There is a growing concern regarding the use of relatively coarse units for the aggregation of various spatial information. Researchers thus suggest that the street segment might be better suited than areal units for carrying out such a task. Furthermore, the street segment has recently become one of the most prominent spatial units, for example, to study street network centrality, retail density, and urban form. In this paper, we thus propose to use the street segment as unit of analysis for calculating the residential valorisation of urban space. To be more specific, we define a protocol that characterises street segments through a measure of central tendency and one of dispersion of prices. Moreover, through Bayesian clustering, it classifies street segments according to the most probable combination house type-valuation to provide a picture of local submarkets. We apply this methodology to the housing transactions exchanged in the French Riviera, in the period 2008–2017, and observe that outputs seem to align with local specificities of the housing market of that region. We suggest that the proposed protocol can be useful as an explorative tool to question and interpret the housing market, in any metropolitan region, at a fine level of spatial granularity.