Abstract: Customer Lifetime Value (CLV) is a useful metric that plays a vital role in setting marketing plans because it can be used to answer several questions of business interest such as those related with the number of purchases and the monetary value that the customer can generate to the company in a future time range. There are two main related problems, namely, CLV prediction and customer churn classification of which each has its practical implications although the current study focuses on the former. The relevant literature highlights two broad approaches towards CLV prediction given by the probabilistic approach which represents the traditional approach for CLV prediction, and machine learning approach being the recent one. There are several methods to be implemented within each prediction approach, but no overall conclusion can be drawn concerning their performance.