Advancements in technology have contributed to the new business culture, where the Customer Relationship Management (CRM) is in the centre of a business concern. CRM is a widely implemented strategy for managing and fostering long term, profitable relationships with specific customers (Ling and Yen, 2001). The automated data mining tools made it possible to move beyond the analyses of the past events and data mining tools can be used to address problems that were seen as too time-consuming in the past, providing new opportunities for businesses within relationship management. This report identified that CHAID and neural networks are two of the most commonly used data mining techniques within CRM domain. However, each of these techniques has advantages and drawbacks, which should be taken into consideration when deciding on its appropriateness.
Nowadays, organisations are more concerned with increasing customer value, and realise that customers are more predictable than ever thought before. As argued by Written and Frank (1999), consumers engage, negotiate and purchase according to certain patterns engraved into transactional and behavioural records. This report will discuss data mining techniques for CRM. Real life case studies will be analysed and two types of data mining techniques will be discussed, focusing on their appropriateness to CRM.
2. Customer Relationship Management
According to Swift (2001) and Ngai (2005), CRM consists of four dimensions, which can be seen as a closed cycle of customer management system. They share a common goal of creating a deeper understanding of customer behaviours to maximise its value to the business in the long term. Data mining techniques can be used to extract hidden customer characteristics and behaviours from databases. Figure 1 represents a graphical classification framework of data mining techniques in CRM. As argued by Carrer and Povel (2003), the generative aspect of data mining consists of the building of a model by performing one of the following types of data modelling:
• Sequence Discovery
Figure 1: Classification Framework for Data Mining Techniques in CRM (Swift, 2001).
In order to begin the managing and fostering relationships, organisation needs to possess information about customers. The main source of customer data is from internal sources (e.g., billing records, customer surveys, web logs). An enterprise data is a crucial component of a CRM strategy in any organisation that wishes to apply data mining techniques. Most companies have very big databases that contain Human Resources (HR), marketing and financial resources; however, CRM requirements can be limited to a marketing data mart with partial supplies from other corporate systems (Freeman, 1999). Alternatively the information can be acquired from external sources, which might be a key...