Barilla spA case analysis Barilla SpA, an Italian pasta manufacturer, is experiencing amplified levels of inefficiencies and rising

costs due to variability in demand from its distributors. The main problem addressed in this case is how to effectively implement JITD system suggested by Giorgio Magialli, the Director of Logistics by resolving the issue of gaining control over the fluctuating demand. Barilla has a very complex distribution network including independent third party distributors and due to such a multi-echlon network, Barilla has been experiencing large amounts of variability in demand which are resulting in operational inefficiency and increased manufacturing, inventory and distribution costs. The proposed JITD system required the distributors to share their sales data with Barilla, who would then forecast and deliver appropriate amounts of products to the distributors at the right time in order to effectively meet demand. This was a radical change from the current and more traditional supply-chain setup where the distributors were not sharing any data and could place orders at will. Vitali's proposal came under severe criticism from not only the distributors but also Barilla's own Sales and Marketing department for an array of reasons. Main reasons for fluctuating demand: Promotions: The use of promotions in the form of price, transportation, and volume discounts was the main strategy to sell more products to the distriutors. Sales Reps: The compensation system in place at Barilla for Sales reps, made them to push more products into the pipeline during promotional periods and not able to sell sufficient quantities during non promotional periods created wide variation in demand patterns. SKU's: The huge range of SkU's in each product line led to greater complexity Gaming bevavior and wrong forecasting practices: The distributors were having full control of their orders to Barilla and used gaming during stock outs periods. The distributors did not use any sophisticated forecasting models or systems to calculate the order quantities without any threshold minimum order quantities but rather just followed a replenishment ordering. Long lead times: Barilla was dealing with perishable items and the lead times of between 8 and 14 days after it received their orders, the average lead-time being 10 days was high given the nature of products and continuously changing ordering patterns. Barilla's cost implications: The fluctuating demand had adverse cost implications for Barilla: The production costs was rising heavily due to frequent changeover for offering wide range of SKU's, and backorders. It also led to production scheduling difficulties and high labor & transportation costs to satisfy the high ordering during promotional periods. Finally the cash flows were also uneven. Reasons for stiff opposition to JITD by Distributors: The pasta industry distributors were not yet prepared for such an sophisticated information sharing process. The distributors were not ready to share their sales data to Barilla and were not aware of the benefits of supply chain collaboration through collaborative forecasting and planning. The distributors were sacred of the negative implications of sharing the sales data thinking of the loss of control over ordering process would make them loose competitive advantage over Barilla through adjustin their order quantities through their own forecasting & planning for attaining the profit

margins, and Barilla would push their products and cut down their costs. They also felt that, they would loose the opportunity to gain advantage during the promotional periods and they cannot place orders at will. Since the distributors were working with nearly 200 suppliers, they did not have the trust comfort level by disclosing the sales data to Barilla, and their relations with other suppliers could be affected, if they were aware of sharing the proprietary data with Barilla. Besides, the distributors, the internal Barilla's sales force were also reluctant to these JITD implementation, as they were sacred of losing their jobs, if the ordering process was automated and the Barilla would not need them for going to the field and putting their efforts to convince their distributors to order more quantities and get sales incentives. Recommendations: The first approach to effectively implement JITD is to convince and educate the internal sales force and the distributors of the benefits of Collaborative Planning Forecasting and Replenishment (CPFR). In order to condense the Bullwhip effect being experienced by Barilla, their supply chain would have to be centralized. The JITD will work, only if Barilla gains control over the inventory throughout their supply chain. Since the pasta industry in Italy is not yet geared towards tracking the point of sales data through scanners and barcodes, hence each stage has to collaboratively make ordering forecasts based on the end customer sales data. The JITD system will eliminate the fixed costs in ordering for distributors through an effective centralized system through out the supply chain and for Barilla, they can improve their production scheduling and reduce the labor and transportation costs. How to effectively implement JITD at Barilla? Collaborative planning and Aggregate Demand: Barilla and their Distributors can effectively team up and build a comprehensive forecasting model, which will take into account the perishable inventory system model, threshold inventory levels, demand fluctuation factors to capture the spikes in the demand by aggregating the demand throughout the supply chain. Demand Planning: Customer surveys can help Barilla and Distributors to understand the needs of their customers and help them to effectively manage and control the demand patterns. Methodology: To convince the internal staff and sales force, Barilla can run a 6-8 month pilot program or educate them by running a simulation of the whole exercise to show the benefits of sharing the sales data. The upper management should involve the sales managers in incorporating their needs and complications involved in executing the JITD in the field. The sales force should be given additional responsibilities for market development through adopting "Go-to-market" strategy in the event the whole ordering process is automated through the JITD system implementation. The compensation system can be revamped into a new level and could still be in place by assigning target sales revenue through above mentioned market development and territory management programs. The pilot programs and simulation exercise cost results should be circulated to the distributors to enable them to understand the cost cutting and better inventory management practices the JITD system can offer, so that the distributors initially

become ready to adopt the JITD implementation in a phase manner. Confidence and trust building with distributors can be achieved through long term strategic partnership development and having constant operational meetings with Barilla's upper management instead of letting the sales force giving their inputs to the effective application of JITD. Policies governing the profit margin sharing can be formalized as part of building the supply chain collaboration strategies. Since Barilla has more than 22% market share outside Italy in Europe, and it becomes easy to implement JITD in Europe locations where there is significant usage of barcodes and scanners systems to capture the POS data in place. The success story of these European locations can convince Italian distributors of the benefits of JITD. Barilla should conduct a study to track the most fast selling pasta SKU's, region wise, so that they can reduce the amount of SKU's they offer and streamline to offer a particular set of product line segmented across each region to reduce the complexity in the supply chain. The huge # of intermediaries of distributors and selling agents is also one of the hindrances for existing inefficiencies, hence Barilla should consider reducing the # of levels in the logistics network. The Efficient Customer Response (ECR) initiative can be a useful tool for Barilla for sharing the POS data all along the supply chain and understand the customer buying preferences and accordingly adjust the # of SKU's to offer in each pasta product line and plan for new product introductions.

