Barilla SpA (A) Case Report
SCMP Module 5

By

Jamie Brymer

Robert Greene February 12th, 2016
Table of Contents

Executive Summary 3-4

Issue Identification 5-7

Environmental and Root Cause Analysis 8-9

Alternatives or Options 10-12

Recommendations and Implementation 13

Monitor and Control 14

Exhibits Attached

Executive Summary

Barilla SpA, is an Italian pasta manufacturer and distributor and even with an attractive piece of the market (35% in Italy, 22% in Europe) and diversifying product lines, they are experiencing inefficiencies and rising costs due to irregular and ever changing demand from their distributors. Giorgio Magialli, the Director of Logisitics, has been tasked with resolving these issues by gaining control over their fluctuating demand. Barilla currently has in place, a very complex distribution network and has been experiencing increasing variable changes which is not only hitting operations for their inefficiency but increasing costs across the board for inventory, distribution and manufacturing. After much backlash the first time of trying to implement a Just-In-Time Distribution (JITD), the plan is to regroup and decide whether this time of program can be implemented successfully and if so, which customers they should target with this type of system. The first attempt at implementing the JITD strategy was done by Brando Vitali, the former Director of Logistics, he had proposed that a JITD system would put an end to demand variation and enable their distributors to reduce their inventory on hand, therefore reducing costs, thereby giving them additional service tools at no additional cost. The only thing required would be that the distributors would need to share their sales data, on hand inventory, etc with Barilla who would then be able to forecast and deliver appropriate amounts of products to the distributors in a Just-In-Time approach. Meaning inventory should be arriving at their warehouse at the appropriate time to effectively meet the demand of their customers. However, change is not easy to implement with companies that have been around a long time and are used to one way of doing things. In fact, Marconi initially was warm to the idea on the first try, however once the idea started to move through the company, resistance started bubbling to the surface. At the end, the objections continued to pile on and Marconi would only agree to sell the information that Barilla would need, and they would continue to make all decisions on replenishments and when the product would be coming in to the warehouse. After examining the situation and reading through the challenges outlined by both Barilla and their distribution network, the recommendation would be to once again attempt to implement the JITD system. This system will work as it’s based on Barilla’s information that is shared by their distributors when determining their production schedule instead of relying on distributor orders to drive production. The Just-In-Time Distribution system will also alleviate other issues within Barilla such as decreasing costs in manufacturing, inventory, and distribution as they will be using objective data to plan out production runs. This will also allow distributors to reduce their inventory and reduce stock outs as inventory will arrive at the same time it will be needed. In addition to cost reduction and better production planning, a JITD implementation would increase a distributor’s need for Barilla product and should result in a better working relationship between the two parties. Once the system is in place, it will be closely monitored to ensure its success, co-operation between Barilla and their distributors will be the key to it succeeding.

Issue Identification The underlying issue is that Barilla is facing increasing costs revolving around the requirement to react to their ever changing demand from week to week of product. Sales data in the 1980’s would simply be not easy to obtain due to technological disadvantages over today. Reacting to market demand for inventory would be near impossible in an efficient manner. Due to the fluctuating inventory needs resulting from the distributor’s weekly variations (See attached Exhibit 12 from Barilla SpA Case Study), Barilla finds they have to build up a huge inventory or suffer a low fill rate, neither of which is acceptable or profitable. There is several reasons for the demand fluctuating as much as it does, and will be outlined below to identify the main causes.
Forecasting - Without a doubt, the largest issue facing Barilla right now, as distributors have mainly been using physical count methods to determine what they should be ordering, this can cause manufacturing and logistics nightmares for Barilla. With no available POS figures at the customers site, lack of actual sales forecasting from distributors, and no accounting system in place, Barilla are at the mercy of the person doing the counts, which may or may not be accurate at the time but only gives one small piece of the forecasting puzzle.
Promotional Activities - The use of promotions in the form of price, transportation, and volume discounts was the main strategy to sell more products to the distributors. The year is divided into 10-12 periods, during which different products are used for the promotions at discounts of anywhere between 1.4% and 10% which caused a major impact to demand variations. These fluctuations are the main cause of the supply issues, Barilla had to overproduce during these times creating higher demand and inventory buildup in their distribution centers. The DC’s would then take measures to try and clear out the inventory buildup such as pushing the discounted product to distributors which started the cycle all over again as the price was driving the demand up.
Transportation Incentives – Barilla would offer discounts to distributors/customers based on full truckloads of product. The downside to this is that the distributors would invest heavily on product, overloading them on certain items that may not be hot sellers and force them to carry excess inventory for some time, forcing them to buy different products the next time the truckload incentives were used, and again, may be carrying excess inventory on that product. This is a vicious cycle that will eventually catch up to the distributor. Since Barilla controls the product shipping from their plant to the distribution centers, their transportation costs would have been reduced however other costs would have gone up such as inventory holding costs, transportation, and extra labor. (See attached Exhibit 8 from Barilla SpA Case Study)
Order Lead Times – The distributors would place order with Barilla once per week, and the average time it takes Barilla to ship the product is 10 calendar days from the date the order is received. This longer lead time causes distributors to place orders for more product than they actually need knowing that it takes 10 business days until they get their next one.
Sales Representatives – Similar to the transportation incentives where distributors would get discount pricing on truckloads, the sales reps would push more products out during promotional periods in order to up their compensation; however this would also lead to a lower need from distributors when outside the promo periods as they would be all stocked up. This would cause a wide variation in demand patterns for Barilla and would be almost impossible to forecast.

