Barilla SpA – Harvard Business Case Analysis

* Mature product by 1990. Little or no growth domestically (Italy)

* Export market experiencing large growth (20-25% per yr in European countries)

Expect 2/3 of new demand from Eastern European market. Demand is high because they wanted low-priced basic food products

* Limited or no seasonal demand

* Highly competitive domestically over 2000 pasta manufacturers in Italy. Declining margins.

* Pasta market is extremely price sensitive.

* Barilla is the market leader in Italy and is the industry’s largest manufacturer.

* Supply chain represents approximately 55-58 calendar DOH of dry goods (30 in the CDC, 14 at the Distributor, 11-14 at the Supermarket or Independent Shops)

Competitive Strategy

* Barilla is the largest pasta manufacturer in the world (35% of market in Italy and 22% of market in Europe)

* Also 29% of the Italian bakery product market

* Vertically integrated – owned mills, pasta/bread plants, and distribution network

* High quality product

* Largest and most advanced manufacturing facility (Pedrignano plant)

Focus is on developing most efficient production methods for existing and new products.

* Will increase focus on export opportunities.

* Creative marketing campaigns

Large scale advertising programs promotions. Also, selling in cardboard box with a recognizable color pattern created strong brand name recognition and image for their pasta.

Enjoyed a strong brand image in Italy

* Product mix

Fresh products – 21 day shelf life

Dry products – long shelf life 18-24 months, medium shelf life 10-12 weeks (75% of Barilla’s sales)

800 dry SKU’s; 470 pasta SKU’s from 200 pasta shapes.

Pasta SKU differentiation by packaging.

* Plans to exploit new market opportunities in Eastern Europe.

* Focus on high volume manufacturing.

Distribution Network

Barilla Manufacturing Plants

To

Barilla CDC or Barilla depots

To

Distributors (GD’s or DO’s)

To

Supermarkets, Small Independent Shops

1. What do you think are the main causes for large fluctuations in orders observed at the Pedrignano CDC?

We believe the primary reason for the huge fluctuations in demand is because of Barilla’s trade promotions and the frequency in which they occur (10 – 12 Canvass periods per year, each 4-5 weeks). These trade promotions create incentive conflicts by causing customers to purchase in quantities that do not reflect their immediate needs (forward buying) but instead focus their stocking decisions on acquisition price. They buy in bigger quantities and ‘stock-up’ which takes them out of their natural demand based ordering patterns. Normal reordering does not happen until the product has either been depleted to the normal reorder point or another pricing incentive has been offered. This is especially true for products with long shelf lives (i.e. Dry Products). Therefore, at the end of each canvass period orders are artificially inflated and total supply chain inventories are out of balance.

Other factors that cause fluctuations in orders are:

* The combination of weekly orders and lead time required when placing an order. When the distributor only orders once a week, order demand variability is increased. Also, the average lead time of 10 days increases forecasting inaccuracy, and requires the distributors hold more safety stock. The longer the lead time, the larger the safety stock, the larger the fluctuation in orders.

* The sales representative’s incentives were based upon achieving sales targets for each canvass period. This could cause a conflict of interest for the sales reps where they push product at the end of the canvass period to make their numbers. This would create a large fluctuation at the end of each canvass period.

* The huge number of SKU’s that each distributor maintains could cause an increase in the demand uncertainty for a particular Barilla product.

* Customers respond to other companies’ trade promotions and this impacts their Barilla order quantities.

* Regional preferences add to the fluctuation in demand.

* Lastly, most distributors do not have forecasting systems to help determine order quantities.

2. What do you think of the JITD Program? What actions should Barilla take to reduce the fluctuations in demand? For what kind of products would such a program be best suited?

We believe that the JITD program is a good one which could help Barilla, the Distributors, the supermarkets, and ultimately the consumer. The primary focus of JITD is information sharing upstream and downstream in the supply chain. Knowing the distributors sales and shipping data could help Barilla and the distributors in many ways:

* It would smooth the manufacturing cycle. Barilla would know the demand for each product as it is sold / shipped.

* Use the sales information from the distributors to create replenishment orders on an as needed basis. Thus, it would reduce inventory levels at both Barilla and the distributors. Also, shortening the time between orders reduces demand variability.

* Barilla would also know when stock items are low and generate replenishment orders reducing stock-outs, which lead to increased sales.

We believe Barilla’s high volume dry goods products would be best suited for JITD.

a. The JITD program would help reduce demand variation for one of the two order streams; Barilla Distributor warehouse. The variation reduction will help stabilize ad hoc scheduling changes and should help improve Distributor Market fill rates.

b. Barilla should work with Distributors and Markets on an event basis given the promotion-centric environment. Baseline forecasts could be maintained statistically while events could be updated manually based on the most current market conditions and promotions.

c. Promotions should be based on a broader range of factors that include inventory throughout the pipeline and competitive promotions.

d. Barilla should focus these efforts on dry goods because they represent the biggest problems with long lead-times and process related demand fluctuations.

e. Barilla could also consider delaying manufacturing scheduling until each part was ready for production. For example, Barilla could determine the optimal changeover sequence for their dry goods and then calculate the production quantity right before the part was due for production. The delay would lead to a more accurate replenishment quantity since the most recent information would be used.

