A Cointegration Model with Structure Breaks for Customer Migration Analysis

As new technologies, products, and services emerge, consumers actively or passively migrate from legacy products to new substitutes while enjoying the adventures and benefits of the new everyday technology. However, identifying customers who migrate presents a challenge for service providers in understanding customer needs and business trends. This paper proposes a cointegrated linear regression model with structure breaks for customer migration analysis. The purpose of the cointegration model is to extract important cause-effect relationships for the migrations. The unique characteristic of the proposed cointegration model encompasses structure breaks that represent significant changes of the relationship among multiple data streams of customer activities. The proposed method can help marketing researchers identify migration customers, better understand customer needs, and evaluate their business impacts so that corresponding marketing campaigns can be initiated. An example in the telecommunications industry is discussed to demonstrate the application of the proposed method for migration analysis.