Demand forecast sharing in a dual-channel supply chain

Abstract

We assess the benefits of sharing demand forecast information in a manufacturer–retailer supply chain, consisting of a traditional retail channel and a direct channel. The demand is a linear function of price with a Gaussian primary demand (i.e., zero-price market potential). Both the manufacturer and the retailer set their price based on their forecast of the primary demand. In this setting, we investigate the value of sharing demand forecasts. We analyze the ‘make-to-order’ scenario, in which prices are set before and production takes place after the primary demand is known, and the ‘make-to-stock’ scenario, in which production takes place and prices are set before the primary demand is known. We also compare the supply chain performance with and without the direct channel under some assumptions (production cost is zero, and each demand function has the same slope of price). We find that the direct channel has a negative impact on the retailer’s performance, and, under some conditions, the manufacturer and the whole supply chain are better off. Our research extends and complements prior research that has investigated only the inventory and replenishment-related benefits of information sharing.

Introduction

Information technology has significantly reshaped supply chain behavior in the last two decades. Particularly, the commercial blossoming of the Internet has introduced tremendous opportunities and has underscored the importance of effective supply management. Companies such as Dell use direct sales, primarily through the Internet, to bring products to market faster than competitors, thus enjoying a huge early-to-market advantage. Moreover, direct sales allow the manufacturer to eliminate distributor and retailer margins, thus increasing its own profit margin. According to one survey reported in the New York Times, about 42% of top suppliers in a variety of industries have begun to sell directly to consumers over the Internet.

One significant benefit of information technology is to allow firms to share information (i.e., point-of-sale data, inventory, forecast data, and sales trends) quickly and inexpensively. Information sharing helps the supply chain in two fundamental ways. First, it enables the manufacturer to respond to consumer demand more quickly by appropriately scheduling production and replenishing retailer inventory. Innovations such as CRP and Vendor Managed Inventory (VMI) are efforts in this direction. Second, information sharing improves the accuracy of demand forecasts. As we know, forecasts are essential to the supply chain’s decision making and planning processes. Better forecasting can contribute to better price structuring and better inventory management. Recently, schemes such as Collaborative Planning, Forecasting, and Replenishment (CPFR) facilitate sharing of demand forecasts among the supply chain players in this direction. Conventional wisdom suggests that sharing of forecasts within a supply chain improves the forecast accuracy and leads to higher profitability.

One objective of this paper is not only to analyze the value of demand forecast sharing in a supply chain with direct channel but also to understand how to share information, e.g. under what conditions is information sharing mutually beneficial to both the manufacturer and the retailer and under what conditions is one player better off and the other is worse off with information sharing, as one player may use the shared information against the other.

So far, most research on information sharing has focused solely on inventory and replenishment related savings (i.e. Bourland et al., 1996, Gavirneni et al., 1999, Lee et al., 2000, Cachon and Fisher, 2000). In their models, price and demand are considered exogenous. By focusing on inventory and replenishment decisions alone, these models underestimate the benefit of information sharing, and also do not address the strategic role forecasts play in pricing. Our paper contributes to existing literature by addressing information sharing’s impacts on two components: pricing and inventory.

Another objective of this paper is to analyze the direct channel’s impact on supply chain performance. Intuitively, a direct channel enables the manufacturer to be the upstream product provider for the retailer and also enables the manufacturer to become the competitor of the retailer. Thus, questions may arise naturally, such as is the retailer the loser due to the resulting channel conflict (competition)? What about the manufacturer’s performance? We know that the direct channel simultaneously brings opportunity and threats to the manufacturer. Specifically, a direct channel is able to attract a new customer segment and increase the manufacturer’s profit margin, but on the other hand, a direct channel could lead to conflict with the existing retailer channel. Therefore, what is the overall effect of direct channel on the manufacturer’s performance? We will quantify these effects in this study.