INTRODUCTION Barilla SpA (Barilla), is an Italian manufacturer that sells pasta to retailers largely through third-party distributors. Barilla has been experiencing widely fluctuating demand patterns from these distributors. Such unpredictable patterns are problematic because a specific sequence of pasta production is used that minimizes the incremental changes in kiln temperature in order to keep the changeover costs low and the product quality high. This process makes manufacturing unfortunately unresponsive to changes in anticipated demand. In order to address this issue, Brando Vitalis has proposed a Just-in-Time Distribution (JITD) model, which is a continuous replenishment strategy under which the responsibility for determining shipment quantities to the distributors would shift from the distributors to Barilla. Such a system would result in Barilla pushing its pasta to suppliers based upon its demand forecasts. Implementing a JITD system should have the effect of reducing channel costs improving service levels to distributors for Barilla, and improving service and reducing Distributor inventory. PROBLEMS Barilla has been experiencing significant problems in its implementation of the JITD model. Preventing Barilla from effectively implementing a JITD system are: 1) internal opposition from its sales staff; 2) lack distributor buy-in stemming from a fear of loss of power; 3) an inability to collected needed information; and 4) the traditional Italian trade promotions system. EVALUATION CRITERIA The main decision to be made by Barilla is not whether Barilla should apply the JITD model, but whether it can be applied. The benefits of introducing a successful system are numerous to an industry where both manufacturers are suffering from thinning margins. To Barilla a successful JITD would result in a decrease in distribution, inventory and manufacturing costs. To the distributors it would result in decreased inventory costs and a reduction in stockouts. This decision that Barilla faces is how such JITD would be implemented. In evaluating the solutions to the JITDs major problems we will use four main evaluation criteria: 1.) Cost Savings to Barilla The reduction inventory and transportation costs incurred by Barilla. 2.) Cost Savings to Distributors - The reduction in inventory incurred by the distributors. 3.) Service Level As defined as the percent of retailer orders filled from distributors inventory. 4.) Internal resistance Making adjustments to insure that internal buy-in for the JITD is high. SOLUTIONS TO JITD IMPLEMENTATION PROBLEMS LOSS OF POWER An obvious concern of Distributors in allowing Barilla to implement a JITD system is that it gives Barilla the power to push product into distributor warehouses as it wishes. Such a system in the eyes of the distributor can result in obvious abuses as it could be used by Barilla to move inventory from the CDC to the Distributors inorder to decrease costs to Barilla at the expense of the Distributor. This concern can easily be mitigated by

structuring the contractual supplier-distributor relationship in way that gives the Distributor the power to deny any shipment. In this way the Distributor still has the power to say no if it feels that the order is excessive. As most Distributors currently keep a 2 week supply of product on hand, and the JITD proposes to significantly cut inventories, theoretically there should not be many situations where the Distributor will be forced to veto shipments as they should not be many situations where inventory will build up to the point where the distributor is not comfortable with the level. INFORMATION: In order for Barilla to use a JITD it requires that each distributor provide data on what Barilla products it has shipped out of its warehouse to retailers during the previous day, as well as the current stock level for each Barilla SKU. This information can then be used by Barilla to make its own replenishment decisions based on its forecast. Hitherto distributors have been very reluctant to share this information citing their lack of confidence that Barilla can create better estimates. There are three alternative solutions addressing this problem: Firstly, buy information for Distributors. One of the larger Distributors has already offered to sell the information. It is likely that many other Distributors would similarly part with this information for the right price. Unfortunately such an exercise would create a dangerous precedent that could be difficult to reverse in the long run. Though it might be useful to purchase such information until the JITD method catches on it could create an expectation for such payments to continue in the long run. Secondly, gather information from point of sale. Another information source which JITD forecasts could be generated from is data from retail outlets. This is an interesting proposal as retail outlets already take inventory on a daily basis. However, it is unlikely that such a system can be implemented in the near future due to the lack of relationship between Barilla and the retailers and the fact that many retailers do not have the necessary computer equipment to properly monitor and communicate this information to Barilla efficiently. Thirdly, introducing a pilot project. Obtaining information is much more accessible if Distributors are convinced that Barilla can actually reduce costs and increase service through its own forecasting. This can be demonstrated using a pilot project that would demonstrate the difference decrease in stockout rates, and the amount of warehouse space used due to the implementation of the JITD strategy. In order to implement such a pilot project, Barilla should contact a distributor and offer significant discounts in order to get them to sign on. We recommend this third option as it best satisfies the aforementioned criteria, and it is likely to be the most effective in the long run. DISCOUNTS/PROMOTIONS Well established in the Italian grocery industry, promotions and volume discount cause significant fluctuations in demand for Barilla. The promotion system encourages distributors to deplete inventory and buy during the appropriate canvas period, or to make purchases by the multiple truckload to receive further sales representative induced discounts. There are two alternative solutions addressing this problem: Firstly, keep prices fixed/ constant. Instead of constantly adjusting prices through promotions, Barilla instead could decrease all prices equally to all distributors by incorporating the savings that result from the elimination of trade promotions and the

