The problem argued in the literature of direct mailing decisions generally contains three parts: 1) forecasting customers' future purchase/non-purchase responses; 2) evaluating the effectiveness of various strategies for increasing customers purchase responses; 3) prioritising the customers in terms of their values. A significant body of the literature has been dedicated to the first two components, and in particular, to purchase/non-purchase prediction modelling. However, in the current paper, we do not address these two components, but rather we focus on the third component. To this end, data envelopment analysis (DEA) technique and particularly cross-efficiency formulation of the best practice frontier Charnes, Cooper and Rhodes (CCR) (Charnes et al., 1978) (BPF-CCR) is used to determine those customers who should be put on the first priorities of marketing mailing list. In addition, the cross-efficiency formulation of worst practice frontier CCR (WPF-CCR) is developed to exclude the worst customers from mailing list and save the mailing expenses for the best practice ones. Using a numerical example, the application of the proposed model is demonstrated.