Chapter 4 The Value of Information Qi Xu Professor of Donghua University Tel: 021-62378860 E-mail: xuqi@dhu.edu.cn

Case: Barilla SpA Part A Barilla SpA is the world’s largest pasta manufacturer The company sells to a wide range of Italian retailers, primarily through third party distributors During the late 1980s, Barilla suffered increasing operational inefficiencies and cost penalties that resulted from large week-to-week variations in its distributors’ order patterns

Barilla distribution patterns Barilla dry product factories 65% TL Barilla CDCs TL 35% 10% TL 18 Barilla run depots 90% 70% 30% Grande Distribution Distribution Organization LTL LTL LTL Chain Supermarkets Independent Supermarkets Signora Maria shops TL=Delivery in truckload quantities LTL=Delivery in less-than-truckload quantities

Weekly Demand for Barilla Dry Products from Northeast DC to CDC

Distribution Distributor Ordering Procedures. Most distributors checked their inventory levels and placed orders with Barilla once a week. Barilla products would then be shipped to the distributor over the course of the week that started 8 days after the order was placed and ended 14 days—the average lead time was 10 days. Most distributors used simple periodic review inventory systems. Nearly all of them had computer-supported ordering systems but few had forecasting systems or sophisticated analytical tools for determining order quantities.

Barilla increasingly felt the effects of fluctuating demand Barilla increasingly felt the effects of fluctuating demand. Orders for Barilla dry products often swung wildly from week to week( see Figure 1). Distributors’ service level to the retailers were unacceptable.Asking distributors or retailers to carry additional inventory to check the fluctuation. Distributors and retailers were already carrying too much inventory.

Questions: What exactly is causing the distributor’s order pattern to look this way? What are the underlying drivers of the fluctuations?

Causes of Demand Fluctuations Transportation discounts Volume discount Promotional activity No minimum or maximum order quantities Product proliferation(激增) Long order lead times Poor customer service rates Poor communication

Demand Fluctuations The extreme fluctuation in Exhibit 1 is truly remarkable when one considers the underlying aggregate demand for pasta in Italy. What does the underlying consumer demand pattern for pasta look like in Italy?

What is the impact of demand fluctuation seen in Exhibit 1? Because the plant has high product change over costs, Barilla has either inefficient production or excess finished goods inventory Utilization of central distribution is low Workers Equipment Transportation costs are higher than necessary

What is the impact of demand fluctuation The distributor must build excess capacity to hold goods bought on any type of promotion, including quantity discounts, truckload discounts and canvass period discounts What if the distributor passes the discount along to the retailers? What is the value of the promotion game?

Barilla SpA Part A (continued) To address this problem, the director of logistics suggests: Implementation of Just-in-Time Distribution (JITD) with Barilla’s distributors. What is the JITD System? Decision-making authority for determining shipments from Barilla to a distributor would transfer from the distributor to Barilla Rather than simply filling orders specified by the distributor, Barilla would monitor the flow of its product through the distributor’s warehouse, and then decide what to ship to the distributor and when to ship it.

Evaluation of the JITD Proposal Clearly the variation in demand is imposing additional costs on the channel. What do you think of the JITD proposal as a mechanism for reducing these costs? Why should this work? How does it work? What makes Barilla think that it can do a better job of determining a good product/delivery sequence than its distributors?

Two Key Concepts Behind JITD Replace sequential optimization with global optimization Who will optimize? Eliminate some of the ‘false’ economics that drive traditional ordering processes What does this mean?

Implementation Issues Resistance from the Distributors “Managing stock is my job; I don’t need you to see my warehouse or my figures.” “I could improve my inventory and service level myself if you would deliver my orders more quickly; I would place my order and you would deliver within 36 hours.” “We would be giving Barilla the power to push products into our warehouse just so that Barilla can reduce its costs.” ?

Implementation Issues Resistance from Sales and Marketing “Our sales levels would flatten if we put this program in place.” “How can we get the trade to push Barilla product to retailers if we don’t offer some sort of incentive?” “If space is freed up in our distributors’ warehouses…the distributors would then push our competitors’ product more than ours.” “…the distribution organization is not yet ready to handle such a sophisticated relationship.” ?

How Can Barilla Solve the Implementation Problems? Demonstrate that JITD benefits the distributors (lowering inventory, improving their service levels and increasing their returns on assets); Run experiment at one or more of Barilla’s 18 depots Barilla needs to look at JITD not as a logistics program, but as a company-wide effort; Get top management closely involved Trust

Barilla (B) Case What did Barilla learn from the experiments in Florence and Milan? How should Barilla change the way it attempts to sell the JITD concept to its distributors? If you were a Barilla distributor, would you sign onto the program after seeing these results?

Evaluation How do you evaluate the implementation process Barilla used with Cortese?