increased efficiencies resulting from the JITD system. Such a strategy would be unfair for the larger distributors who buy more and are accustomed to greater discounts, and such a strategy would also not be aligned with Italys trade promotion culture and may result in significant pushback from Distributors. Secondly, negotiate volume discounts and push canvas discounts. Barilla can still keep its canvas discount system by building it into their forecasting model. Thus allowing Distributors discounts during the appropriate canvas period, but by not allowing them to stock up inventory to maximize these discounts (as only the amount recommended by the forecast is discounted). Furthermore, volume discounts can still be offered in the form of refunds. Instead of giving discounts for each order, the amount purchased over a monthly period can be used to calculate the size of discount the distributor receives. Thus larger purchasers still receive their expected discounts. INTERNAL PUSHBACK Within the current system, the sales representatives are playing important roles, and many fear that their responsibilities in managing inventory and setting up promotion would be reduced if JITD system is implemented. The lack of buy-in especially be the sales team can be a huge problem as they are the ones would are in constant communication with the Distributors. Negative perceptions of the salespeople will reflect poorly on Barillas ability to implement a JITD, and may result in poor Distributor buy-in. Barilla could require that each sales person be trained to monitor the JITD system, and give commissions to those salespeople who convince their distributors to buy-in to the program. In so doing, the sales representative will still play an important role in the new system, and therefore encourages them to participate and support the JITD system. JITD DISTRIBUTOR TARGET The JITD program eventually should be implemented throughout the entire distribution system. However in order to make implementation as smooth of a process as possible it is important to target smaller sections of the distribution channel intially. Three alternatives for an initial distributor implementation exist; the Grande Distribuzione(GD), Distributione Organizzata (DO) and Barilla run depots. The benefits of using GDs are that they are the largest part of the supply chain and presumably they follow more standardized processes within the company. The large size may make it difficult to implement thus it may make more sense to initially use one or two GDs. DOs would be more difficult as they serve independent supermarkets and thus they use less sophisticated and less standardized inventory procedures and are also more vulnerable to pressure from other suppliers. If the GDs do not immediately sign on running a pilot project with the 18 Barilla Run Depots would make the most sense as information would be more available. However as it is an internal part of the distribution chain, its success may be less convincing to sceptical distributors. As such we would suggest that the JITD target the Depots first, the GDs second and the DOs last. Conclusion: The JITD system has the potential to substantial reduce costs if it is implemented correctly. In order to do so Barilla should begin implementing JITD within its own Depots and expand with pilot projects with the Distributors. The aforementioned analysis discussed a variety of methods bywhich to make such implementation more viable than

it was in the past. However, such a list is in no way conclusive. Ideas such as reducing the number of SKUs and rearranging distribution channels should also be explored in the long term as they can also result in substantial cost savings.

Executive Summary Barilla, the leading pasta manufacturer in Italy, faces increasing problems related to demand fluctuation. Their distributors also suffer from high inventory holding costs and low service levels on the other hand. This report explains, why the company and their distributors are troubled with this situation and how Barilla intends to solve it. The problem Barilla experiences is called the Bullwhip Effect, i.e. that demand variability increases when moving up the supply chain. Several factors enforce this Bullwhip Effect, e.g. high lead times, poor demand forecasting, and batch ordering. In this report we will point out, that exactly those aspects can be identified as the underlying reasons for Barillas problems. In a next step, we will analyze the cost dimensions of the JITD concept Barilla developed in order to reduce demand variability. We identified five cost categories where the JITD approach would lead to a decrease in costs, e.g. inventory holding costs, penalty costs and transportation costs. However, cost increases, especially implementation costs, have to be taken into consideration as well. In the last part of the report, we will address the issue of implementation hurdles of the JITD program with regard to distributor resistance, sales force refusal and the problematic handling of trade promotions. The resistance Barilla faces from their distributors can be lessened by convincing them of their competence to forecast so pre-cisely, that inventory levels will decrease and out of stock situations will be damped down as well. Therefore, Barilla might conduct a test phase or offer a warranty. Furthermore, the reluctance from within Barilla, especially from the sales force, has to be taken down by changing the incentive system and clearly defining new tasks for the sales representatives. Concerning trade promotions, the possibility of abolishing promotions and quantity discounts has to be discussed. As conclusion, Barilla needs to improve the overall supply chain cooperation in order share information, reduce the Bullwhip Effect and realize cost savings. Exercise 1 - Underlying causes of difficulties The Barilla company faced increasing problems of demand fluctuations. As you can see in figure 1, the orders of one of their distributors (Corteses Northeast Distribution Center), var-ied widely with high peaks in one week, followed by very low levels of shipments in the next week. These de-mand fluctuations resulted in high inventory costs for Barillas Central Distribution Centers as well as for the dis-tributors in their warehouses. Furthermore, stock out levels at the distributor were quite high (see figure 2). This implies, that the service level towards the retailer suffered, as products were not always available and iretailers had to wait for replenishment. What are the reasons for these demand fluctuations? Pasta in general has a very stable demand, except from some seasonal peaks at Easter and Christmas. That means, consumers buy about the same amount, so that demand patterns should be without high fluctuations. But why does Barilla then face problems of demand variability? This problem is referred to as the Bullwhip Effect. On each level in the supply chain, meaning retailer, distributor and manufacturer, order quantities increase. The retailer will order at his dis-tributor the amount he forecasted to be the consumers demand plus an additional amount of safety stock. The distributor then receives a biased order quantity from the retailer (a higher quantity than the actual forecast). The distributor on the other