Environmental and Root Cause Analysis Pasta is a staple in Italy and very significant on the food chain. “Per capita pasta consumption in Italy averaged 18 kilos per year, greatly exceeding that of other western European countries.” (pg. 2, Barilla SpA case study) Because of this, the consumers who would be purchasing these products would be very cost conscience and would be prepared to go a little further out of their way to save a few dollars. Also, quality is very important, again, the customer is willing to travel a bit further in order to get the best possible tasting product for their money. Based on the fact that Barilla has the market share they do, we have to assume they would be recognized as an industry leader and favorite among their customers. This is why forecasting the consumer demand is extremely important and a necessity with JITD. Currently, the traditional way of filling orders is leading to stock out or excess inventory, neither is benefiting anyone in the supply chain. With Barilla’s current process to make pasta, they cannot change production at a moment’s notice. The plant needs to keep the kiln’s humidity and temperature at precise levels for the different types of pasta that it produces. Based on those requirements, sequential production would be hugely beneficial to Barilla to reduce downtime of production and keep their costs low throughout the process. Implementing the JITD system would give them the top down perspective for production/inventory decisions instead of the bottom to top reactionary system they have currently. In the case study, there is a mention that competition is a factor, with over 2000 Italian pasta manufacturers. This means that even though Barilla has a huge share in the area and has the name behind the product, they are constantly competing or have competitors going after more and more shelf space in supermarkets, small chain stores, and restaurants. Success in this market place is going to depend on pricing, brand recognition, brand availability, and customer service. Recognition is something they already have, however in the past, having control of their pricing and availability was out of their hands. By implementing a JITD system, they would have the ability to bring production back into their hands and gaining control over distribution. Adding to the fluctuating demand is also a strain on raw materials, meaning that if Barilla cannot get a handle on their own raw material needs, they may find it impossible to get what they need, when they need it. For example, if they continue to allow distribution to set production, they will have weeks where quantities needed can be 100, and then next week that can skyrocket to 1000, then back to 400 the week after. There is no forecasting model that can keep them on track for raw materials if this variation continues, especially in a perishable food market. If they decide to stock up on raw materials to prevent stock outs, they could find themselves with capital tied up in raw materials and no orders to sustain, or in inventory holding costs. If they can have their sales representatives working more closely with people placing the orders and the end users themselves, they would find they would have an “ear on the ground” that may assist them in forecasting more accurately as well.

