Machine learning-based classification of competitors performance: evidence from Chinese logistics companies

Abstract

Digital technology is drastically altering business patterns. Enterprises strive to use these technologies to discover the worth of their consumers while understanding their rivals is required to get a decisive advantage in competitive marketplaces. In this situation, existing competitive intelligence practices require more fundamental theories to underpin them and the awareness to apply these strategies. This study will propose theory-based indicators as criteria to evaluate competitors intelligently through machine learning techniques and depict the competitive landscape using the experts' suggestions. The results reveal that the XGBoost model provides the best performance in evaluating competitors and will be used for a competitive landscape depiction with experts' suggestions. The competitive landscape shows that companies with medium R&D and high R&D are mainly located in Eastern China. In all provinces, Beijing is the most competitive. Given the number and proportion of high-R&D logistics companies, the Beijing market is nearly saturated and presents a high market entry barrier for entrants.