Stochastic processes applications in e-commerce.

Abstract

Nowadays, people tend to rely on Internet to do their shopping. Even though shopping online is constantly rising, its channels convert only a small fraction of their visitors into buyers. This might be either because of technical factors (for example, website design or functionality) or because of the initial user’s intentions. By using stochastic processes a shop manager can acquire valuable data that can be used to rise the conversion percentage and the successfully completed sales. In this essay, we study the phenomenon of cart abandonment and we propose a hidden Markov model that detects user’s characteristics, heterogeneity and intentions on the fly. Then this information is proposed to be used by the store in order to adapt properly and reduce the probabilities of cart abandonment and by the store’s manager to gain a better understanding of the way consumers tend to use the website.