Stimulation of productivity increase is a key task at the present stage of development of the economies of both Russia and Eurasian countries. The purpose of this article is to identify quantitative assessments of how various factors impact productivity increase and conduct a cluster analysis of the regions, based on the considered indicators that evaluate the impact of relevant factors on productivity. The authors use general scientific methods such as analysis and synthesis, econometric analysis and multidimensional statistics. To build the model, the authors of the article used statistical data relating to socioeconomic development indicators for 85 Russian regions. As a result of the correlation and regression analysis, the following factors were identified: the average monthly wage, consumption of fixed capital, internal R&D costs, innovative activity of organisations, and tax burden. These factors have both positive and negative impacts on productivity. A cluster analysis was also conducted. It enabled to group the regions in terms of their productivity. Based on the analysis, the authors proposed the directions of improving the policy to increase productivity for each of the three clusters. For the regions included in the first cluster, it is necessary to apply methods of direct state regulation, for the regions of the second cluster-to pursue a policy of improvement of tax incentive mechanisms through the application