Customer loyalty prediction in multimedia Service Provider Company with K-Means segmentation and C4.5 algorithm

The development needs the Internet and cable television entertainment increase per year that affect popping up various multimedia service provider company which is offered a lot of services to win the market. This makes customer has a lot of company choices and makes customer to be more demanded and move easily from a provider to other provider, where company knows that keep customer has the cost that is lower than the cost to get new customer. So, it's important for company to know customer loyalty and company can also project the income as reference in company development planning. Company needs to has accurate model, so researcher uses k-means segmentation and C4.5 classification algorithm, which can be seen that the model has accuracy 79.33% and Area Under Curve (AUC) 0.831. This research contribution is the use of related data using customer potential segmentation based on Recency Frequency Monetary (RFM) model, so can increase accuracy percentage in customer loyalty classification research.