Predicting Subscriber Usage: Analyzing Multidimensional Time-Series Using Convolutional Neural Networks.

Companies operating under the subscription model typically invest significant resources attempting to predict customers' future usage. These predictions can be used to fuel growth: Companies can use them to target individual customers - for example to convert non-paying consumers to begin paying for enhanced services - or to identify customers not maximizing their subscription product. This can allow the company to avoid an increase in the churn rate, and to increase the usage of some customers. In this work, we develop a deep learning model to predict the product usage of a given consumer, based on historical usage. We adapt a Convolutional Neural Network with auxiliary input to time-series data, and demonstrate that this enhanced model effectively predicts future change in usage.