The core problem that is identified in this paper is that no effective empirical operationalisation for the concept "Website" exists up to now. Consequently, progress in the research areas of webometrics is hindered. The authors point out that, although the notion of a Web-page can be operationalised in a straightforward way as the resource corresponding to a particular URL, the same is not true of "Website". There is no consensus as to what a "Website" is, although in practice the term is frequently used to suggest a coherent assemblage of Web-pages that is intended to be considered as a unit by the reader. Furthermore, there is a need to understand the effect that different pragmatic analytic choices may have on the results of such Web-graph analyses in order to aid interpretation of their results.