Spatial clustering finds groups of neighbouring objects with similar attributes, revealing patterns of spatial interaction and influence. However, not all similarities in spatial data are due to areal effects. Confounders can mask similarities and hide the spatial signal in the data. We see this, for example, in cultural evolution where language similarities
due to shared ancestry mask similarities due to contact and interaction. In this article, we present sBayes a Bayesian mixture model for spatial clustering in the presence of confounders. sBayes learns which similarities in a set of spatial point objects are explained by confounding effects and assigns objects to clusters based on the remaining similarities in the data. We introduce the algorithm to a geographic audience on the example of a fictional mobility analysis. We discuss how sBayes can be applied to ecology, health, and economy problems, revealing hidden geographic structures and patterns.