Data Driven Web Experimentation on Design and Personalization

Abstract

In today's world for we use online medium for virtually every aspect of our lives. Companies run controlled web experiments to make data driven decisions, to provide an intuitive online experience. We see a big correlation between online customer behaviors and designs and personal treatment, which could be used to create better customer engagement. In this paper we have studied the impact of design elements on chat invites*, by running experiments on a small population, using machine learning algorithms. Based on this we identify significant elements and build the most opportune personalized messages on invites. Statistical results show that, more visitors on the website accept chat invites which are personalized and optimized for the design. At [24]7, we have experimented extensively on user interface designs and journey based personalization which resulted in positive impact on our annual revenue.