The high uncertainty of the pharmaceutical market environment justifies the relevance of using predictive analysis methods based on open data. The authors propose a methodological approach that provides for the construction of the patent landscape in conjunction with a categorical analysis of the pharmaceutical organizations product portfolio. The practical implementation of the proposed approach is illustrated by analyzing behavioral models of pharmaceutical organizations specializing in the development and market launch of antidiabetic drugs. The results allow us to draw reasonable conclusions about the models of competitive behavior of Russian and foreign organizations that have different rank of production facilities localization without using commercial data