hand places his order with the manufacturer for the quantity the retailer ordered plus safety stock for his warehouse. This results in an even higher order quantity than the actual demand forecast on retail level has been. There are several factors influencing the demand variability over the whole supply chain. First of all, the lead time aspect contributes enormously to that problem. Lead time refers to the time between placing an order and receiving a shipment. Barilla had an average lead time of 10 days for shipments to their distributors. Lead times act as a multiplier with regard to demand fluctuations. Even small changes in demand result in either high stock outs or high holding costs due to the time delay in delivery, since the order quantity cannot be adjusted during the lead time. Another factor is demand forecasting. Retailer, distributor and manu-facturer use different forecasting methods (in most cases not very sophisticated), leading to different results and biasing the demand forecast at the end of the supply chain. A third factor contributing to the Bullwhip Effect is the specific production process of pasta. Barilla cannot re-act to demand changes quickly enough, due to production restrictions as e.g. heat and humidity specifications. That means, if a product is out of stock it cannot be produced and delivered immediately. Additionally, forward buying activities from distributors due to frequent trade promotions and volume dis-counts enforce the effect of demand fluctuations. During promotions, distributors buy huge amounts of pasta at a lower price. This forward buying leads to lower demand in future periods and contributes to high peaks in or-der quantity and low levels of orders in the following periods. Connected to this problem is the incentive system of the sales force, which relies heavily on trade promotions in order to achieve quarterly targets. Another factor related to these aspects is batch ordering and order fix costs. During promotions huge amounts are ordered, followed by periods with low order levels in order to benefit from the lower promotion price, avoid fixed ordering costs and fulfill sales targets. Due to the wide variety in product and packaging options Barilla offers (for dry products about 800 different SKUs), demand forecasting on that disaggregated level is complicated. Finally, Barilla follows no global optimization approach, where the costs of the whole supply chain are opti-mized and forecast information are shared. When collaborating with distributors and retailers, demand variability could be reduced and cost savings realized. Exercise 2 - Cost savings and increases When assessing cost savings, that can be realized by implementing the JITD program, the most obvious cost category would be inventory holding costs. The JITD approach offers a huge inventory cost reduction potential. The delivered quantities from manufacturer to distributor are optimized, so that the inventory level at the manufacturers site as well as the inventory level at the distributors warehouses decrease significantly (see figure 3). Thus, inventory holding costs decrease as well. At the same time, the service level can be increased with the help of the JITD program. Stock out rates de-crease, leading to lower back order or penalty costs. Connected to this point are also indirect costs savings as a reduced risk of lost orders and an

improved reputation of the manufacturer Another cost aspect that can be op-timized with the JITD program is transportation. Trucks are filled according to customers demand with a higher fill rate (mix of heavy and light products). Furthermore, Barilla can hire trucks more flexible with less than one week notice in advance. Additionally, a priority of emergency goods has been established, ensuring the supply to out-of-stock retailers (reduction of backorder costs). Finally, the costs for the coordination of forecasting, ordering, producing and shipping goods can be facilitated and optimized by ensuring an efficient collaboration. On the other hand, the JITD approach entails some increases in cost as well. The major cost block would be all costs occurring during the implementation of the forecasting system. The data exchange between the dis-tributor Cortese and Barilla has to be based on a computer system, the Electronic Data Interchange (EDI). Both parties have to implement special programs, employees have to be trained to use these programs, product codes have to be matched etc. The whole process has to be coordinated, responsibilities have to be determined and decision rules have to be decided. Secondly, since Barilla receives shipment data from Cortese, Barilla has to work with this data in order to plan forecasts. The work shifts from the distributor to the vendor, resulting in more effort for Barilla (maybe they have to hire additional employees to evaluate and assess shipment data). Furthermore, transport, truck loads and loading have to be planned more precisely. A third cost block are spendings for presenting the JITD system, acquiring new distributors, demonstrating that Barilla has the competencies to conduct an efficient forecasting and trust building. All these factors have to be considered when analyzing cost aspects of the JITD program. Exercise 3 - Respond to the JITD proposal Vitalis JITD plan faces several conflicts with us, the distributors. As a distributor, I am afraid of losing core com-petencies by accepting the implementation of the JITD plan. I am unwilling to give up control and that is the rea-son why I am reluctant to hand out internal data. Honestly, I am afraid of a high level of dependence on Barilla by providing detailed sales data. Most of the distributors including me are only using simple periodic review sys-tems and no advanced analytical tools so far. So the implementation of the JITD plan might be a very costly ac-tion. The costs of the implementation include training costs and change-over costs. From my point of view, the loss of core competencies might also cost jobs in the distribution centre. Being very dependant on Barilla, weakens our own power in the supply chain: We are no longer a decision-making actor but a logistic agency. Our customers, the retailers, might capitalize on the weak power of the distributor in their daily business by call-ing for even more accurate and fast delivery times. With the implementation of the JITD plan, inventory will decrease and costs will be reduced. Nevertheless, dis-tributors might fear the risk of running out of stock. The reaction of the distributors might vary depending on the management level within the company. Especially people that are responsible for the daily business in logistics might be the most sceptic ones. Exercise 4 Advantages of the JITD Programm

