Customer lifetime value (LTV) is one of the most important metrics used to analyze any business projects. The knowledge of LTV makes it possible to organize efficiently marketing campaigns, to forecast projects revenues and to simulate customers' lifecycle.
The most of the known methods of LTV calculation are based on the knowledge of the total revenue from a customer and their lifetime. However, many of the modern tasks require calculating LTV until the end of customer’s lifetime. In this case the only solution is to predict the LTV based on the available data.The aim of this report is to perform a comparative analysis of existing models of LTV calculation and to describe the modified predictive model used in the game project "The Heavens".