In the snowmobile example, the served market consisted of one segment. But conceivably, the served market could be much broader in scope. For example, the company could decide to serve all industrial customers (large, medium, small) by offering diesel-driven snowmobiles for delivery use. The “broader” served market, however, must be segmented because the market is not homogeneous; that is, it cannot be served by one type of product/service offering. Currently, the United States represents the largest market in the world for most products; it is not a homogeneous market, however. Not all customers want the same thing. Particularly in well-supplied markets, customers generally prefer products or services that are tailored to their needs. Differences can be expressed in terms of product or service features, service levels, quality levels, or something else. In other words, the large market has a variety of submarkets, or segments, that vary substantially. One of the crucial elements of marketing strategy is to choose the segment or segments that are to be served. This, however, is not always easy because different methods for dissecting a market may be employed and deciding which method to use may pose a problem. Virtually all strategists segment their markets. Typically, they use SIC codes, annual purchase volume, age, and income as differentiating variables.

Categories based on these variables, however, may not suffice as far as the development of strategy is concerned. RCA, for example, initially classified potential customers for color television sets according to age, income, and social class. The company soon realized that these segments were not crucial for continued growth because potential buyers were not confined to those groups. Later analysis discovered that there were “innovators” and “followers” in each of the above groups. This finding led the company to tailor its marketing strategy to various segments according to their “innovativeness.” Mass acceptance of color television might have been delayed substantially if RCA had followed a more traditional approach. An American food processor achieved rapid success in the French market after discovering that “modern” Frenchwomen liked processed foods while “traditional” French housewives looked upon them as a threat. A leading industrial manufacturer discovered that its critical variable was the amount of annual usage per item, not per order or per any other conventional variable. This proved to be critical since heavy users can be expected to be more sensitive to price and may be more aware of and responsive to promotional perspectives. Segmentation aims at increasing the scope of business by closely aligning a product or brand with an identifiable customer group. Take, for example, cigarettes. Thirty years ago, most cigarette smokers chose from among three brands: Camel, Chesterfield, and Lucky Strike. Today more than 160 brands adorn retail shelves. In order to sell more cigarettes, tobacco companies have been dividing the smoking public into relatively tiny sociological groups and then aiming one or more brands at each group. Vantage and Merit, for example, are aimed at young women; Camel and Winston are aimed mostly at rural smokers.

Cigarette marketing success hinges on how effectively a company can design a brand to appeal to a particular type of smoker and then on how well it can reach that smoker with sharply focused packaging, product design, and advertising. What is true of cigarettes applies to many, many products; it applies even to services. Banks, for example, have been vying with one another for important customers by offering innovative services that set each bank apart from its competition. These illustrations underscore not only the significance of segmenting the market but also the importance of carefully choosing segmentation criteria.

Segmentation Criteria

Segmentation criteria vary depending on the nature of the market. In consumergoods marketing, one may use simple demographic and socioeconomic variables, personality and lifestyle variables, or situation-specific events (such as use intensity, brand loyalty, and attitudes) as the bases of segmentation. In industrial marketing, segmentation is achieved by forming end use segments, product segments, geographic segments, common buying factor segments, and customer size segments. For a detailed account, however, reference may be made to a textbook on marketing management. In addition to these criteria, creative analysts may well identify others. For example, a shipbuilding company dissects its tanker market into large, medium, and small markets; similarly, its cargo ship market is classified into high-, medium-, and low-grade markets. A forklift manufacturer divides its market on the basis of product performance requirements. Many consumer-goods companies, General Foods, Procter & Gamble, and Coca-Cola among them, base their segments on lifestyle analysis. Data for forming customer segments may be analyzed with the use of simple statistical techniques (e.g., averages) or multivariate methods. Conceptually, the following procedure may be adopted to choose a criterion for segmentation:

1. Identify potential customers and the nature of their needs.

2. Segment all customers into groups having

a. Common requirements.

b. The same value system with respect to the importance of these requirements.

3. Determine the theoretically most efficient means of serving each market segment, making sure that the distribution system selected differentiates each segment with respect to cost and price.

4. Adjust this ideal system to the constraints of the real world: existing commitments, legal restrictions, practicality, and so forth.

