In this paper, the effect of matrix properties on electrical properties of the composite material is investigated. The limited diffusion aggregation model is utilized to study the agglomeration-deagglomeration processes of the filler with consideration of a stochastic component of material properties. The model input parameters are the volumetric concentration of free particles, the probability of free particle sticking to an aggregate, the velocity of free particles. The material viscosity is selected as a parameter of a polymer matrix that affects the level of agglomeration of the filler and manufacturing process. The impact of viscosity on a distribution of disperse filler aggregates over the polymer matrix is simulated with the following model parameters: the probability of free particle sticking to an aggregate, the velocity of free particles. The modeling results reveal that increasing of mixing velocity (decreasing of viscosity) causes only slight changes in the composite material conductivity. The modeling results and dependencies are validated by experimental studies of carbon-filled rubbers. It is confirmed that impact of viscosity matrix on composite material conductivity reduces the effect of highly structured carbon. Therefore, taking into consideration the complexity of filling viscous rubbers with highly structured carbon, the proposed technique of conductivity control can be considered ineffective. For materials with high viscosity, concentration changes affect properties of materials way more than the agglomeration-deagglomeration processes affect properties of materials with constant concentration.