Decades after services outdistanced manufacturing from an employment perspective, manufacturing has continued to dominate economic analyses, including innovation studies. As a reaction to this a new strand of service innovation studies has emerged within the last decade. These studies do not aim to compare innovation in services directly with innovation in manufacturing, rather they aim a studying distinctive features of service innovation. This has among other things resulted in the development of new, service-specific innovation concepts. However, as this paper seeks to demonstrate, these concepts imply a merging of actual innovation with activities such as learning and codification of knowledge. Whereas learning and codification of knowledge are closely related to innovation, the inclusion of activities that e.g. require or result in learning, but neither result in new products, processes, markets nor organisational structures, in the definition of innovation, moves these studies away from the Schumpeterian heritage of innovation studies. This further implies that the meaning of innovation as an economic concept becomes unclear. There is thus a need for a conceptual strengthening of the new service innovation studies in order for them to contribute to the development of a so-called ‘synthesis approach’ to innovation, which has a broad and conceptually solid – perspective on innovation, whether this is carried out in manufacturing, in services, or in a grey area embracing both.