From the late 1990s, the establishment of a new relational 'turn' in the study of world city connectedness in globalisation has run parallel to the wider relational turn occurring in economic geography. Early work, built firmly upon a qualitative approach to the collection and analyses of new intercity datasets, considered cities as being constituted by their relations with other cities. Subsequent research, however, would take a strong quantitative turn, best demonstrated through the articulation of the inter-locking world city network (ILWCN) 'model' for measuring relations between cities. In this paper, we develop a critique of research based around the ILWCN model, arguing that this 'top down' quantitative approach has now reached a theoretical impasse. To address this impasse, we argue for a move away from structural approaches in which the firm is the main unit of analysis, towards qualitative approaches in which individual agency and practice are afforded greater importance.