The recent increase in everyday life processes oriented to services has modified the classic conception of traditional economic, management, organizational, relations and computer systems, supporting the development of Service-oriented Systems. Following the assumptions of Service Science, a Service System is today conceived as a system composed of heterogeneous entities, including people, technologies and business activities, which interact through the exchanging of services. However, the nature of such an exchanging of services, which could be static, dynamic or adaptive, modifies the degree of interaction among these entities, thus assessing the capability of the entire service system as being smart. The aim of this work is first, outlining the most representative service logics and systems theories, to highlight the interpretative evolution of a service system, and next, focusing on the nature of the interaction process, to describe the conditions that could make service systems smarter. In order to make as evident as possible the originality of our proposal, we have illustrated these theoretical considerations by focusing on the real example of an e-commerce platform, Amazon, that, as we will see, can be considered as a service system