Home › Guides

Market-sizing guides

Everything below is the method behind a SectorIntel report, written out in plain English with worked examples you can reproduce on paper. No black boxes, no paywalled research — this is textbook practice, which is exactly the point: a market size only holds up if the person reading it can rebuild it.

Foundations

How to calculate TAM, SAM, and SOM (with a worked example)

The full funnel, step by step: buyer universe → served market → reachable slice, with the exact arithmetic our engine runs on a warehouse-robotics example.

Read the guide →
Method

Top-down vs bottom-up market sizing: which to trust

Why each method fails on its own, how they anchor to different public numbers, and how a cross-check catches an inflated TAM before a committee does.

Read the guide →
Method

Triangulating a market size across three methods

Why one number is a single point of failure, how the median beats the average, and how a spread ratio turns disagreement into an honest confidence band.

Read the guide →
Method

The value-theory (value-pool) sizing method

Sizing a market by the economic value it removes — labour saved, losses avoided — and the fraction a vendor can actually charge for. The third leg of the triangle.

Read the guide →
Discipline

Five sanity checks that catch a wrong market size

The automated gates every figure passes before it ships: funnel monotonicity, value-pool containment, plausible capture share, price coherence, and a realistic SOM.

Read the guide →
Framework

Competitor-mapping frameworks that survive an IC

The 2×2 quadrant done properly, why the status-quo row is mandatory, and how to score players from public evidence without hand-waving.

Read the guide →
Sourcing

Where citable market claims come from

The public-source hierarchy — official statistics, filings, standards bodies — ranked by trust, plus the rule that a figure without a source never ships.

Read the guide →
Application

Sizing a market for a board deck or investment committee

What a committee actually challenges, the eight-part structure that pre-empts every question, and how to present a range instead of false precision.

Read the guide →
Want the math run on your sector instead of a sample? Try the free instant estimate — it runs a simplified version of this engine on your own inputs in ten seconds — or commission a cited report where every assumption is replaced by a dated public source.