A market can also be segmented by level of customer service, stage of production, price/performance characteristics, credit arrangements with customers, location of plants, characteristics of manufacturing equipment, channels of distribution, and financial policies. The key is to choose a variable or variables that so divide the market that customers in a segment respond similarly to some aspect of the marketer’s strategy. The variable should be measurable; that is, it should represent an objective value, such as income, rate of consumption, or frequency of buying, not simply a qualitative viewpoint, such as the degree of customer happiness. Also, the variable should create segments that may be accessible through promotion. Even if it is feasible to measure happiness, segments based on the happiness variable cannot be reached by a specific promotional medium. Finally, segments should be substantial in size; that is, they should be sufficiently large to warrant a separate marketing effort. Once segments have been formed, the next strategic issue is deciding which segment should be selected. The selected segment should comply with the following conditions:

1. It should be one in which the maximum differential in competitive strategy can be developed.

2. It must be capable of being isolated so that competitive advantage can be preserved.

3. It must be valid even though imitated.

The success of Volkswagen in the United States in 1960 can be attributed to its fit into a market segment that had two unique characteristics. First, the segment served by VW could not be adequately served by a modification to conventional U.S. cars. Second, U.S. manufacturers’ economies of scale could not be brought to bear to the disadvantage of VW. In contrast, American Motors was equally successful in identifying a special segment to serve with its compact car, the Rambler. The critical difference was that American Motors could not protect that segment from the superior scale of manufacturing volume of the other three U.S. automobile producers. The choice of strategically critical segments is not straightforward. It requires careful evaluation of business strengths as compared with the competition. It also requires analytical marketing research to uncover market segments in which these competitive strengths can be significant. Rarely do market segments conveniently coincide with such obvious categories as religion, age, profession, or family income; or, in the industrial sector, with the size of company. For this reason, market segmentation is emphatically not a job for statisticians.

Rather, it is a task that can be mastered only by the creative strategist. For example, an industrial company found that the key to segmenting customers is by the phase of the purchase decision process that they experienced. Accordingly, three segments were identified: (a) first-time prospects, (b) novices, and (c) sophisticates. These three segments valued different benefits, bought from different channels, and carried varying impressions of providers. A technology-consulting firm, Forrester Research Inc., separates people into ten categories: “fast forwards, techno-strivers, hand-shakers, new age nurturers, digital hopefuls, traditionalists, mouse potatoes, gadget-grabbers, media junkies, and sidelined citizens.” For example, “Fast forwards” own on an average 20 technology products per household. Several of their clients have found this kind of classification useful in identifying segments to serve. Market segmentation has recently undergone several changes. These include: 16

• Increased emphasis on segmentation criteria that represent “softer” data such as attitudes and needs. This is the case in both consumer and business-to-business marketing.

• Increased awareness that the bases of segmentation depend on its purpose. For example, the same bank customers could be segmented by account ownership profiles, attitudes towards risk-taking, and socioeconomic variables. Each segmentation could be useful for a different purpose, such as product cross-selling, preparation of advertising messages, and media selection.

• A move towards “letting the data speak for themselves,” that is finding segments through the detection of patterns in survey or in-house data. So-called “data mining” methods have become much more versatile over the past decade.

• Greater usage of “hybrid” segmentation methods. For example, a beer producer might first segment consumers according to favorite brand. Then, within each brand group, consumers could be further segmented according to similarities in attitudes towards beer drinking, occasions where beer is consumed, and so on.

• A closer connection between segmentation methods and new product development. Computer choice models (using information about the attribute trade-offs that consumers make) can now find the best segments for a given product profile or the best product profile for a given market segment.

• The growing availability of computer models (based on conjoint data) to find optimal additions to product lines products that best balance the possibility of cannibalization of current products with competitive draw.

• Research on dynamic product/segment models that consider the possibility of competitive retaliation. Such models examine a company’s vulnerability to competitive reactions over the short term and choose product/segment combinations that are most resistant to competitive encroachment.

• The development of pattern-recognition and consumer-clustering methods that seek segments on the basis of data but also respect managerial constraints on minimal segment size and managerial weightings of selected clustering variables.

• The development of flexible segmentations that permit the manager to loosen a clustering based only on buyer needs (by shifting a small number of people between clusters); the aim might be to increase the predictability of some external criterion, such as household profitability to a company, say, selling mutual funds.