Two important issues: Variations in distributors’ order patterns have caused severe operational inefficiencies and cost penalties for Barilla. Indeed, while variability in aggregate demand for pasta is quite small, orders placed by the distributors have a huge variability. Unlike traditional SC in which distributors place orders and manufacturers try to satisfy these orders as much as possible, in JITD Barilla would specify the appropriate delivery quantities—those that would more effectively meet the end consumer’s needs yet would also

more evenly distribute the workload on Barilla’s manufacturing and logistics systems. Such a strategy has been referred to as vendor managed inventory(VMI).

Questions What are the reasons for the increase in variability in Barilla’s SC? How can the firm cope with the increase in variability? What is the impact of transferring demand information across the SC? Can the VMI strategy solve the operational problems faced by Barilla?

Chapter 5 Value of Information “In modern supply chains, information replaces inventory” Why is this true? Why is this false? Information is always better than no information. Why? Information Helps reduce variability Helps improve forecasts Enables coordination of systems and strategies Improves customer service Facilitates lead time reductions Enables firms to react more quickly to changing market conditions.

The Bullwhip Effect and its Impact on the Supply Chain Consider the order pattern of a color television model sold by a large electronics manufacturer to one of its accounts, a national retailer. Fig 1. Order Stream

The Bullwhip Effect and its Impact on the Supply Chain Fig 2. Point-of-sales Data-Original Figure 3. POS Data After Removing Promotions

The Bullwhip Effect and its Impact on the Supply Chain Figure 4. POS Data After Removing Promotion & Trend

Higher Variability in Orders Placed by Computer Retailer to Manufacturer Than Actual Sales

Increasing Variability of Orders Up the Supply Chain Lee, H, P. Padmanabhan and S. Wang (1997), Sloan Management Review

We Conclude …. Order variability is amplified up the supply chain; upstream echelons face higher variability.

Consequences…. Increased safety stock Reduced service level Inefficient allocation of resources Increased transportation costs

What are the Causes…. Promotional sales Forward buying Volume and transportation discounts Batching Inflated orders IBM Aptiva orders increased by 2-3 times when retailers thought that IBM would be out of stock over Christmas

What are the Causes…. Demand forecasting Long cycle times Order-up-to points are modified as forecasts change – orders increase more than forecasts Long cycle times Long lead times magnify this effect

Consider a simple supply chain… Single retailer, single manufacturer. Retailer observes customer demand, Dt. Retailer orders qt from manufacturer. What if there are promotional activities, no transportation discounts and no inflated order? Are we still going to see an increase in variability? Consider a simple supply chain with a single retailer and a single manufacturer qt Retailer Manufacturer Dt L

Consider a simple supply chain… Average annual demand is 5,200 units, and shipments are made to the store each week. If the variability in orders placed is low, so that the shipment every week is capacity need by only about 100 units. If weekly variability is very high, so that during certain weeks makers must make and ship 400 units and weeks no units at all, it is easy to see that production and shipping capacity must be much higher and that some weeks is capacity will be idle.

Quantifying the Bullwhip Effect Suppose a P period moving average is used. (移动平均法) If the variance of the customer demand seen by the retailer is Var(D), then the variance of the orders placed by that retailer to the manufacturer, Var(Q), relative to the variance of customer demand satisfies: 零售商向制造商发出的订单需求的方差Var(Q)与顾客需求的方差之比满足： 零售商向制造商发出的订单需求的方差Var(Q)与顾客需求的方差之比

Var(q)/Var(D): For Various Lead Times A lower bound on the increase in variability given as a function of p 14 L=5 L=5 12 10 8 L=3 L=3 6 4 L=1 L=1 2 P 5 10 15 20 25 30

Figure shows the lower bound (下限)on the increase in variability as a function of p for various values of the lead time,L. When p is large, and L is small, the bullwhip effect due to forecasting error is negligible. The bullwhip effect is magnified as we increase the lead time and decrease p. Assume p=5, L=1 The variance of the orders placed by the retailer to the manufacturer will be at least 40 percent larger than the variance of the customer demand.

Multi stage SC systems

Multi-Stage Supply Chains Consider a multi-stage supply chain: Stage i places order qi to stage i+1. Li is lead time between stage i and i+1. qo=D q1 Retailer Stage 1 Manufacturer Stage 2 q2 Supplier Stage 3 L1 L2

SC with centralized Demand Information Centralized: each stage bases orders on retailer’s forecast demand. The retailer observes customer demand, forecasts the mean demand using a moving average with p demand observations, finds his target inventory level based on the forecast mean demand, and places an order to the wholesaler. The wholesaler receives order along with the retailer’s forecast mean demand, uses this forecast to determine his target inventory level, and place an order to the distributor. Similarly, the distributor places order to the factory.

