We provide an O(n polylog n) bound on the expected complexity of the randomly weighted multiplicative Voronoi diagram of a set of n sites in the plane, where the sites can be either points, interior-disjoint convex sets, or other more general objects. Here the randomness is on the weight of the sites, not their location. This compares favorably with the worst case complexity of these diagrams, which is quadratic. As a consequence we get an alternative proof to that of Agarwal et al.[AHKS14] of the near linear complexity of the union of randomly expanded disjoint segments or convex sets (with an improved bound on the latter). The technique we develop is elegant and should be applicable to other problems.Generalizations of Voronoi diagrams. In the additive weighted Voronoi diagram, the distance to a Voronoi site is the regular Euclidean distance plus some constant (which depends on the site). Additive Voronoi diagrams have linear descriptive complexity in the plane, as their cells are star shaped (and thus simply connected), as can be easily verified. This holds even if the sites are arbitrary convex sets. In the multiplicative weighted Voronoi diagram, for each site one multiplies the Euclidean distance by a constant (again, that depends on the site). However, unlike the additive case, the worst case complexity for multiplicative weighted Voronoi diagrams is Θ(n 2 ) [AE84], even in the plane. In the weighted case, the cells are not necessarily connected, and a bisector of two sites is either a line or an (Apollonius) circle.In the Power diagram, each site s i has an associated radius r i , and the distance of a point p to this site is s i − p 2 − r 2 i ; that is, the squared length of the tangent from p to the disk of radius r i centered at s i . As such, Power diagrams allow including weight in the distance function, while still having bisectors that are straight lines and having linear combinatorial complexity overall.Klein [Kle88] introduced (and this was further refined by Klein et al. [KLN09]) the notion of abstract Voronoi diagrams to help unify the ever growing list of variants of Voronoi diagrams which have been considered. Specifically, a simple set of axioms was identified, focusing on the bisectors and the regions they define, which classifies a large class of Voronoi diagrams with linear complexity (hence such axioms are not intended to model, for example, multiplicative diagrams).Randomization and Expected Complexity. In many cases, there is a big discrepancy between the worst case analysis of a structure (or an algorithm) and its average case behavior. This suggests that in practice, the worst case is seldom encountered. For example, recently, Agarwal et al. [AKS13, AHKS14], showed that the expected union complexity of a set of randomly expanded disjoint segments is O(n log n), while in the worst case the union complexity can be quadratic. In other words, Agarwal et al. bounded the expected complexity of a level set of the randomly weighted Voronoi diagram of disjoint segments.If the sites are placed...