Maintaining customer relationships is a key to business success in today’s competitive environment. But all markets contain many subgroups of customers that behave differently, have different hopes, fears and ambitions, and have different purchasing behaviors. So, each subgroup must be behaved differently in order to build these relationships. On the road to this goal, customer segmentation is the first step. The goal of a segmentation system is to identify groups in which the customers are as much alike as possible and greatly differentiated from customers in other segments. If the segmentation system is well designed, members of a segment have similar interests, attitudes and behaviors, and they will respond similarly to elements of the marketing mix such as pricing, promotion and sales channel. Properly developed, segmentation insights inform a strategic roadmap intended to take advantage of key profit driving opportunities within each unique customer group. This could be shortening customer purchase cycles, driving higher spend, building greater customer loyalty, deepening cross-product penetration or lowering service and support costs.Internet service providers are one of the most active companies in today’s business world and a key element of development. They provide various services and products for their customers who are growing so rapidly. The number of their customers is different in countries depending on the level of development of countries. But it can be said that in a close future, almost all of the people will be customers of Internet service providers. Furthermore, in today’s world, where the market is highly competitive, customers face with various providers with different marketing strategies. These companies can be successful in the competitive environment by customer segmentation and designing proper strategies in each segment.The goal of this project is to mine the customer data to perform customer segmentation and consequently defining proper and useful strategies to win in the competitive environment.In this study Recency, Frequency and Monetary method which also known as RFM method has been used for customer segmentation in an Iranian internet service provider. By definition of some new variables in RFM method, two new RFM variant methods have been proposed which have some advantages with respect to simple RFM model. The results of applying these new methods show their effectiveness for customer segmentation and also their ability in identification of customer behaviors especially the risk of cancelling company services.