As a distributor, I can tell you what would convince me to sign in: Decreasing costs, no additional risk considerably improved service level, insurance in case of data abuse. Costs are always a big issue, especially in these days. Margins are thinning, additionally Italians react very sen-sitive to pasta prices. I cannot afford high implementing and changeover costs. There has to be a fair cost split-ting between me and Barilla for the implementation of the software system. They say, with JITD I can decrease my inventory level, as well as out of stock situations. If I can really reduce these, it would be a great improve-ment, also with regard to a higher service level for my clients. Usually, less inventory entails a higher risk for out of stock situations. To minimize the risk when switching to JITD I would prefer a test run, e.g. for half a year. So Barilla has the opportunity to prove they have the compe-tencies for JITD. In case of out of stock situations, there has to be a sophisticated emergency plan. Referring to trade promotions, I do not want to have any disadvantage to other distributors, who have not im-plemented JITD and still benefit from trade promotions. Another fact is, that I as a top manager want to be in-volved in the development and decision processes. I have to give up one of my major core competencies, so I want to be able to influence the decision rules they are establishing. There has to be an active cooperation be-tween Barilla and me, so that I can oversee and comment the whole process at every stage. Another big issue is the revelation of my internal data. There must be a kind of insurance against the abuse of my data. If anything happens, I want to be legally protected. Exercise 5 a) Dispirited sales force The Barilla sales force has to understand that the implementation resulting in cost savings, is essential for the companys success. Besides helping retailers with their orders, the Barilla sales force has several other daily tasks they can now intensively address to. The JITD plan is an instrument intended to increase sales rather than decrease sales. Sales representatives have a good basis of negotiation by providing a high service level for the retailers. The former bonus system has to be changed by promoting new incentives to the sales force. Sales force should focus more on POS promotions and on strengthening the relationship to the retailers in order to op-timize the whole supply chain by building up trust. The new responsibilities have to be communicated consis-tently over the whole supply chain. In order to give the Barilla sales force new tasks, a stronger collaboration with the marketing, in particular with the trade marketing, can be pushed forward. With their knowledge of the POS and with their detailed information of their daily business, the sales force is going to remain an important part of Barillas success. This has to be communicated in order to motivate the sales representatives again. Exercise 5 b) - Empty shelf space Truly, this is not a trivial problem. Many distributors may tend to fill free shelf space by buying other products, in the worst case pasta from Barillas competitors. The following approaches can prevent that the JITD program entails negative consequences for Barilla.

Firstly, distributors should be convinced to extend the range and the relative share of Barilla products they are offering to their customers. The high service level and the lower inventory costs are a good argument to prefer Barillas products to competitors products in the long run. The fast inventory-to-sales ratio is another argument to stock more Barilla units, demonstrating the development of a classical distributor towards a modern cross-docking station. Additionally, Barilla marketing and sales force should strengthen their efforts to promote Barilla so that customers and retailer demand the products. Exercise 5 c) - The disgruntled distributor In this case, the distributor felt disadvantaged, because as a member of the JITD program he did not get any discounts, while his competitor was able to swamp the market Barilla products. Two aspects have to be taken into consideration in order to solve this problem. Firstly, in the long run, joining the JITD program has to be more profitable for distributors regarding cost savings than the short term benefits realized by discounts or promo-tions. The huge cost reduction potential has to be the argument to participate in the program and abandon pro-motional discounts. Secondly, Barilla should rethink its order policies in general. There are three options in changing the policy, ei-ther abolishing quantity discounts and promotions, restricting them or adjusting the system. When abolishing quantity discounts, Barilla risks losing distributors that do not participate in the JITD program, as they are reluctant to buy without discounts. However, in the long run these distributors will be forced to buy with Barilla, since Barilla is market leader for pasta and retailers demand their products. When there is no possi-bility to make a bargain through promotions or discounts, those distributors might decide to join JITD in order to realize cost savings instead. On the other hand, restricting promotions entails an administrative effort to control these restrictions and would not exactly solve the situation. Thirdly, the promotional system can be adjusted by offering overall discounts (e.g. on a yearly basis). This avoids irregular peaks in order quantities. Exercise 5 d) - Trade Promotions If every distributor participates in the JITD program, sales promotion will no longer be an issue. The distributors demand will be flattened, costs are decreasing at every level of the supply chain. Until then there has to be a decision rule, how to deal with distributors, who want to ob-tain promotional discounts. In order to place distributors, who have implemented JITD, not at a disadvantage, trade promotions should be abolished. But this will assuredly lead to a revolt of the other distributors. A second possibility is only allowing the not participating distributors to get promotions. As a result, the others are discriminated for implementing JITD. They will lose customers, who will transfer to the retailer with the lowest pasta price. If there are two distribu-tors in a target area and only one is using JITD, you can allow the JITD- user to participate in trade promotions over a certain period of time. There is the possibility to introduce an interim solution timely graded discount, from high percentage, e.g. 4% at the beginning to low per-centage, e.g. 1% at the end of the transfer period. The abolition of the trade promotions has to be associated with increasing POS promotions and the implementation of a yearly bonus sys-tem for distributors. We have to make clear to all of our distributors, that in the long

run they will realize cost savings with JITD exceeding the saving they gain with sales promotions. Until then conflicts cannot be avoided.

Executive Summary

Barilla SpA, an Italian pasta manufacturer, is experiencing amplified levels of inefficiencies and rising costs due to variability in demand from its distributors. In order to bring things back in order and to improve margins, Giorgio Magialli, the Director of Logistics at Barilla wants to implement a Just-In-Time Distribution (JITD) system that was proposed by his predecessor Brando Vitali. This system is entirely different from the existing setup and is being opposed by both the distributors and Barillas Sales and Marketing Department.

