Pricing Perishables

A key feature of food products is their perishability. Within the short marketing window that characterizes most food and ag products, demand is typically highly stochastic and difficult to predict. This combination of features poses substantial challenges to retailers when pricing products and has implications for performance that ripples through vertical food chains. For many food products, processing to forms that can be preserved and held in inventory has traditionally been used as a means of coping with these conditions, despite its high costs and ancillary risks introduced such as change in product attributes and deterioration. This paper presents an alternative ERM strategy that focuses on dynamic pricing to control the rate of sale for perishable products. The paper considers a retailer that has market power to price and supplies perishable products to a market with substitute products and demand originating from heterogeneous consumers. Perishability implies a finite horizon for the marketing of the products over which demand across market segments of consumers is both dynamic and stochastic. Faced with uncertainty, we suppose the firm has limited information about the stochastic properties of demand and must choose a pricing strategy that projects over the market horizon. This price trajectory represents a key control mechanism to cope with uncertainty of both the perishability of the product and of demand. A variety of mechanisms for setting prices has been pursued in the past and can be imagined. Given substantial waste associated with food retailing, it seems evident that retailers may not incorporate the social or food supply network interests in optimal performance. Uniform pricing within the marketing horizon is typical for most food retailers. At the horizon, or shelf-life, the product is often removed from the shelves and either disposed of, or diverted into a secondary market. Despite such practice, lessons exist from other industries where dynamic pricing approaches have been pursued. Here, we consider pricing rules derived from robust optimization that sets price trajectories over the market horizon that explicitly consider two features that appear to be of particular interest for food: 1) sales of all available supply (i.e. eliminate disposal) and 2) existence of close substitutes (i.e. fresher product). An important innovation in many industries for dynamic is the concept of price assurance. We consider two types of price assurance. Under ex-post price assurance, prices are set subject to the constraint that refunds will be paid if future prices are reduced below levels paid by consumers. This is an intriguing variation on “everyday low pricing”. Next, we introduce a novel alternative that we label as ex-ante price assurance where the firm sets the dynamic price trajectory subject to the Robert. D. Weaver and Yongma Moon 213 constraint that prices will not decrease. Though perhaps counter-intuitive in perishable good context, we show this novel approach has merit under particular demand conditions. Thus, we compare three dynamic pricing strategies to manage uncertain demand given perishability: i) robust dynamic pricing, ii) ex-post price assurance, and iii) ex-ante price assurance. Numerical experiments show that our robust optimization model prevents loss when a firm encounters the worst case demand and outperforms a deterministic pricing model. Comparison across different pricing strategies identifies conditions under which particular strategies are superior to the others.