Visual representation of information is ubiquitous, from traditional management reports via dynamic management cockpits to modern 3D and 4D visualizations of 'big data'. Little is known about how the perception in terms of efficiency of such a visual representation can be predicted ex-ante. The visualization/task complexity nexus is widely accepted to be a solid foundation of such a predictive model, although evidence points to a variety of additional factors, including data complexity, visual complexity, spatial ability, and experience. While most of these factors can be readily operationalized, the construct of experience is vastly under-researched in this context. In this paper we test the influence of the aforementioned factors on the perceptive efficiency of visualization - given a certain task complexity - by using experiments and structural equation modelling. In addition we look at four different conceptualizations of experience. The adjusted R2 of the model is 39% and all four proxies of experience are significant at the p.