In this report we have studied the reasons for this opposition by various quarters and have suggested recommendations, which will allay this lack of support. We consider that in order to condense the Bullwhip effect being experienced by Barilla, their supply chain would have to be Centralized. We have given our rationale for the JITD system to work and have suggested recommendations to resolve all existing issues. We think that a centralized supply chain with Barilla controlling the orders will result in better margins for all the partners.

We believe that by following our recommendations, Barilla will succeed in influencing its distributors and Sales personnel to work together and implement the JITD program. This will not only result in better performance in terms of time and money but also promote trust and good relations among all the partners in the supply chain.

Introduction

Barilla SpA, an Italy based company, is the worlds largest Pasta manufacturer. It has a 35% market share in Italy and a 22% market share in Europe. In addition to the family of pastas (macaroni, spaghetti, fusilli, etc.) it also manufactures bread, cookies, biscuits, rusks, sauces, breadsticks, etc. Barilla has a very complex distribution network consisting of Grand Distributors (owned by large Supermarket chains), Organized Distributors (independent third party distributors) in addition to its own depots. Due to such a complex and multi-echelon network, Barilla has been experiencing large amounts of variability in demand which are resulting in operational inefficiency and increased manufacturing, inventory and distribution costs.

Brando Vitali, Barillas ex-Director of Logistics, had proposed a Just-In-Time Distribution (JITD) system to counter this demand variation. This system required the distributors to share their sales data with Barilla, who would then forecast and deliver appropriate amounts of products to the distributors at the right time in order to effectively meet demand. This was a radical change from the current and more traditional supply-chain setup where the distributors were not sharing any data and could place orders at will. Vitalis proposal came under severe criticism

from not only the distributors but also Barillas own Sales and Marketing department for an array of reasons.

Giorgio Magialli, Barillas current Director of Logistics, is trying to implement this idea since two years but has not made much progress. Our objective in this report is to study the problems and come up with recommendations to resolve them.

Problem Description

We will attempt to first list all the problems that are faced by Barilla and the distributors so that we can better understand them in order to make recommendations that will eliminate them. The main problem in this case is the fluctuating demand. Once this is brought under control, many other problems will be solved. Some of the causes of this fluctuating demand are:

Promotions: Barillas sales strategy relied heavily on the use of promotions, in the form of price, transportation and volume discounts. They divided the year into 10 to 12 canvass or promotional periods, during which different products were offered at discounts. These price discounts ranged from 1.4% to 10%. Barillas volume discounts consisted of carton discounts offered by sales representatives and the transportation discounts consisted of free shipping to the distributors. Sales Representatives: The compensation system for the sales reps was flawed in the sense that they were rewarded based on the amount of the products that they sold to the distributors. This was causing problems as the sales reps would try and push more products during the promotional period to get a bonus and were not able to sell as much during nonpromotional periods. This led to wide variation in demand and made forecasting very difficult. Large number of SKUs: Barillas dry products (the focus of the JITD proposal) were offered in 800 different packaged stock keeping units (SKUs). Most of the popular products were offered in as many as 8 different packaging options. These large numbers led to greater complexity. Gaming Behavior: The distributors were used to having full control of their orders to Barilla and indulged in gaming by ordering different quantities in different periods. This led to variation in demand. Bad forecasting by Distributors: The distributors did not have forecasting systems or sophisticated analytical tools for determining order quantities and this resulted in bad forecasts. Absence of Maximum or Minimum order quantities: Barilla does not require its distributors to order any minimum quantities every time it places an order. This causes the distributors to order fewer quantities more often and increases variability and Barillas

production costs. Also, there are no maximum order quantities during promotional periods, thereby allowing distributors to order large quantities at low prices and thus creating a false demand. Long Lead Times: Barilla supplied its distributors between 8 and 14 days after it received their orders, the average lead-time being 10 days. This was slightly long and a lot could change in the supply chain during this period, causing rise in variability.

This fluctuating demand had many adverse implications for Barilla. Some of them are:

Production Costs: These costs increased drastically as the demand variation increased. This was because pasta production was a very delicate process and the drying of different types of pasta required precise levels of temperature and humidity that could not be changed fast enough to produce different types of pasta in demand. As Barilla struggled to cope up with the fluctuating demand, the production costs rose as changeover was costly and lead times as well as backorders increased. Scheduling Difficulties: It was difficult for Barilla to schedule its resources and facilities like Labor, Machinery and Trucks due to demand variations. They would tend to overuse the resources when the demand was high and on periods of low demand the resources would be scantily used. This was resulting in increased costs of Labor and maintenance. Transportation Costs: As a result of uneven demand, Barilla had to transport more products to its distributors in periods of high demand in lesser time than when compared to periods of low demand. The greater number of trucks, etc that it required in periods of high demand increased its Transportation costs. Cash Flow: The cash flow to Barilla was very uneven due to fluctuations as well as promotions.

