Obstacles to Coordination in a Supply Chain

Any factor that leads to either local optimization by different stages of the supply chain or an increase in information delay, distortion, and variability within the supply chain is an obstacle to coordination. If managers in a supply chain are able to identify the key obstacles, they can then take suitable actions to help achieve coordination. We divide the major obstacles into five categories:

• Incentive obstacles
• Information-processing obstacles
• Operational obstacles
• Pricing obstacles
• Behavioral obstacles

1. Incentive Obstacles

Incentive obstacles occur in situations when incentives offered to different stages or participants in a supply chain lead to actions that increase variability and reduce total supply chain profits.

local optimization within functions or stages of a supply chain Incentives that focus only on the local impact of an action result in decisions that do not maximize total supply chain surplus. For example, if the compensation of a transportation manager at a firm is linked to the average transportation cost per unit, the manager is likely to take actions that lower transpor­tation costs even if they increase inventory costs or hurt customer service. It is natural for any participant in the supply chain to take actions that optimize performance measures along which they are evaluated. For example, managers at a retailer such as Kmart make all their purchasing and inventory decisions to maximize Kmart profits, not total supply chain profits. Buying deci­sions based on maximizing profits at a single stage of the supply chain lead to ordering policies that do not maximize supply chain profits (see Chapters 11, 13, and 15).

SALES FORCE incentives Improperly structured sales force incentives are a significant obsta­cle to coordination in a supply chain. In many firms, sales force incentives are based on exceed­ing sales thresholds during an evaluation period of a month or quarter. The sales typically measured by a manufacturer are the quantity sold to distributors or retailers (sell-in), not the quantity sold to final customers (sell-through). Measuring performance based on sell-in is often justified on the grounds that the manufacturer’s sales force does not control sell-through. For example, Barilla offered its sales force incentives based on the quantity sold to distributors dur­ing a four- to six-week promotion period. To maximize their bonuses, the Barilla sales force urged distributors to buy more pasta toward the end of the evaluation period, even if distributors were not selling as much to retailers. The sales force offered discounts they controlled to spur end-of-period sales. This increased variability in the order pattern, with a jump in orders toward the end of the evaluation period followed by few orders at the beginning of the next evaluation period. Order sizes from distributors to Barilla fluctuated by a factor of up to 70 from one week to the next. A sales force incentive based on sell-in thus results in order variability being larger than customer demand variability because the sales force tends to push product toward the end of the incentive period.

2. Information-processing Obstacles

Information-processing obstacles occur when demand information is distorted as it moves between different stages of the supply chain, leading to increased variability in orders within the supply chain.

FORECASTING based ON orders and NOT CUSTOMER Demand When stages within a supply chain make forecasts that are based on orders they receive, any variability in customer demand is magnified as orders move up the supply chain to manufacturers and suppliers. In sup­ply chains where the fundamental means of communication among different stages are the orders that are placed, information is distorted as it moves up the supply chain (see Chen, Drezner, Ryan, and Simchi-Levi [2000] for a good quantitative analysis). Each stage views its primary role within the supply chain as one of filling orders placed by its downstream partner. Thus, each stage views its demand as the stream of orders received and produces a forecast based on this information.

In such a scenario, a small change in customer demand becomes magnified as it moves up the supply chain in the form of customer orders. Consider the impact of a random increase in customer demand at a retailer. The retailer may interpret part of this random increase as a growth trend. This interpretation will lead the retailer to order more than the observed increase in demand because the retailer expects growth to continue into the future and thus orders to cover for future anticipated growth. The increase in the order placed with the wholesaler is thus larger than the observed increase in demand at the retailer. Part of the increase is a one-time increase. The wholesaler, however, has no way to interpret the order increase correctly. The wholesaler simply observes a jump in the order size and infers a growth trend. The growth trend inferred by the wholesaler will be larger than that inferred by the retailer (recall that the retailer increased the order size to account for future growth). The wholesaler will thus place an even larger order with the manufacturer. As we go farther up the supply chain, the order size is magnified.

Now assume that periods of random increase are followed by periods of random decrease in demand. Using the same forecasting logic as earlier, the retailer will now anticipate a declin­ing trend and reduce order size. This reduction will also become magnified as we move up the supply chain.

LACK OF INFORMATION SHARING The lack of information sharing between stages of the sup­ply chain magnifies the information distortion. A retailer such as Walmart may increase the size of a particular order because of a planned promotion. If the manufacturer is not aware of the planned promotion, it may interpret the larger order as a permanent increase in demand and place orders with suppliers accordingly. The manufacturer and suppliers thus have much inventory right after Walmart finishes its promotion. Given the excess inventory, as future Walmart orders return to normal, manufacturer orders will be smaller than before. The lack of information shar­ing between the retailer and manufacturer thus leads to a large fluctuation in manufacturer orders.

3. Operational Obstacles

Operational obstacles occur when actions taken in the course of placing and filling orders lead to an increase in variability.

