Customer Modeling & Segmentation: A Practical Approach for Telecom

Abstract

Globally, the Telecommunication sector has grown up leaps and bounds, which has effect on several other business verticals as well, due to change in technology in the recent years like introduction of 5G technology, edge computing, cloud technology, IOT, Automation & AI etc. Earlier telecom sector was mostly for p-2-p communication like voice and text but these new technologies contributing havoc data disruption and inorganic growth of data every day. According to Forbes, in the end of 2019, we can expect 44 zettabytes data generated everyday which makes the analytics much more difficult to generate insight and take the further actions. When we talk about telecom services, we mostly talk about the data services to meet the daily need like banking transaction, insurance payment, reading newspaper, virtual classes, watching movies & other television programs, mobile money, virtual credit card etc. In today’s date, measuring customer experience in the light of new technologies, attribute of networks, device experience, personalized and contextual product & service offerings would be the key to sustenance.

 As competition is all time high, understanding the market need is the key to any service business, as well asthe market segmentation and customer modeling. The purpose of this study is to understand how the different parameters are influencing the customer structure and segmentation, internal and/or external dependencies on customer segmentation, why modelling is so imprtant while preparing the segementation, what analytics can be applied etc. Many modelling techniques like predictive or classical, available in the market e.g.; RFM (Recency Frequency Monetary Method), CRISP (Customer Response?based Iterative Segmentation Procedures) etc. and also the statistical technique like K-means clustering. In this study, a practical approach for telecom sector is shown and also the detailing of the parameters which are the influencing factor.