Segmentation

Market segmentation groups customers or prospects whose needs or behaviour are similar enough to support a distinct strategy. Demographic and firmographic categories can be practical, but needs-based segmentation seeks the deeper differences that shape choice, value and service requirements.

When to use it

• Identify customer groups whose differences should change the offer, channel, service or message.
• Choose which groups the organisation can serve distinctively and profitably.
• Avoid treating a large or familiar account category as if every member wants the same thing.
• Refresh market understanding when behaviour, technology or needs change.

Origins

Businesses have always differentiated among customers, but segmentation became a formal marketing concept as mass markets expanded. Wendell R. Smith’s 1956 article “Product Differentiation and Market Segmentation as Alternative Marketing Strategies” gave the field its defining formulation. He contrasted differentiation across a broad market with carving out a segment whose distinct preferences warranted a tailored offer, while showing that the two strategies can reinforce one another.

What it is

Demographic consumer variables include age, household or life stage; firmographic variables include size, industry and location. These characteristics make groups identifiable and reachable, but they do not necessarily explain why customers choose.

Needs, attitudes, jobs and behaviours can provide stronger strategic insight. A large business customer may demand a stripped-down low-cost offer or extensive technical support; size alone cannot distinguish the two.

Research should therefore ask not only what customers say they prefer but what problem they are solving, what trade-offs they make and what evidence appears in behaviour. Self-report is valuable but incomplete, especially for status, habit or sensitive motives.

Needs-based work often combines interviews and observation with survey analysis. Researchers may ask respondents to rate several statements, use factor analysis to identify related attributes and then use cluster analysis to group similar response patterns. With 200 interviews and many variables, validation is essential: algorithms will produce clusters even when the segments are unstable or strategically meaningless.

Useful segments should be:

• distinct in needs or response;
• measurable and large enough for the decision;
• reachable through a viable route;
• stable enough to act on but monitored for change; and
• compatible with a differentiated, economically sound response.

Developments of the model

Plotting segment attractiveness against the supplier’s ability to compete can support selection. A large “price fighter” segment may be unattractive if margins are poor, while a smaller group may justify investment because the company has a strong fit.

Quality fanatics

Segments that prioritise superior performance and are willing to recognise and pay for it may be attractive when the supplier can prove distinctive quality.

Traditionalists

Segments that value familiarity, continuity or trusted relationships require a different proposition from novelty-seeking groups.

Low

A low score on attractiveness or competitive position is a warning to deselect, redesign or gather better evidence; it is not a judgement about the customers themselves.

Needs-based segmentation is difficult in business-to-business markets because several people influence a purchase and the decision unit changes by situation. Digital personalisation can narrow groups dramatically, but more granularity is not automatically better. Privacy, discrimination, operational complexity and measurement error increase as segments become smaller.