Information sharing has been a major issue in supply chain literature. This literature discusses how a manufacturer can elicit information from retailers through inventory, lead time, and shortage allocation policies (Bourland et al., 1996, Gavirneni et al., 1999, Gallego et al., 2000, Cachon and Fisher, 2000). Bourland et al. (1996) show that information sharing offers significant benefits to the manufacturer and retailer when their ordering cycles are significantly out of phase. Gavirneni et al. (1999) study the holding and penalty cost of a finite capacity supplier facing demands from a single retailer following a (s, S) policy. By considering various types of demand distributions in their numerical experiments, Gavirneni et al. examine the conditions under which gaining information about the retailer’s inventory level is beneficial. Lee et al. (2000) study the benefit of demand information sharing when the underlying demand process faced by the retailer is an AR (1) process. They assume that the manufacturer and the retailer incur linear holding and backlogging costs, experience constant lead times, and follow base-stock policies. Chen et al. (2000) quantify the bullwhip effect for a simple two-stage supply chain and demonstrate that centralizing demand information can significantly reduce the increase in variability. Cachon and Fisher (2000) study the value of sharing data in a model with one supplier, N identical retailers, and stationary stochastic consumer demand. They conclude that implementing information technology to accelerate and smooth the physical flow of goods through a supply chain is significantly more valuable than using information technology to expand the flow of information. Gallego et al. (2000) show that delay base-stock policies can capture a significant portion of the benefits of demand information sharing. They compare the costs under fixed and random delay base-stock policies against the cost under demand information sharing. Aviv (2001) investigates the value of collaborative forecasting and integrating the retailer into the manufacturer’s replenishment process. Cachon and Lariviere (2001) study forecast sharing in a two-stage supply chain between a manufacturer and a supplier, where the manufacturer provides an initial forecast and a contract to the supplier. They find that it is always in the interest of the manufacturer to truthfully share the forecast with supplier, particularly when the demand forecast is high. Li (2002) considers a supply chain model with a manufacturer and multiple symmetric retailers. He shows that vertical information sharing has two effects: a direct effect and an indirect effect (leakage effect). He identifies conditions under which information can be traded and examines the impact of information sharing on the total supply chain profits.

The impact of forecasting on pricing has been studied in marketing and economics literature. This literature typically considers the competition in a duopoly that uses different forecasts of the market demand. Vives (1984) identifies conditions under which the sharing of private information between competitors is profitable. Raju and Roy (2000) analyze the impact of firm size, product substitutability, and intensity and mode of competition on the value of forecast information. Researchers have also looked at how to combine information from different sources (Winkler, 1981). Sarvary and Parker (1997) show that information from different sources can be substitutes or complements depending on characteristics such as variance and correlation. Mishra et al. (2001) examine the value of demand forecast sharing in a simple supply chain consisting of a manufacturer and a retailer. They find that demand forecast sharing is beneficial to the manufacturer under any condition and beneficiary to the retailer under some conditions.

Another stream of research related to this work is about direct channel: Balasubramanian (1998) models the competition in the multiple-channel environment with direct marketers and obtains the price equilibrium; Chiang et al. (2003) study the dual channel supply chain design for goods that do not provide a large service (value). Their results show the manufacturer can mitigate the profit loss by adding a direct channel. Geyskens et al. (2002) find that powerful firms with a few direct channels achieve better financial performance than less powerful firms with broader direct channel offerings. Yao and Liu (2003) study the customer diffusion between an e-tail channel and a retailer channel. They find that demands on both channels are stable under certain conditions.

In this paper, we consider a manufacturer–retailer supply chain, which consists of a mix of the traditional retail channel and a direct channel. Customers can purchase products from either the retailer channel or the direct channel. The demand is a linear function of price. The retailer and the manufacturer forecast the primary demand, which is unknown. When forecasts are not shared, the manufacturer and the retailer set the wholesale price and retail price respectively, using their own forecasts. In the information sharing case, the forecasts are shared prior to setting the prices. In order to focus on the pricing aspect of the supply chain as opposed to the inventory aspect considered in prior literature, we first assume that the retailer places an order and the manufacturer makes production after demand is known. We show that the manufacturer and the retailer benefit from information sharing when the manufacturer is optimistic about the demand, i.e., the manufacturer’s forecast is sufficiently larger than the retailer’s. Under this situation, the manufacturer would like to give a price discount to the retailer for sharing the forecast information. This represents a win–win situation for the manufacturer and the retailer. Alternatively, when the manufacturer’s forecast is sufficiently lower (pessimistic) than the retailer’s, information sharing increases the manufacturer’s profit but decreases the retailer’s profit due to double marginalization. In this case, it is not in the interest of the retailer to share forecast information. If the supplier and the retailer jointly maximize their profits, the misalignment of incentives for information sharing caused by double marginalization can be alleviated.

We also compare the manufacturer and the retailer’s profits between the cases with and without the direct channel under some assumptions (production cost is zero, and each demand function has the same slope of price). We find that the direct channel has a negative impact on the retailer’s performance, and the manufacturer and the whole supply chain are better off under some conditions.

We extend our model to consider the case when production takes place before demand is known. In such cases, we find that the manufacturer can offer an appropriate price discount or side payment, such as buying the information from the retailer, to obtain the information from the retailer when the retailer has no incentive to share the information voluntarily. Through a simulation example, we illustrate information sharing’s impact on the supply chain performance.

The rest of the paper is organized as follows: we discuss the model framework in Section 2. In Section 3, we analyze the value of information sharing when production and sale takes place after the demand is known (prices are set before the demand is known). In Section 4, we analyze the value of information sharing when both price and production decisions are made before the demand is known. We present our simulation result in Section 5 and conclude the paper in Section 6.