The most common way to model space syntax is to use an even distribution of lines or isovists. In positioning analysis, an uneven distribution of isovists is used in order to integrate social structures into the spatial analysis. These models can be used to analyse spaces highly influenced by social or cultural organisations. In offices, organisational structures can produce a constant bias in patterns of, for example, face-to-face interaction. Furthermore, labelling space according to organisational structure has sometimes been found to be as powerful a barrier as physical walls. In this study, these ‘constant biases’ are treated as being more important as origins rather than as mere biases. Positioning analysis is based on examination of spatial configuration of selected positions that enable these biases to be turned into parts of the analytic model. The selected positions should represent the organised entity of interest and form a subset within the general integration model. This set of points can be analysed configurationally through their spatial relationships, which can then be used together with traditional space syntax models. Thus the subset is described both from internal and external relationships, a fruitful approach in many cases. Applying this methodology when analysing face-to-face interaction in offices has provided strong correlations, which have produced two interesting results: face-to-face interaction correlates both with subsets of spatial distribution and with the integration patterns within the subset. At the same time, our analyses of how department stores and libraries spatially organise commodities and categories provide further support for how an analysis of spatial positioning answers key spatial strategy questions. Brands, types, and other categories position themselves in relation to one another as well as to a general integration and control patterns using modulations of a small set of spatial relations. Basing the configurative models on spatial relations of categories rather than on general spatial relations of an evenly distributed grid provides remarkably clear information. These findings are used to argue for working with positioning analysis parallel to traditional space syntax analysis, especially regarding examining spaces that are highly influenced by social or cultural organisation, such as offices, department stores, libraries, and museums.