A Practical Framework for Real Life Webshop Sales Promotion Targeting

Abstract

In recent years, online marketing has become increasingly extensive and effective. Product recommender systems are often deployed by e-commerce websites to improve user experience and increase sales. To address this, more and more e-commerce started to use machine learning models to predict customers’ purchase behaviors. In the scientific literature there are only few real-life studies to date which give solutions for recommendation systems for online advertising. The demand from the owners of such websites is given, however, it is hard for them to choose a method or model to predict from an endless number of options for some specific circumstances. The aim of this paper is to propose a practical guideline as a hybrid approach that predicts customers’ purchase behaviors and helps to target advertisement, sales form in user level. To this end, we have designed a robust hybrid model to predict interested sales form based on user behavior within a large e-commerce website. The paper details a real-life practical solution and build a structure that can be used in a large variety of e-commerce systems.