Market Data Exploitation: Exemplified by the Battery Electric Vehicle Market

We propose a systematic data-driven approach for complexity reduction for metric data given as a table where the columns reference variables (features, attributes, characteristics, etc.) and the rows to samples (cases, products, machines, etc.). The approach introduces new variables, called principal components, which are a linear superposition of the original variables and capture most of the information to differentiate the several samples. The number of principal components is less than that of the original variables (complexity reduction). The analytical method is based on the Principal Component Analysis (PCA). Utilizing spectral theory, the PCA finds orthogonal directions in the original variable space where the highest variance occurs. These directions form the principal components. The proposed method is applied to the battery electric vehicle (BEV) market. Three principal components, interpretable as smallness, efficiency, and fun per euro, are identified, which cover 98 % of all the information on the market to differentiate offered electric vehicles. Furthermore, the positioning according to the original equipment manufacturers' smallness, efficiency, and fun per euro is illustrated. The developed method can be applied to various problems, such as finding the most relevant aspects, gaining insights into complex situations, or identifying the most important drivers for system optimization.