3. Do you anticipate any problems if the JITD program is implemented?

Yes, besides the normal problems associated with any change, you may still have distributors that do not want to share information. It will be difficult convincing distributors that sharing their internal sales/shipping information and becoming a passive partner in the supply chain will benefit them in the long run. Even if most distributors agreed to share sales/shipping data, a single large distributor still could create an order that would interrupt Barilla’s manufacturing and logistics operation. Also, Barilla would be required to limit trade promotions for JITD to be successful. They would need to renegotiate contracts with the distributors agreeing on an everyday low price (ELP). The problem here is trade promotions are ingrained in the industry and distributors may not accept the change.

Other problems we could see happening:

* Sales representatives would have issue with the change. Their incentives are based upon meeting targets for each canvass period. These incentives would need to change to be aligned with JITD. No longer will sales rep’s incentives be based upon sales of a particular product during a canvass period, no more pushing product and sacrificing future demand. This dramatic difference would likely lead to serious resistance in the field and a long implementation, thus delaying the potential benefits of JITD.

* Barilla and distributors may need to invest in IT to support the EDI transactions needed to share information.

* Distribution costs may increase because of smaller more frequent shipments. We would anticipate these costs would be counteracted by the reduced inventory costs and increased sales via the reduction of stock-outs.

* Manufacturing flexibility is another key exposure as schedules would be created over a shorter ordering lead-time. Some of the cost of this exposure is already built into current operations in the form of last minute schedule changes.

4. As one of Barilla’s customers, what would your response to JITD be? Why?

We would invest time working with Barilla to figure out the specific impacts on people, processes, and systems. This big of a shift in managing the supply chain would require many changes that could create new problems with the removal of other problems. Examples include the receiving and put-away processes with more frequent deliveries, and how to manage promotions not originating from Barilla. Additionally, we would want to understand how the price promotions would work given the removal of buy-in incentives. We would anticipate volume discounts over longer horizons and possible rebate programs because it would be difficult to rationalize a decrease in purchase price offset by only an inventory reduction.

Especially as a DO, we would be skeptical because we would have to now treat Barilla differently from our other suppliers in both how we order product and how we control price. This loss of economies of scale would have to be offset.

5. Critique Magialli’s plan.

Magialli’s plan capitalizes on an opportunity to change how the heavily competitive pasta industry manages its supply chain activities. From this vantage, Barilla could establish a first mover position in developing ways to increase responsiveness to competitive moves and recognizing and rectifying supply and demand imbalances.

In spite of the potential to develop first mover advantages, the plan suffers from five significant drawbacks. First, the plan is incomplete in how it addresses the price promotion environment. Specific ideas need to be communicated that reinforce the need to continue promotions but change how they are derived and managed. Sales Rep buy-in should be the easiest to address since it is internal to Barilla but the Sales Reps are campaigning against it outside the company. Improving service and reducing inventory should have no bearing on how Sales Reps are compensated so Barilla should direct the Reps to develop and sell JITD instead of spending too much time trying to convince them. Second, the plan is incomplete in how it addresses expectations on the manufacturing operations which are constrained by high changeover costs. Manufacturing will have to eventually focus on finding ways to continuously improve changeover times but no quick fix will be available if increased responsiveness is required immediately.

Barilla will have to consider a pilot or phase approach to create enough opportunity to evaluate the effectiveness of each change. Third, the plan needs a starting point but lacks one. The ideas are intuitive but have to be proven in a representative setting that is non-threatening to everyone. The GD’s seem to be good candidates for a pilot project that could address each of the people, process, and system issues. It seems that organizational levels too low in the organization are being allowed to detract from the potential of a tightly managed pilot program. Fourth, Magialli does not refer to the financial impact of reducing inventory in the supply chain which will have a direct impact to the profit and loss statements of Barilla. There needs to be an agreement from management that a reduction in supply chain inventory will impact the P&L unfavorably. Finally, the plan ignores how to handle returns if Distributors become more passive in their ordering but must still work with stores on managing excess inventory. Planning for a returns process can significantly help avoid costly processing in the future.

Compensating for the drawbacks are three key areas of improvement. First, the plan could help determine which price promotions to run based on current inventory throughout the supply chain combined with the most recent competitor situation. The current promotions might be predictable and could be incenting unprofitable behaviors with both the distributors and the sales force. Switching to a variable model both optimizes the supply chain and helps to build anticipation. Second, the plan allows Manufacturing to find ways to improve changeovers with fewer distractions. Continuous improvement participation is likely hampered in the current environment given so many schedule changes and special requests. Third, the plan can be put into place in any geographic environment, not just in a couple of segments in Italy. This fact should give Barilla the opportunity to pilot the key concepts in a less hostile environment.