An Application of Fuzzy Clustering to Customer Portfolio Analysis in Automotive Industry

An Application of Fuzzy Clustering to Customer Portfolio Analysis in Automotive Industry

Abdulkadir Hiziroglu (Yildirim Beyazit University, Cankaya, Turkey) and Umit Dursun Senbas (Sakarya University, Adapazari, Turkey)
Copyright: © 2016 |Pages: 13
DOI: 10.4018/IJFSA.2016040102


Having achieved an optimized customer portfolio has been of significant importance for companies. The literature provides several portfolio models and vast majority of them are in matrix form where several descriptors are used as dimensions of the matrix. These dimensions are characterized in ambiguity and require specific methods to tackle with it. The aim of this paper is to utilize fuzzy clustering in customer portfolio analysis to reduce this uncertainty and to make a comparison with a traditional customer portfolio model. A dataset of 130 customers of an automotive supplier in Turkey is used to perform the analyses and the results are compared with a conventional customer portfolio matrix. By making use of substantiality and balance of portfolio parameters, a qualitative and quantitative assessment of categorization generated by both approaches are evaluated. The use of fuzzy clustering gives more substantial clusters and a more balanced customer portfolio compared to the traditional matrix form of portfolio. Marketing managers can understand their overall customer portfolio better and reduce the effect of descriptive indicators via benefiting the fuzzy clustering results.
Article Preview

1. Motivation And Background

Studies on customer portfolio analysis have been done since the relationship approach is an issue of discussion in the literature (Gok, 2009). Ittner and Larcker (1998) have suggested that companies increase their profitability by means of proactive customer relations enabled by an effective portfolio structure. Some researchers have presented that it is necessary to establish different customer and manager relations in changing situations (Terho & Halinen, 2007). Therefore, it is possible to apply different approaches in different situations. There have been many studies since 1980s regarding customer portfolio modelling. Fiocca (1982) has developed a portfolio model known as “Account Portfolio Analysis for Strategy Development” in which customer account and buyer/seller relationship are used as portfolio variables. Turnbull and Topcu (1997) have tested this model in a Minerals Company. Cunningham and Homse (1982) have also proposed a model that uses technical interaction and sales volume as the dimensions of the portfolio matrix. Customers in this portfolio model are classified into four categories; technical development customers, high involvement customers, low commitment customers, cash flow customers. This model is tested by Eng (1999) in service business. Campbell and Cunningham (1983)’s model focus on industrial markets and life-cycle related attributes, indicators pertaining to customer/competitor analysis are the main variables of this model. Shapiro et al. (1987) have developed a portfolio model based on two dimensions; net price and cost to serve in which the customers are classified as passive, aggressive, carriage trade and bargain basement customers. Turnbull and Zolkiewski (1997) have empirically tested the last two models in an IT company. The theoretical base of this study is adapted from Shapiro et al. (1987) and the applied portfolio model utilizes three dimensions, ‘Ordering Frequency’, ‘Turnover’ and ‘Customer Relation Cost’. This study uses turnover and relationship cost variables in place for net price and cost to serve dimensions of the Shapiro’s portfolio model and it also extends the model via adding another variable (ordering frequency) as an additional portfolio dimension. In other words, Shapiro et al. (1987)’s two-dimensional portfolio model was extended to a third-dimension portfolio model as it has been adapted by Hiziroglu et al. (2013). The aforementioned study also provides a comparison for different portfolio models together with the existing empirical work, therefore more interested readers can refer to the corresponding work for more information.

Complete Article List

Search this Journal:
Open Access Articles: Forthcoming
Volume 9: 4 Issues (2020): Forthcoming, Available for Pre-Order
Volume 8: 4 Issues (2019): 3 Released, 1 Forthcoming
Volume 7: 4 Issues (2018)
Volume 6: 4 Issues (2017)
Volume 5: 4 Issues (2016)
Volume 4: 4 Issues (2015)
Volume 3: 4 Issues (2013)
Volume 2: 4 Issues (2012)
Volume 1: 4 Issues (2011)
View Complete Journal Contents Listing