SC with centralized Demand Information (cont’) In this centralized SC, each stage of the SC receives the retailer’s forecast mean demand and follows and order-up-to inventory policy based on this mean demand. The variance of the orders placed by the kth stage of the SC, Var(Qk), relative to the variance of the customer demand, Var(D), is just: 供应链上第k阶段发出订单的方差Var(qk)相对于顾客需求的方差Var(D)的关系， 供应链上第k阶段发出订单的方差Var(qk)相对于顾客需求的方差Var(D)的关系:

SC with centralized Demand Information (cont’) For example, if the lead time from the retailer to the wholesaler is two periods, then L1=2. Similarly, if the lead time from the wholesaler to the distributor is two periods, then L2=2, and if the lead time from the distributor to the factory is also two periods, then L3=2. The total lead time from the retailer to the factory is L1+L2+L3=6 This expression for the variance of the orders placed by the kth stage is very similar to the expression in the previous section, with the single stage lead time.

Decentralized Demand information Decentralized: each stage bases orders on previous stage’s demand. The retailer does not make its forecast mean demand available to the remainder of the SC. Instead, the wholesaler must estimate the mean demand based on the orders received from the retailer. The variance of the orders placed by the kth stage of the SC, Var(Qk),relative to the variance of the customer demand, Var(D) satisfies: The variance increases multiplicatively at each stage of the SC.

Multi-Stage Systems:Var(qk)/Var(D) Increase in variability for centralized and decentralized system Dec, k=5 Cen, k=5 Dec, k=3 Cen, k=3 k=1

It is now clear that by sharing demand information with each stage of the SC, we can significantly reduce the bullwhip effect. When demand information is centralized, each stage of the SC can use the actual customer demand data to estimate the average demand. When demand information is not shared, each stage must use the orders placed by the previous stage to estimate the average demand. These orders are more variable than the actual customer demand data, thus, the forecasts created using these orders are more variable, leading to more variable orders.

The Bullwhip Effect: Managerial Insights Exists, in part, due to the retailer’s need to estimate the mean and variance of demand. The increase in variability is an increasing function of the lead time. Centralized demand information can significantly reduce the bullwhip effect, but will not eliminate it.

Coping with the Bullwhip Effect in Leading Companies Reduce uncertainty POS Sharing information Sharing forecasts and policies Reduce variability Eliminate promotions Year-round low pricing Reduce lead times EDI Cross docking Strategic partnerships Vendor managed inventory (VMI) Data sharing

Example: Quick Response at Benetton Benetton, the Italian sportswear manufacturer, was founded in 1964. In 1975 Benetton had 200 stores across Italy. Ten years later, the company expanded to the U.S., Japan and Eastern Europe. Sales in 1991 reached 2 trillion. Many attribute Benetton’s success to successful use of communication and information technologies.

Example: Quick Response at Benetton Benetton uses an effective strategy, referred to as Quick Response, in which manufacturing, warehousing, sales and retailers are linked together. In this strategy a Benetton retailer reorders a product through a direct link with Benetton’s mainframe computer in Italy. Using this strategy, Benetton is capable of shipping a new order in only 4 weeks, several week earlier than most of its competitors.

How Does Benetton Cope with the Bullwhip Effect? 1. Integrated Information Systems • Global EDI network that links agents with production and inventory information • EDI order transmission to HQ • EDI linkage with air carriers • Data linked to manufacturing 2. Coordinated Planning • Frequent review allows fast reaction • Integrated distribution strategy

Information for Effective Forecasts Pricing, promotion, new products Different parties have this information Retailers may set pricing or promotion without telling distributor Distributor/Manufacturer might have new product or availability information Collaborative Forecasting addresses these issues.

Information for Coordination of Systems Information is required to move from local to global optimization Questions: Who will optimize? How will savings be split? Information is needed : Production status and costs Transportation availability and costs Inventory information Capacity information Demand information

Locating Desired Products How can demand be met if products are not in inventory? Locating products at other stores What about at other dealers? What level of customer service will be perceived?

Lead-Time Reduction Why? How? Customer orders are filled quickly Bullwhip effect is reduced Forecasts are more accurate Inventory levels are reduced How? EDI POS data leading to anticipating incoming orders.

Information to Address Conflicts Lot Size – Inventory: Advanced manufacturing systems POS data for advance warnings Inventory -- Transportation: Lead time reduction for batching Information systems for combining shipments Cross docking Advanced DSS Lead Time – Transportation: Lower transportation costs Improved forecasting Lower order lead times Product Variety – Inventory: Delayed differentiation Cost – Customer Service: Transshipment

Supply Sources: plants vendors ports Regional Warehouses: stocking points Field Customers, demand centers sinks Production/ purchase costs Inventory & warehousing Transportation