Optimizing store-based picking in the e- grocery: a model to assess costs and benefits

Despite e-grocery is achieving significant volumes in Europe, it is still underdeveloped in Italy, where it accounts for a low percentage on the total sales of the sector. However, its growth is the most important one among those performed online by the other industries and the number of e-grocery initiatives has been proliferating in the last three years. The main e-commerce challenges, most of all speed and quality of the service, are amplified by the peculiarities of the grocery sector. On the one hand, products are characterized by high levels of perishability and fragility; on the other hand, differently from the online sales in other sectors, demand is characterized by a significant number of lines per order and a low number of items per order line. Therefore, the order fulfilment process becomes more challenging. Moreover, if compared to offline purchases, the online channel requires new activities, i.e. picking and delivery, thus introducing additional costs for the grocers. Currently, e-grocers implement either the warehouse-based or the store-based logistic model. The warehouse-based solution, where online orders are fulfilled in a warehouse dedicated to e-commerce, requires significant investments. The stored-based solution, where items are picked from the shelves in traditional stores, implies high picking time (and thus high picking costs), and a reduction in the item fill rate. This study is aimed at proposing an intermediate logistic solution based on an access index based allocation: A-class items are picked from a dedicated back area, while B and C ones are picked from the shelves in the stores. The results of the model application show a reduction of about 38% in the picking time per order if compared to the traditional store-based solution.