOM)
Purpose
Quantitatively sizes the market and the genuinely reachable share of it — not to produce an impressively large headline number, but to give a defensible, traceable estimate that a business case or investment conversation can actually stand on.
Anchored in research
- The standard TAM/SAM/SOM framework (Total Addressable Market / Serviceable Addressable Market / Serviceable Obtainable Market), widely used in venture and strategic planning to move from "the size of the whole market" down to "what we could realistically capture" in three explicit, progressively narrower steps.
- stratarts — market-opportunity-analyzer (a public reference implementation of a market-sizing workflow)
- Business Opportunity Scout (buYoung) (a public reference implementation of opportunity scouting and sizing)
Method
- Define TAM (Total Addressable Market) using two independent
approaches, not one.
- Top-down: start from a recognized industry or market-research figure for the broadest relevant category, and state the source.
- Bottom-up: start from a realistic unit economic (price per
customer, or revenue per transaction) and multiply by a defensible
estimate of the total number of potential buyers.
Running both and comparing them is the single most effective way to
catch an inflated or nonsensical TAM before it enters a business case —
see
../agents/market-sizing-cross-validator.md, which exists specifically to do this cross-check.
- Narrow TAM to SAM (Serviceable Addressable Market) by applying the real constraints of the business: geography actually served, product segment actually addressed, language/regulatory/channel constraints. Every narrowing filter should be named explicitly — "we only serve [region/segment] because [reason]" — not folded silently into a lower percentage.
- Narrow SAM to SOM (Serviceable Obtainable Market) using realistic go-to-market constraints over a defined time horizon: sales capacity, competitive response, adoption curve, channel access. SOM should be expressed with an explicit time frame ("reachable within 3 years"), not as a static number — reachable share changes as the business matures.
- Make every multiplier and percentage traceable to its source. For
each number used in the chain, note whether it's: user-supplied, an
industry benchmark (name the source), an assumption
(
[assumption — verify]), or pulled from a connected external data source (see../../../meta/external-data-mcp.md). A TAM/SAM/SOM chain is only as credible as its weakest, least-traceable multiplier. - Watch for double counting when multiple narrowing factors are chained together (e.g. "30% of the market × 15% conversion") — check whether the two percentages are actually independent of each other, or whether one already implicitly contains the other.
- Sanity-check the result against an external anchor if one is available: a known industry revenue figure, a comparable company's disclosed revenue, or a connected external data source. This is a magnitude check, not a precise validation — is the number the right order of magnitude, or off by 10x without an explanation?
- Route the result through cross-validation before it's used. Before
the TAM/SAM/SOM figures go into a business case or in front of a
decision-maker, run them through
../agents/market-sizing-cross-validator.md, which independently re-derives the number using a second method and reports a confidence level.
Available scripts
scripts/market_sizing.py-- computes TAM from top-down and/or bottom-up inputs (Method step 1), flags a divergence when the two methods disagree by more than a configurable ratio (default 3x), then narrows to SAM and SOM using explicitly named filters/constraints (steps 2-3) -- from EACH available TAM basis, so you see a range instead of one arbitrarily-picked number. Run it once you have at least one TAM input; supplying both top-down and bottom-up is strongly preferred, since the cross-check is the point.python3 scripts/market_sizing.py --example > input.json # see the input shape python3 scripts/market_sizing.py input.jsonStdlib only (json, argparse) -- no install needed. Warns (not errors) if a divergence is detected; errors only on missing/invalid required fields.
What this skill does NOT do
- Doesn't make the final decision for you — it produces a structured draft to support a human decision.
- Doesn't confirm figures, market data, or competitor data from memory — it
uses the inputs you provide, or clearly flags an assumption
(
[assumption — verify]). - Doesn't rely on a single calculation path as sufficient proof — a
top-down-only or bottom-up-only estimate should be treated as provisional
until cross-checked by the other method or by
../agents/market-sizing-cross-validator.md.
Refinement notes
Areas to keep deepening with real practice:
- your own rules of thumb and heuristics for this technique
- concrete templates (into
../../references/) - reference cases / your own examples
- what this skill deliberately does not do (guardrails, common mistakes) — add to the list above
Once this section is filled in and validated in practice, update the
maturity field in skills_index.json to draft, validated, or
canonical (see ../../../meta/maturity_levels.md). Don't add new fields
to the frontmatter — name and description are the only ones allowed
(see ../../../meta/frontmatter_schema.md).
Continue from here
- Next in this pack:
../competitive-and-five-forces-mapping/SKILL.md— Maps the competitive dynamics and structural forces of the industry. - This pack's shared guardrails:
../../CLAUDE.md
References
../../references/— the pack's shared background material../../CLAUDE.md— this pack's shared guardrails../../../meta/external-data-mcp.md— an optional external data MCP (Market Sizing MCP Server / TAM-MCP-Server) as underlying data for the calculation or for cross-checking, if one is connected in the user's environment. Not a dependency — the skill works without it.../agents/market-sizing-cross-validator.md— a delegatable agent that cross-checks the calculation this skill produces before the figure is used in a business case