Cluster and Factor Analysis

A sporting goods manufacturer attempted to identify the market segments that existed for all types of sporting equipment. A large sample of sports active individuals was interviewed to determine attitudes toward sporting activities. They were asked to respond to dozens of sports related statements and to record their answers to each statement on a 10-point scale. Researchers at the sporting goods company had to decide how they could analyze such a large body of data in order to identify the number and types of market segments that existed for sporting equipment.

In the first example the sporting goods company is trying to identify the different market segments for sporting goods. Their intent is to identify one set of respondents with nearly identical attitudes and consider them one market segment. They hope to identify a second set of respondents, all of whom have attitudes similar to each other but different from those possessed by the first set of respondents. These respondents will be considered a second market segment. The company will also try to identify a third and a fourth market segment and still more, if they exist. Identifying market segments on the basis of similar attitudes is a commonly encountered situation in marketing, and cluster analysis is the most widely used method of analysis in such situations.

An automobile manufacturer wanted to know which automobile characteristics were considered to be very important by potential buyers of compact cars. Company researchers chose not to ask undisguised multiple choice questions about the importance of horsepower, interior roominess, luggage space, and so forth because they felt that consumers could not accurately report the relative importance of these5 items in their buying decisions. They also chose not to ask open questions about what characteristics were important, because they felt that respondents could not verbalize the characteristics of automobiles that were of importance to them. Instead, the researchers prepared 100 statements that related to all characteristics of automobiles that they believed could be important to consumers. Three hundred potential buyers of compact cars were asked to report on a seven point scale the extent to which they agreed or disagreed with each statement. Researchers at the automobile company then had to determine how they could analyze the responses to 100 scaled statements from each of 300 individuals in order to identify the major characteristics of compact cars that were important to potential consumers of such cars.

In the second example, the automobile company wants to identify the characteristics of compact cars that potential purchasers consider very important. But this is not easy because automobiles are complex products with many features and characteristics. The automobile company would like to identify only those features and characteristics that all or most potential purchasers agree are either important or very important. This type of situation is faced by many firms that market complex products or services such as homes, boats, large kitchen appliances, insurance policies, financial services, retirement communities, recreational vehicles, and other products. Factor analysis is one of the most widely used methods of analysis whenever accompany is trying to identify the most important characteristics of a complex product or service.

The two examples described above represent situations commonly encountered by marketing researchers. They are similar insofar as the researchers would have a great deal of data to be analyzed. They are also dissimilar, because a different method of analysis has to be used in each example if the data from that example are to be analyzed in the most appropriate manner.