Final Remarks for the Case Study
This chapter presents the final conclusions of the case study analyzing the telecommunications industry. Although there are several distinct statistical and mathematical methods used to analyze data and, therefore, increase knowledge of customer behavior, most if not all of these methods miss relevant information about the relationships that make up the network.
The regular data analysis method isolates customers’ attributes, such as usage and demographic information. Neither of these describes how customers relate to each other or, most important, how relevant these relationships are to some business events.
The case study described in Part II was developed based on telecommunications data concerning two particular business events, churn and bundle diffusion.
The premise for the case study was that both events are impacted in a cascade time scale occurrence. Based on this hypothesis, a framework of analysis was built using several months of data in order to determine average customer behavior and analyze later events caused by the cascade concept.
If a particular customer decides to leave the company, the impact depends on his level of influence. If he is a highly influential customer, the likelihood that he will lead other customers to leave the company is very high. However, if he is not influential, he will probably not lead many customers to follow him. Bundle diffusion is quite similar to churn; influential customers are more likely ...