Market & Segmentation

Defining who you sell to and how coverage is assigned — and the test that separates a real segment from a reporting label.

The test of a segment

A segment is real when it justifies a different operating decision — a different motion, price, coverage model or support structure. If enterprise and mid-market accounts receive the same motion at the same price with the same coverage, the segmentation is a reporting label rather than a strategy.

This test matters because blended metrics across segments describe no segment. A single win rate spanning self-serve and enterprise is arithmetic, not information, and capacity planning built on it will be wrong for both.

The ICP has to be able to exclude

An ideal customer profile asserted in a planning session is a hypothesis. One derived from the observed characteristics of customers who buy quickly, stay and expand is evidence — and the second frequently contradicts the first, which is the point of doing it.

The test is whether it disqualifies. A definition no account can fail cannot route leads, prioritise territories or focus a team; it is a description of the market rather than a profile. The most useful and most commonly omitted half is the negative criteria: what makes an account a poor fit despite looking right. Derive those from churned customers, not from retained ones.

Where to look

Take your best twenty customers by retention and expansion, and your twenty worst by churn. The characteristics that separate the two lists are your ICP. Anything not present in that comparison is speculation.

Market sizing, and what each figure is for

Total addressable market is everyone with the problem regardless of whether you can reach them. Serviceable addressable market is the portion your product and go-to-market can serve today. Serviceable obtainable market is what you can plausibly capture in the planning horizon.

TAM is largely a fundraising number. SAM and SOM are the planning numbers — they determine territory design, coverage models and whether a quota is achievable in a given patch. A TAM built by multiplying an industry-wide figure by a guessed percentage is unfalsifiable; count addressable accounts and multiply by realistic contract value instead.

Vertical and horizontal, as a trade

Vertical focus buys a sharper fit, less competition and usually better retention, at the cost of a smaller addressable market. The narrower TAM is the point rather than a flaw: depth of fit is what makes the product hard to displace, and it makes the ICP unusually clear, which makes targeting and territory design considerably easier.

Horizontal breadth buys a larger market at the cost of a vaguer ICP — and a vague ICP degrades routing, scoring and territory design simultaneously. The temptation a large TAM creates is selling to anyone, which produces a customer base with no coherent pattern and makes both retention and product prioritisation harder.

Coverage

Territories should be balanced on potential rather than account count. A hundred accounts with no budget and twenty with real need are not equivalent patches. Overlaps surface as conflicts and get fixed; gaps surface as nothing at all, which is why they have to be looked for deliberately.

10 terms

COMMON QUESTIONS

How do you define an ideal customer profile?
Derive it from your own retained and expanding customers rather than asserting it. Compare your best twenty by retention against your worst twenty by churn — the characteristics separating the lists are the profile. Include negative criteria, which are usually the more useful half.
What is the difference between TAM, SAM and SOM?
TAM is everyone with the problem; SAM is the portion you can serve today; SOM is what you can realistically capture in the planning horizon. TAM is mostly a fundraising figure — SAM and SOM are what territory and quota design actually rest on.
When is a segment real rather than a label?
When it justifies a different operating decision — motion, pricing, coverage or support. If two segments get identical treatment, the split exists only in reporting and blended metrics will conceal problems in both.
How should sales territories be balanced?
On potential rather than account count, because equal counts with unequal potential produce attainment differences that look like performance differences. Audit deliberately for gaps as well as overlaps — gaps generate no complaint and so are never reported.

Course: RevOps Audit

See the course