This chapter explores how the structure of networks, in terms of complexity, mediates the propensity to advertise ‘extreme services’. The central hypothesis is that network structure predicts the advertisement of extreme services. To test this hypothesis, associations between key variables are first explored through a bivariate analysis, followed by a series of multilevel, or random effects, models. The results from the analysis suggest that the structure of networks predicts the advertisement of specific services, and that more complex networks are significantly more likely to advertise ‘extreme services’. The variability between networks, in relation to the propensity to advertise different services, appears much more important than the variability within networks and between adverts. This, in turn, suggests that the assessment of vulnerability and exploitation in online escort adverts may be more feasible when conducted on a network rather than advert level.