Abstract
Bookis is a C2C online marketplace for physical books. It is a product-orientated platform on web and mobile apps that enables customers to buy used books from other customers on the platform. Insights into the customer behavior and how customers experience their customer journey is crucial to improve their marketplace. Service providers, such as Bookis, are accumulating more and more data from their systems that could be leveraged for analysis. Yet, turning such data into comprehensive individual customer journeys is a significant challenge. This difficulty is compounded by several issues among which: (1) converting or connecting raw events to discernible touchpoints, (2) defining the boundaries of journeys in a C2C marketplace, and (3) acquiring and integrating data from external service providers such as delivery services. We provide a case study describing two analysis iterations in an attempt to obtain a comprehensive view on Bookis’ main customer journey by combining an “outside in” view (mystery shopping) with an “inside out” view (process mining). We leverage a novel event knowledge graph approach to data integration and conclude that the combination of both viewpoints is important to obtain a comprehensive analysis of the customer journey. In particular, relevant parts of the journey may not be covered by data-driven method, despite the large amounts of event data available.