Major Project Report On Analysis of Customer Lifetime Value Modeling Techniques

Abstract

ABSTRACT

Customer lifetime Value has become an important parameter in order to find and reach out to the customers who tend to contribute more heavily and frequently. This parameter therefore depends on the marketing industry or even we can say they are interdependent. It is important to know about any customer’s purchase value and continuously monitor his transaction frequency and value for accurately determining CLV. One of the most important aspects or parameters that relates marketing and CLV is CAC (Customer Acquisition Cost) . It is definitely important to know how potentially a customer will contribute to the company and is it really worth spending money on an aspect to boost his purchases. In this project, we aim to study the different CLV Modelling techniques to predict the customer lifetime value, analyse them and find the best suitable model for our dataset. We have analysed in detail about both the Probabilistic models and Machine Learning models, Our study includes the detailed analysis of the two most common and widely used probabilistic model i.e. BG/NBD and in Machine Learning, we used DNN to develop the model and to predict the future value of both existing customers with more transaction history and also for the new customers with very few purchases.