All these factors prompted Vitali to propose the JITD system. But this led to great uproar and opposition among the distributors and even Barillas own Sales and Marketing Department. The perceived causes of this opposition among the distributors are:

Untraditional and Revolutionary idea: The JITD system proposed something very new and unheard of. It entailed disclosure of sales data by the distributors to Barilla who would then forecast for the entire supply chain and provide products in the right quantity and at the right time to meet demand. The distributors had reservations to this idea, as they did not want to share their sales data and did not believe that this system would work. Loss of Control: The distributors felt that by allowing Barilla to access their sales data and order for them, they would be giving control of their business to Barilla who would push

their products and cut their own costs. The distributors were scared that they would lose their capacity to game if they could not place orders at will. Lack of Trust: There were major trust issues that needed to be resolved between the distributors and Barilla. The distributors did not trust Barilla enough to let them in their own business and reveal proprietary data. This was because the distributors typically handled products from around 200 different suppliers and they felt that by disclosing their sales data to Barilla, their relations with other suppliers would be adversely affected. Fear of Disintermediation: The distributors felt that if they allowed Barilla to control the order process, they would be of no use in the supply chain and would eventually be disintermediated (removed).

The reasons for internal opposition in Barilla are:

Job Cuts: The Sales personnel feared that if Barilla would forecast and place orders for the distributors, they would not be required anymore to persuade the distributors to buy products and would be laid off. Compensation System: The Sales force feared that they would lose control over their compensation methods. Presently, they were compensated on the quantity of stock that they sold to the distributors but with the JITD system in effect they would lose this opportunity to make more money, as they would not be responsible for selling products to the distributors. Lack of Belief: Moreover the Sales department did not believe that the JITD system would work. They felt that it was too complex a relationship to work and that the distributors would not buy if incentives were not offered to them.

In order to make the JITD system work, Magialli had to convince first his own Sales force and then the external distributors.

Recommendations

After listing out the problems and studying them in detail, we have come up with our recommendations to reduce uncertainty in demand. We think that the best way to do this will be to implement the JITD proposal. In order to do this, the distributors as well as Barillas Sales and Marketing will have to be convinced that this proposal will do more good to them than bad. The distributors will have to be offered collaboration incentives to join the bandwagon. Our recommendations are geared towards assuaging their doubts and worries and getting them on

board. In addition to this we have some other recommendations to improve the supply chain performance, which are listed in the end of the report.

Firstly, we try and answer the whys and hows of the process in order to convince the skeptics. Why should JITD work?: We think that the JITD system will work because of the following reasons: Centralization: As Barilla will have control on the inventory through the entire supply chain, it will be its responsibility to ensure global optimization and benefits for everyone. Centralization will ensure that each stage in the chain has actual customer data making forecasting more accurate. In this case as it is very difficult to get the actual point-of-sales data from the many retailers due to their sheer number and geographical difficulties, the forecasting will have to be done based on the sales data of the distributors. Elimination of excess costs: A centralized system will eliminate needless costs throughout the supply chain. For the distributors, it will mean no more fixed ordering costs as Barilla can plan its production schedule better. For Barilla it will result in cheaper production and distribution costs.

How will JITD work?: The JITD system will work using the following innovative techniques: Collaborative Planning: Barilla and the distributors can work as a team and plan on different issues to smoothen out the creases in the supply chain. Demand Management: Barilla should not only forecast demand but also should try to shape and influence it. This can be done using market surveys to determine the needs and expectations of the consumers and designing products to satisfy them. Forecasting would become much easier and reliable if the demand can be managed. Aggregate Demand: As Barilla is going to decide order quantities for all distributors, it can aggregate the demand and thus ensure better forecasts and reduced costs. This is due to the fact that even if some of the distributors demand more than what was forecasted, the slack in demand from others would offset this increase. Forecasting Techniques: Barilla being a big company has access to the latest and most sophisticated forecasting techniques. It should use these methods to ensure accurate forecasts thereby reducing uncertainty.

Overcoming Internal Resistance: Before, Magialli can try and convince the distributors, he has to convince everyone at Barilla that the JITD system will be a success. If everyone within the company is convinced and they provide a single front to the distributors, Barilla has a better chance of convincing the distributors. Some of the steps that Barilla will have to take in this regard are:

 Convince and use the Upper Level Management: Magiallis task will be to first convince the Upper management that the JITD plan will work. If they are convinced of its feasibility, they will offer full support in trying to convince the Sales department. It will make a lot of difference if the Upper management talk to the Sales people than if just their Logistics peers talk to them. Involve the Sales Department in all stages: To make the Sales people feel important and wanted, Barilla should involve them in all stages of the JITD proposal. They should be made to feel that they would have lots to do even after the system comes into effect. No Job Cuts: The Sales personnel should be assured that they would not be losing their jobs due to the JITD system coming into effect. They should be shown that even if they will not be doing the work that they were up to this point, they will still have lots of work to do even with the system in effect. Barilla should explain to them that they (the Sales force) would be needed to share the extra responsibility that the Logistics department will have to take as a result of planning for all the distributors. Also, the sales people will still be needed to maintain close relationships with the distributors, to solve any problems the distributors have and also to work with the retailers to ensure complete satisfaction of Barillas customers. They would also be needed to work in the stores as usual to set up in-store promotions, discuss strategy and monitor demand for Barillas products to help the company forecast better. New Compensation System: The Sales department should be informed that they would still be rewarded for good performance, except that in this case it was the performance of the entire company as a whole and not just they individually. The new system would award a percentage of Barillas profits to the employees as a bonus. They should be shown that the reduction in costs due to the JITD system would ensure greater margins for Barilla that would then be passed on to the employees. Trial run of the JITD system: This recommendation is common with the next section and has been explained in detail there.

