The issue of products missing from the shelf is of high importance in a retailing supply chain. A decision system enabling store managers to decrease the out-of-shelf rate would be valuable. This paper examines two different approaches in building such a decision system: (a) utilization of inventory control models and (b) incorporation of classification methods. Both share criticism regarding their applicability, thus a further exploration of this issue is valuable. This paper examines four different experiments in order to evaluate each approach and record its applicability in the grocery sector. It comes out that both research approaches have limitations and strengths, thus it is suggested that such a decision system should be designed by utilizing both of them.