Defining the main concepts is the first step in theory development in marketing (Bartels 1970, 6). However, studies only rarely explicitly define their concepts, although they measure them Jacoby (1978). Jacoby (1978) sees this as a quite serious deficit, since it is clear that concepts must precede their measurement. This paper argues that in spite of these very clear calls for rigorous theory-driven research, much has no changed since the 1970’s and 80’s. We take an example from a research stream on business-to-business relationship marketing and the basic concept of relationship and ask; has existing research started with the definition before it is measured? If not, why is it so difficult?