Project topics and materials
Get Complete Project Material File(s) Now! »
Segment identification method selection
This aspect deals with the actual formation of market segments. Guidelines for formation indicate that a segment must consist of “an economical minimum number of customers [buyers] to offer an economical value or volume of sales” (Foedermayr & Diamantopoulus, 2008:253). In addition, the number of segments formed must be such that it is manageable. The formation of too many segments could lead to difficulties in terms of inter-segment heterogeneity. In other words, the number of segments is too many and almost too granular. On the other hand, too few segments could lead to a lack of intrasegment homogeneity; in other words, the diversity between the members in a group being too great. McDonald and Dunbar (2004:57) advise that each segment should be subjected to a “reality check” based on the size of each segment, the differentiation between the offers they require, the business’s ability to identify and reach the different buyers found in each segment, and the compatibility of these segments with the business. However, this process takes place without any consideration of a segment’s attractiveness (Foedermayr & Diamantopoulus, 2008:253). The employment of statistical tools to aid in segment identification, and for the determination of membership in market segments ranges from the construction of basic cross-tabulation to advanced multivariate statistical techniques. The following sections describe some of the multivariate statistical techniques that could assist in segment identification. The researcher, however, realises that new statistical methods and tools emerge and evolve constantly and describing all of these would go beyond the scope of this study. The intention is merely to note, through a brief discussion, the value of the contributions that these types of techniques have made in assisting in market segmentation. The application and extent of techniques in practice go far beyond these descriptions. Some of these techniques directly result in the formation of groups or segments, while others only provide insights into the existence of underlying relations that could be used by researchers and marketing practitioners in market segmentation studies. This suggests the combining of multivariate statistics techniques, as part of segmentation analysis. However, as noted, the use of statistical techniques requires various decisions about the selection of a clustering algorithm, and the determination of the number of segments, as well as the responsibility for these choices. This responsibility would lie with the researcher, the statistician and the marketing practitioner.
Cluster analysis
Wiid and Digginess (2009:250) describe cluster analysis as a multivariate technique that is used to group similar objects. Everitt, Landau and Leese (in Abeyasekera, 2005:370) describe it as identifying natural groupings among sampling units, for example, respondents, households or businesses, so that units within each group (cluster) are similar to one another, while being dissimilar from any other units, which are to be found in different groups. Alexander et al. (2005:113), for example, employed cluster analysis to identify five distinct buyer segments for expendable input purchases for crop and livestock commercial producers in the United States. The goal of cluster analysis is, therefore, to explore patterns in complex population and to identify homogeneous groups of clusters (Alexander et al., 2005:113; Franke et al., 2009:273). Key considerations, therefore, include the selection of variables that could serve as a basis for cluster formation, the number of variables, the measurement level of data, as well as the criteria for combining cases into clusters. Dolnicar and Lazarevski (2009:359) noted specifically the challenge associated with having too many variables in the segmentation base, given the sample size. A common approach that researchers have used to address this challenge is to first subject the variables to an exploratory factor analysis or principal component analysis as a data reduction technique, before clustering the resulting factor scores. A potential disadvantage of this approach is when the factor analytical solution explains a very low percentage of the variance in the raw data. Consequently, a large proportion of the information contained in the data is essentially discarded.
Optimally, researchers should measure only a small number of conceptually well-developed items from the start.
CHAPTER 1: INTRODUCTION
1.1 BACKGROUND
1.2 PROBLEM STATEMENT
1.3 PURPOSE STATEMENT
1.4 RESEARCH OBJECTIVES
1.5 IMPORTANCE AND BENEFITS OF THE STUDY
1.6 RESEARCH RESOURCES AND METHODS
1.7 STUDY DELIMITATIONS AND ASSUMPTIONS
1.8 DEFINITION OF KEY TERMS
1.9 STRUCTURE OF THE THESIS
CHAPTER 2: MARKET SEGMENTATION
2.1 INTRODUCTION
2.2 MARKET SEGMENTATION AS PART OF MARKETING MANAGEMENT
2.3 DEFINING MARKET SEGMENTATION
2.4 THE MARKET SEGMENTATION PROCESS
2.5 THE PURPOSE AND ANTECEDENTS OF MARKET SEGMENTATION
2.6 THE KEY SUCCESS FACTORS OF MARKET SEGMENTATION
2.7 WHAT A MARKET SEGMENTATION LOGIC OFFERS MARKETING THEORY
2.8 BASES USED IN MARKET SEGMENTATION STUDIES
2.9 CHAPTER SUMMARY
CHAPTER 3: INDEX CONSTRUCTION
3.1 INTRODUCTION
3.2 CHARACTERISTICS OF AN INDEX
3.3 STEPS IN INDEX CONSTRUCTION
3.4 SELECTED APPROACHES TO INDEX CONSTRUCTION
3.5 CHAPTER SUMMARY
CHAPTER 4: RESEARCH DESIGN AND METHODS
4.1 INTRODUCTION
4.2 DESCRIPTION OF INQUIRY STRATEGY AND BROAD RESEARCH DESIGN
4.3 SOURCE OF DATA
4.4 SAMPLING
4.5 DATA ANALYSIS
4.6 ASSESSING AND DEMONSTRATING THE QUALITY AND RIGOUR OF THE RESEARCH DESIGN
4.7 ETHICAL CONSIDERATIONS
4.8 CHAPTER SUMMARY
CHAPTER 5: RESEARCH RESULTS
5.1 INTRODUCTION
5.2 INDEX CONSTRUCTION AND SEGMENTATION
5.3 CHAPTER SUMMARY
CHAPTER 6: CONCLUSIONS, IMPLICATIONS & IMPERATIVES FOR FUTURE RESEARCH
6.1 INTRODUCTION
6.2 CONCLUSION
6.3 IMPLICATIONS
6.4 LIMITATIONS
6.5 IMPERATIVES FOR FUTURE RESEARCH
6.6 CHAPTER SUMMARY
LIST OF REFERENCES
