Data envelopment analysis (DEA) is a non-parametric data-driven approach for evaluating the efficiency of a set of homogeneous decision-making units (DMUs) with multiple inputs and multiple outputs. The number of performance factors (inputs and outputs) plays a crucial role when applying DEA to real-world applications. In other words, if the number of performance factors is significantly greater than the number of DMUs, it is highly possible to arrive at a large portion of efficient DMUs, which practically may become problematic due to the lack of ample discrimination among DMUs. The current research aims to develop an array of selecting DEA models to narrow down the performance factors based upon a rule of thumb. To this end, we show that the input- and output-oriented selecting DEA models may select different factors and then present the integrated models to identify a set of common factors for both orientations. In addition to efficiency evaluation at the individual level, we study structural efficiency with a single production unit at the industry level. Finally, a case study on the EU countries is presented to give insight into business innovation, social economy and growth with regard to the efficiency of the EU countries and entire EU.