The fresh food supply chain management has experienced great changes over the last years, and it has now become the major strategic issue for food firms (Bourlakis and Weightman 2004). In this context, one of the main goals for a company is to individuate a combination of purchasing, transportation, physical distribution and logistics to get a position to achieve economies of scale; the control of the supply chain tends to shift from producers to retailers, while the increased pressure for higher quality and cost efficiency affects all members of the food chain. Consequently, the food chain is called to move towards a more vertically integrated structure that includes joint partnerships, strategic alliances and more vertical coordination among different supply chain players; and retailers are asking for new value-added logistics services. In fact, traditional logistics firms cannot always fully meet the demands of the retailers because they did not possess the adequate logistics and information technology solutions. Thus, retailers started to cooperate also with logistics information technology suppliers as new actors in the field. The resulting value-added services increase the operational complexity of the food retail chain. Today’s food supply chain is required to be reliable and agile enough to respond to consumer demand and preferences and it is expected that e-logistics practices and e-commerce will have a growing relevance to support a customer-focused and more responsive organisation. Sales forecasts and order planning play a central role in this customer- and data-centred process. They constitute a process that is dual to the physical process of delivering goods from the producer to the retailer. The whole process can be viewed as an integrated demand and supply chain management (Dotoli et al. 2005) which integrates both the pull action from the customer’s demand and the push effect from the producer and the retailer in the form of promotions and advertising actions. The forecasting techniques to be employed on the store level are called to fulfil the requirements of the so-called microforecasting (i.e. a specific forecast for every item in each store), while aggregation across items and/or across stores is considered not meaningful or even misleading. The order plan per store must be based on the individual sales forecast for the individual store. Since calendar events, advertising and promotional events may have very different effects in different stores depending on the respective locations, the forecasting techniques should not be based only on the observed time series of sales but should also be able to take this information about external factors into account also including the possibility of judgmental adjustments (Wang and Petropoulos 2016).