Fuzzy Target Groups in Analytic Customer Relationship Management

Fuzzy Target Groups in Analytic Customer Relationship Management

Michael Kaufmann, Cédric Graf
DOI: 10.4018/978-1-4666-0095-9.ch008
(Individual Chapters)
No Current Special Offers


Scoring models yield continuous predictions instead of sharp classifications. Scoring customers for profitability, loyalty, or product affinity corresponds to an inductive fuzzy classification: The model represents a continuous membership function mapping the set of customers into the fuzzy set of interesting customers – the fuzzy target group. This chapter presents a method for membership function induction based on normalized likelihood ratios. Applications of this method are proposed for selection, visualization, and prediction in the field of analytics in general, and for customer profiling, target group definition and customer scoring specifically for analytic customer relationship management. A real world case study is described. Furthermore, an implementation of the proposed method, developed at the research center for fuzzy marketing methods (FMsquare1), is presented.
Chapter Preview


Management Summary

In this Chapter, the application of analytics (quantitative decision support) to customer relationship management (CRM) is discussed. It is shown how three analytic techniques, selection (finding relevant attributes), visualization (plotting relevant associations) and prediction (estimating relevant class membership) help increase success of CRM-based marketing activities. Furthermore, it is shown how inductive fuzzy logic techniques provide a technical means to support these three types of analyses, and a prototype implementation is presented.

The benefit of analytic CRM is its focusing on relevant information in order to increase efficiency. Selection, visualization and prediction can help optimize efficiency of CRM and Marketing resource investment by quantitative decision support. The fuzzy logic method presented in this chapter, inductive fuzzy classification (IFC), is a research approach which can be applied to support decisions in CRM-based marketing either in a human oriented or automated approach. Its benefit is (1) a selection and relevance filtering algorithm which works for both numeric and symbolic data; (2) a standardized semantically intuitive visualization technique for human decision makers; and (3) a method to increase predictive accuracy of existing models by transforming enterprise information into the fuzzy domain.

The basic idea is to compute fuzzy membership degrees to desirable targets from existing data (likelihood-based inductive approach). The resulting models, called membership functions, can be used in targeted database marketing in order to identify relevant customer attributes regarding a target such as product affinity; to plot the corresponding data-based target associations of relevant customer attributes in two dimensions; and to improve campaign response rates and maximizing efficacy of predictive models by transforming customer attributes into membership degrees to target likelihood. Specific visual examples can be found in Figures 2, 3, and 5. The section Case Study shows a real world instance.

Figure 2.

Example of attribute ranking for the German credit data

Figure 3.

Visualization of a relevant attribute and its association with the target as an inductive membership function

Figure 5.

Schema of a fuzzy customer profile based on IFC-NLR, together with two instances presented by Kaufmann (2009)


Complete Chapter List

Search this Book: