Classifying decision-making units into efficient and inefficient classes is a common procedure in nonparametric data envelopment analysis approach. The inefficient units can be projected onto the production frontier by decreasing their inputs or increasing their outputs. However, if the closer projection point is clarified, the units will achieve their best situation. Previous methods have only focused on one objective, and other features have been ignored. This paper presents an alternative definition for the best projection by considering three main aspects: cost, revenue and closest projection points. The proposed procedure provides the closest possible distance, the lowest cost and the highest revenue, simultaneously. The goal programming has been used in this model. The results of implementing this approach on China's textile industry have shown the model applicability.