Alternatives or Options
Alternative 1 – While a JITD system is relying on external numbers, this alternative would be solely based on internal operations. If Barilla is unable to convince distributors to share their information, they can move to a strategy that would set their forecasting to match their sales history. They would also need to incorporate a growth percentage as well as safety stock to ensure they are being as accurate as possible with production. For example, if they example a product that sold 1000 units in January of last year, they would have to assume the same production would be needed this year, along with a safety stock, say 10%, plus a year over year growth percentage, again 10%. They would also take in consideration if there are promotions on that product currently or upcoming shortly, another 10% for promotions. This would give them a monthly total of 1300 units, or 325 units per week. That should be their determined production schedule for January for that product to meet expected demand. By repeating this process month over month for the year and by product, they should be able to estimate a production run for the entire year by product. They can then task their sales teams to allocate monthly sales numbers to each distributor based on the production they know will be happening. This alternative will also allow them to better control costs on raw materials.
This alternative while it will reduce costs across the board in all areas, and will allow Barilla to control their own production schedule, does not come without its negatives. For example, with a preproduction schedule in place, it will be difficult to respond to change in the market trend, they will still have no POS data incorporated into this strategy, and market demand can still be higher or lower than their estimates for production.
Alternative 2 – The 2nd alternative is the most favorable option, it is the implementation of the Just-in-Time Distribution (JITD) system. As mentioned throughout this analysis, Barilla is facing a very challenging situation within their supply chain. The constant fluctuation in their weekly orders is causing the same fluctuation in their production line which is having a detrimental effect on their rising costs. Aligned with that, the distributors are placing weekly orders but suffering from stock outs (See attached Exhibit 13 from Barilla SpA Case Study) for several reasons outlined in this analysis including not using a forecasting model, sales representatives pushing product during promo period, long lead times, and others. To try and combat this originally, Barilla attempted to alleviate the issue by holding buffer inventory and asking distributors to do the same, however this actually would make things worse. The implementation of the JITD system should alleviate all issues for everyone in the supply chain, this will however take time and fine tuning to ensure it meets its full effectiveness and efficiency. The biggest hurdle will be getting through to the distributors/retailers and be able to show them how this system will not only save them by reducing on hand inventory, but will actually assist in getting faster deliveries and reduce stock outs greatly in the future. The main piece to this solution revolves around the sharing of information, mainly the sales and inventory on hand from distributors. This can be done in several ways, including some sort of POS system that automatically uploads the information to Barilla. By getting this information in real time, they should be able to reduce shipping time from an average of 10 days, to a fraction of this; a goal should be in the 48 hour range for time of receipt to the time of delivery of the product needed for replenishment. They will also be able to begin a 12 month sales history for every distributor to look for trends and be able to forecast future demand. Same as with the first alternative, Barilla will need to keep in mind safety stock, promotions and potential growth when determining their production runs. While this alternative has huge benefits to all involved in the supply chain, it does come with a few points that will need to be acknowledged. This will be a very time consuming project for Barilla, they will need to process tons of information coming in through some type of EDI system and they will need to turn it around quickly. They will have to look at hiring a team in place to process incoming information and get it back to production/shipping to reduce the turnaround time. It will also be costly to implement the system, not only to hire more employees, but the technological equipment needed, not only at Barilla but with all their distributors will be expensive. Time will be needed to train people to use the equipment efficiently, and errors will be made at the beginning. It’s important to realize that the whole JITD system will take several months to implement to its full capacity.

Recommendation and Implementation Based on the analysis of the case, looking at the history and facts, the recommendation is to implement alternative 2, the JITD system. This option, while having more upfront costs, is in the best interests of the entire supply chain in the long run. It carries the most benefits, potential for cost savings, and end user satisfaction. Barilla will be able to obtain real time sales information and have the ability to react to that in a way that will shorten time of delivery of product to their distributors. They will also streamline their own production facility by preplanning a production run and be able to plan ahead for raw materials. By implementing the JITD system, they will not only improve customer satisfaction by greatly reducing stock outs, but they should also see improved sales/margins while reducing inventory holding costs for themselves and their distributors. For implementation, Barilla will need to pick a date and then begin selling the idea to their supply chain. Time will be needed to purchase all POS equipment and implement and EDI system that will allow information to flow between distributors and Barilla. Barilla will need a team to implement this process in house, as well as act as customer service to all distributors who need to talk to someone live, or are having issues with their POS equipment. They will need to implement internal software that will not only determine what products were sold by the distributor and needs to be shipped out, but also once product leaves their own facility, a plan to replace their own stock levels to ensure product is always available to ship out. This will tie in to the raw material needs to keep stock at a min/max level to ensure they are meeting needs of all distributors.
Monitor and Control Implementing the JITD system will require time, understanding, and flexibility of all those within the supply chain. There will be bugs in the system all throughout the process that will need to be worked out until things begin to run smoothly. The entire process will need to be closely monitored for 6-8 months to ensure the data is flowing accurately and efficiently. Once they have everything running properly, they can begin to look for opportunities for improvement to streamline the process even further. Suggestion would be to take monthly reports on stock outs, inventory, and sales to see what improvements have been made over the previous time when the JITD system was not implemented to ensure the process is improving the situation. If the system is given a fair shot, Barilla and their distributors should see continuous improvement month over month, year over year in inventory planning, stock levels at the distributors, and an overall reduction in costs. A Process Improvement Associate should be implemented to ensure smooth transition of the new system and continuously look for opportunities to make the entire system easier and more efficient for everyone involved.