Abstract

Employing virtual online sales assistants is one way to provide smart services to online customers during their online shopping experience, while visiting an online store. Virtual online sales assistants can resolve the absence of social experienced by visitors to the majority of online stores. Each customer#x2019;s online shopping behaviour is unique; therefore, the motivations that attract customers to shop electronically are also likely to be different. Employing a virtual online sales assistant can fulfil different needs for the online customers visiting the store. Many developed virtual online sales assistants have disappeared due to not meet the expectations of customers. Therefore, it would be helpful to develop artificial actors to test such an assistant to determine if they can serve online customers satisfactorily. This paper details a novel way to test such an assistant by using Agentbased modelling and simulation (ABMS), which is a powerful simulation tool to simulate autonomous and interactive entities. For our research, we developed four virtual online customer actors with different personalities; e.g. enthusiastic, apathetic, basic, and impatient virtual online customers. These actors then interacted with a representative virtual online sales assistant. The result exhibited the different behaviour patterns that different virtual online customer types might have.