ORDERING IN LARGE LOTS When a firm places orders in lot sizes that are much larger than those in which demand arises, variability of orders is magnified up the supply chain. Firms may order in large lots because a significant fixed cost is associated with placing, receiving, or trans­porting an order (see Chapter 11). Large lots may also occur if the supplier offers quantity dis­counts based on lot size (see Chapter 11). Figure 10-2 shows both the demand and the order stream for a firm that places an order every five weeks. Observe that the order stream is far more erratic than the demand stream.

Because orders are batched and placed every five weeks, the order stream has four weeks without orders followed by a large order that equals five weeks of demand. A manufacturer sup­plying several retailers that batch their orders faces an order stream that is much more variable than the demand the retailers experience. If the manufacturer batches its orders to suppliers, the effect is further magnified. In many instances, there are certain focal-point periods, such as the first or the last week of a month, when a majority of the orders arrive. This synchronization of orders further exacerbates the impact of batching.

LARGE REPLENISHMENT LEAD TIMES Information distortion is magnified if replenishment lead times between stages are long. Consider a situation in which a retailer has misinterpreted a random increase in demand as a growth trend. If the retailer faces a lead time of two weeks, it will incorporate the anticipated growth over two weeks when placing the order. In contrast, if the retailer faces a lead time of two months, it will incorporate into its order the anticipated growth over two months (which will be much larger). The same applies when a random decrease in demand is interpreted as a declining trend.

RATIONING AND SHORTAGE GAMING Rationing schemes that allocate limited production in proportion to the orders placed by retailers lead to a magnification of information distortion. This can occur when a high-demand product is in short supply. In such a situation, manufacturers come up with a variety of mechanisms to ration the scarce supply of product among various dis­tributors or retailers. One commonly used rationing scheme is to allocate the available supply of product based on orders placed. Under this rationing scheme, if the supply available is 75 percent of the total orders received, each retailer receives 75 percent of its order.

This rationing scheme results in a game in which retailers try to increase the size of their orders to increase the amount supplied to them. A retailer needing 75 units orders 100 units in the hope of getting 75. The net impact of this rationing scheme is to artificially inflate orders for the product. In addition, a retailer ordering based on what it expects to sell gets less and as a result loses sales, whereas a retailer that inflates its order is rewarded.

If the manufacturer is using orders to forecast future demand, it will interpret the increase in orders as an increase in demand, even though customer demand is unchanged. The manufac­turer may respond by building enough capacity to be able to fill all orders received. Once suffi­cient capacity becomes available, orders return to their normal level because they were inflated in response to the rationing scheme. The manufacturer is now left with a surplus of product and capacity. These boom-and-bust cycles thus tend to alternate. This phenomenon is fairly common in the electronics industry, in which alternating periods of component shortages followed by a component surplus are often observed.

4. Pricing Obstacles

Pricing obstacles arise when the pricing policies for a product lead to an increase in variability of orders placed.

LOT-SIZE-BASED QUANTITY DISCOUNTS Lot-size-based quantity discounts increase the lot size of orders placed within the supply chain (see Chapter 11) because lower prices are offered for larger lots. As discussed earlier, the resulting large lots magnify the bullwhip effect within the supply chain.

price fluctuations Trade promotions and other short-term discounts offered by a manufac­turer result in forward buying, by which a wholesaler or retailer purchases large lots during the discounting period to cover demand during future periods. Forward buying results in large orders during the promotion period followed by very small orders after that (see Chapter 11), as shown in Figure 10-3 for chicken noodle soup.

Observe that the shipments during the peak period are higher than the sales during the peak period because of a promotion offered. The peak shipment period is followed by a period of low shipments from the manufacturer, indicating significant forward buying by distributors. The pro­motion thus results in a variability in manufacturer shipments that is significantly higher than the variability in retailer sales.

5. Behavioral obstacles

Behavioral obstacles are problems in learning within organizations that contribute to information distortion. These problems are often related to the supply chain structure and the communica­tions among different stages. Some of the behavioral obstacles are as follows:

1. Each stage of the supply chain views its actions locally and is unable to see the impact of its actions on other stages.

2. Different stages of the supply chain react to the current local situation rather than trying to identify the root causes.
3. Based on local analysis, different stages of the supply chain blame one another for the fluc­tuations, with successive stages in the supply chain becoming enemies rather than partners.
4. No stage of the supply chain learns from its actions over time because the most significant consequences of its actions occur elsewhere. The result is a vicious cycle in which actions taken by one stage create the very problems that the stage blames on others.
5. A lack of trust among supply chain partners causes them to be opportunistic at the expense of overall supply chain performance. The lack of trust also results in significant duplication of effort. More important, information available at different stages either is not shared or is ignored because it is not trusted.

Source: Chopra Sunil, Meindl Peter (2014), Supply Chain Management: Strategy, Planning, and Operation, Pearson; 6th edition.