Market Segment

Origin & evolution

The modern concept was named by Wendell R. Smith in Product Differentiation and Market Segmentation as Alternative Marketing Strategies, published in the Journal of Marketing in 1956. Smith argued that demand is diverse, not homogeneous, and that a firm faces a choice: differentiate the product to bend demand toward a single offering, or segment the market and tailor offerings to its natural divisions. He framed segmentation as a deliberate response to heterogeneity rather than an accident of it. That paper is the citation every later treatment traces back to.

Practice then split segmentation into bases. Demographic and, for business markets, firmographic bases describe buyers by observable attributes such as age, income, company size, or industry. They are cheap to measure and easy to act on, which is also their weakness: two companies of identical size can want completely different things. Needs🫀NeedUserA user need, pain, desire, or constraintView reference →-based and behavioural segmentation cut instead by what buyers are trying to do and how they act, which predicts purchasing better but costs more research to build. The field's working consensus is that observable bases are useful proxies, and needs-based bases are the ones that actually explain why someone buys.

Clayton Christensen's Jobs🔨JobUserJob To Be Done: what the user is trying to accomplishView reference →-to-Be-Done argument in Competing Against Luck pushes this further: the relevant unit of demand is not a demographic cohort👥CohortGrowthA group of users sharing a common characteristicView reference → but a functional, emotional, and social job a buyer is trying to get done, with demographic and firmographic attributes serving only as circumstantial context. By that reading, a segment built on observable proxies may cluster the wrong buyers together — two companies of identical size and industry may be hiring for entirely different jobs — while a needs-based segment that groups buyers by the job they share will more reliably predict who buys and why.

Segmentation also acquired a sizing discipline. The TAM, SAM, SOM hierarchy nests three figures: the total addressable market is all demand for the category, the serviceable available market is the slice your offering and reach can actually serve, and the serviceable obtainable market is the share you can realistically win near term. Segmentation is how a credible SAM gets drawn, because a SAM is a TAM filtered down to the segments you have chosen to serve.

How it works in practice

A company selling expense-management software starts with a TAM of every business that processes expenses, a number in the millions. That figure is too coarse to plan with. They segment firmographically by headcount and find three groups behave differently: sole traders want zero setup, mid-market finance teams want approval workflows, and enterprises want procurement integration. The mid-market segment, firms of 50 to 500 staff, has roughly 80,000 companies in their geography, an average contract value that supports a sales-assisted motion, and the least competition. That becomes the SAM.

Naming that segment forces honest trade-offs. The approval-workflow features⭐FeatureProduct SpecificationA product capability or featureView reference → that win mid-market deals add friction the sole trader hates, so the team cannot serve both well with one product. Positioning sharpens immediately, because "expense software for growing finance teams" is a claim a buyer can place themselves inside, and the sole trader self-selects out without a wasted sales call.

Geoffrey Moore's Crossing the Chasm makes segment choice a survival decision rather than a targeting preference. Moore argues that technology products face a structural gap between early adopters, who tolerate incompleteness for the promise of an advantage, and the pragmatist early majority, who want a complete solution🔧SolutionDiscoveryA proposed approach to address an opportunityView reference → and buy from market leaders. Crossing that gap requires concentrating on a single beachhead segment — one specific, bounded group — winning it completely enough to become a reference customer before expanding. The implication for sizing is that a plausible SAM must be a segment you can dominate, not merely enter; diffuse targeting across several segments at once typically fails to build the reference base that the early majority demands.

Market segment vs. its neighbours

• Persona🎭PersonaUserAn archetype representing a user segmentView reference → describes one individual by goals and behaviour, for design decisions. A segment describes a population by shared attributes, for targeting and sizing decisions. A segment can contain several personas, since the same group of companies may hold a buyer, a user, and a blocker who each need different things.
• Ideal Customer Profile👤Ideal Customer ProfileGo-To-MarketThe ideal customer profile (ICP)View reference → narrows further to the accounts most worth acquiring, a go-to-market construct tuned for fit and retention. A segment is broader and more descriptive; an ICP is the high-value core you draw out of it. Every ICP sits inside a segment, and not every segment is worth an ICP.
• Behavioural segment is a segment cut specifically by what users do, such as usage frequency or feature adoption, rather than who they are. It is a segment defined on action, useful once you have a product generating behaviour to observe.

In the graph

In the Unified Product Graph, a market segment anchors the market intelligence region and acts as the join between sizing and design. Personas live inside it (market_segment_includes_personaMarket SegmentincludesPersonacross-domain, persona_belongs_to_market_segmentPersonabelongs toMarket Segmentcross-domain), business models💰Business ModelBusiness ModelThe business model canvas or definitionView reference → point at it (business_model_targets_market_segmentBusiness ModeltargetsMarket Segmentcross-domain), positioning is scoped to it (positioning_within_market_segmentPositioningwithinMarket Segmentcross-domain), an ideal customer profile sharpens it (ideal_customer_profile_targets_market_segmentIdeal Customer ProfiletargetsMarket Segmentcross-domain), and competitive analysis📊Competitive AnalysisMarket IntelligenceA structured analysis of the competitive landscapeView reference → bounds its study to it (competitive_analysis_scopes_market_segmentCompetitive AnalysisscopesMarket Segmenthierarchy). Holding the segment as a node keeps the "who we sell to" decision queryable, so a persona with no segment, or a segment no business model targets, shows up as an unfinished link rather than an assumption🤔AssumptionStrategyA belief taken as true that underpins a strategyView reference → nobody wrote down.