Convincing Barillas Distributors: The distributors are not ready to implement the JTID proposal due to reasons discussed above in the Problem Description section. Barilla will have to convince them that the plan will work and not only benefit Barilla but also the distributors participating in it. Some recommendations to convince the distributors are: Show them the money: This, as always, is the foremost issue on any business enterprises mind. If an enterprise can be convinced that it will make more money, it will be ready to change its current practices. The distributors should be shown that the costs of the system would reduce so much that everyone in the supply chain would save considerable sums of money. Address Trust Issues: This is a major issue between Barilla and its distributors and to solve this Barilla should work out some confidence building measures which would prove to the distributors that Barilla was worth trusting and that they were making the right decision. Some of the steps that could be taken to increase trust are:

Barilla should have a transparent policy and should not hide any of its plans from the distributors. This is to ensure that the distributors are not wary of what Barilla might be planning and how it would affect them. Instead of sending the Logistics personnel to talk to the distributors, Barilla should send the Upper Management. This would make the distributors feel important and would help in convincing them better. Barilla should make compromises in its business in order to help its distributors. Barilla should work to improve the distributors business better thus increasing goodwill with them. Barilla should initiate joint decision-making and planning with its distributors to show them that it cared for each one of them and wanted them to be involved in the process. Barilla should learn and understand the rights and limits of its distributors so that it does not force them to do things that they are not capable of. Trial Run of the JITD system: This is Barillas best chance to convince the skeptics of the system. Barilla should try and implement the system using its 18 depots that supply to the small independent stores as well as some of the supermarkets. This system could be implemented from 6 months to a year depending on the time available to Barilla. If this system were a success, i.e. Barilla and the depots experienced reduction in demand variability and costs, the critics would be effectively silenced. Problems in the Traditional method: Barilla could show its distributors that following the traditional supply chain methods that were in use currently would lead to greater costs and lead times for the distributors if Barilla had to reduce variability and its own costs. The distributors would have to pay additional fixed order costs, would have to operate under minimum and maximum order quantities and would have longer lead times to cope up with as Barilla tried to work with the increased demand fluctuation. Penalty clause: Barilla could sign an agreement with its distributors that if they would be a part of the JITD program and would lose money as a result, Barilla would be penalized and would have to pay the difference in costs to the distributors. This would show the distributors that Barilla meant business and was very confident of the success of the JITD system. Inefficiency of the Distributors: Barilla could prove the inefficiency of the distributors by showing them the mismatch between the inventory held and the actual demand in the current system. For example in week 31 in 1989, the inventory held at the Cortese DC was 1000 quintals while the actual demand was 310 quintals. This mismatch was taking a heavy toll in terms of costs. Partnership Issues: Barilla should convince its distributors that it is the only party in their supply chain that can be the leader and thus they should rally with Barilla. This is because of Barillas sheer size as well as its capability and market share. It will be much easier for Barilla to collect the sales data of all the distributors and forecast the demand than vice versa. Also Barilla would have more chance of accessing competitors information as also the ability to co-ordinate various supply chain partners.

Other Recommendations: We have some other recommendations that we believe will help to reduce demand uncertainty and eventually improve overall supply chain performance. They are: JITD in Exports: Barilla has a 22% market share in Europe and this share is expected to rise as the growth in the export market increases by 20-25% each year. Barilla could try and implement the JITD system with its distributors, both current and future, outside Italy. This could act as a test run to be shown to the distributors in Italy. In addition to this Barilla would ensure that its distributors outside Italy would already be accustomed to this idea so that there would be no opposition offered later. If implemented and successful, this idea would result in savings for Barilla in its export supply chain. Reduce SKUs: Currently, Barillas dry products are offered in 800 different SKUs. This large amount leads to a lot of complexity in the order process, distribution as well as inventory and thus increases costs and variability. Some of its famous products are offered in up to 8 different packaging options. Barilla should reduce the number of SKUs to around 500 for better control. For example the 0.5kg packets could be phased out in favor of the 1kg packets. Actual Point-of-Sales Data: As of now, the JITD system proposes to use the sales data from the distributors as actual customer demand is difficult to obtain from the retailers due to their sheer number. Barilla could eventually invest in technology and provide each of its retailers with computers to have access to point-of-sales data. This would make the supply chain truly centralized and would further reduce variability. However this would require a huge investment form Barilla. Streamlined Distribution System: Barilla could try and make its distribution system leaner. Instead of moving its products through so many levels, it would make more sense to move it through limited levels. It could try and open its own warehouses near centers of demand and could ship directly to its retailers to reduce variability and complexity in the supply chain. CPFR: Barilla could use a CPFR (Collaborative Planning Forecasting and Replenishment) strategy where manufacturers, distributors and retailers work together to plan, forecast and replenish products. This strategy would increase the fill-rates while reducing the inventories. This is due to the fact that for the same level of fill-rate, a CPFR strategy would require lesser inventory due to better forecasting. ECR: Another initiative that could be used by Barilla to reduce the variability in demand is an ECR (Efficient Customer Response) strategy. This idea involves sharing point of sales data between various links in the supply chain ensuring better replenishment, product introductions and promotions.

Conclusions

Barilla can look forward to a much more efficient and profitable future as a result of implementing these recommendations. Their demand would become more steady and easier to produce and transport. Their operating costs will reduce and these savings would result in better margins. The cycle times would decrease and the cash flow would become even.

The JITD system would result in more cooperation between the distributors and Barilla. This would increase trust and encourage collaboration in diverse fields such and Product Planning or Facility Location. A leaner Barilla would be more agile and competitive which could result in capturing more market share both within and outside Italy. More savings for Barilla could translate in growth and diversification in newer areas of business.

The future can only get better for Barilla, in more ways than one.