Reasons for the increase in variability in demand in Barilla’s supply chain The root cause of the increasing variability in Barilla’s supply chain is the demand fluctuation. In fact, Barilla is suffering from what is known as the bullwhip effect: demand fluctuations increase as one moves up the supply chain from retailer to manufacturer (Boute and Lambrecht, 2009). The large impact of the bullwhip effect at Barilla has a number of causes, which will be explored in this section.
In the paragraphs that follow, the following main underlying causes of the variability in the demand fluctuation at Barilla will be discussed: * Distributed inventories, local optimization * Lack of inventory information and sales forecast information * Promotions and quantity discounts * Lack of sophistication at the retailer and distributor level * Barilla’s high number of SKU’s * Barilla’s production process 1. 1Distributed inventories, local optimization
Regarding its dry products, which represent 75% of Barilla’s total sales, Barilla uses two central distribution centers (CDCs), one in the north and one in the south to channel 65% of its dry products. After storage at one of the CDCs, 90% of the dry products are distributed to grande distribuzione (GD) or distribuzione organizzata (DO), where GDs are part of Barilla’s and DOs are independent/external distributors that act as central buying organizations for a large number of independent supermarkets.
The remaining 10% of the dry products is distributed to Barilla-run depots (18 in total) which in turn distribute mainly to small grocery stores and partly to both chain supermarkets and independent supermarkets. All distribution centers, supermarkets and small grocery stores keep a safety stock to avoid stock outs. This increase in safety stock from retailer to distributor and from distributor to factory is a well-known effect of the previously mentioned bullwhip effect.
The isolated way in which Barilla’s distribution channel is currently organized implies that each partner in the channel is organizing its own information gathering and uses the information gathered only internally. 1. 2Lack of inventory information and sales forecast information In the current situation, Barilla clearly lacks vital inventory information as Barilla does not know what the safety stock levels and base stock levels of each individual warehouse are (from retailer to distributor). Moreover, Barilla does not have insight into the actual stock levels on a day to day basis.
Next to the lack of inventory information, Barilla currently has very limited sales forecast information as customers are unwilling to provide detailed sales data upon which Barilla could make delivery decisions and improve its demand forecasts. The lack of inventory and sales forecast information across the entire supply chain of Barilla increases the bullwhip effect. As very little data is observed, the estimates of the mean and standard deviation of customer demand are not regularly modified.
Since safety stock and base stock levels are dependent on the estimates of the mean and standard deviation of customer demand, sales orders have to be changed regularly which increases variability. 1. 3Promotions and quantity discounts Barilla deploys a sales strategy that relies on the use of trade promotions to push its products into its distribution network. On top of the promotional discounts, Barilla offers volume discounts providing incentives to order full truck loads. Finally, Barilla heavily uses advertising campaigns to stimulate end-customer demand, the effects of which are currently hard to measure.
The effects of the trade promotions and the quantity discounts are twofold. First, retailers tend to stock up inventory when the price of a certain Barilla product is low, a phenomenon also known as forward buying. Second, the quantity discounts further magnify the effect since buyers are pushed to order large quantities and/or full truck load in order to obtain additional discounts. Next to this, Barilla does not apply minimum or maximum order quantities, so there is no incentive for the buyers to limit their order quantity.
In case it turns out that too little of a certain product has been purchased, buyers can easily order small additional quantities to fulfill current needs since a minimum order quantity does not apply. 1. 4Lack of sophistication at the retailer and distributor level Most of the supermarket owners physically walk up and down the aisles to determine which products need replenishing and place orders accordingly at a daily basis with one of Barilla’s distributors. After that, it takes typically 1-2 days before the distributor has delivered the order at the supermarket.
At the level of the retailer, there is hardly any sophistication when it comes to forecasting orders and order quantities. Despite the fact that nearly all of Barilla’s distributors do have computerized ordering systems, very few have forecasting systems or sophisticated analytical tools in place to determine their ordering quantities. On top of that, most of Barilla’s distributors (both DOs and GDs) check their inventory weekly to place orders with Barilla after that. As result, the lead time from Barilla to the distributors is on average 10 days.
The lack of sophistication leads to inflated orders and heavily impacts the sales forecast information of Barilla itself (which is limited). 1. 5Barilla’s large number of SKU’s Barilla’s dry products represent about 75% of its total sales and have shelf lives of 2. 5 to 3 months (cookies) and 18 to 24 months (pasta). The dry products are offered in 800 different packaged SKU’s. A closer look at the dry pasta Barilla sells reveals that pasta is made in 200 different sizes and shapes and offered in more than 470 different packaged SKU’s.
The packages represent different quantities (kgs) and motifs that are country or region-specific. However, despite the fact that Barilla offers many different types of packages, most retailers carry a Barilla product in only one packaging option (or two at most). 1. 6Barilla’s production process Finally, the production process at Barilla plays an important role in the way Barilla’s supply chain currently functions. To keep changeover costs low and product quality high, Barilla follows a carefully chosen production sequence that minimizes the incremental changes in kiln temperature and humidity between pasta shapes.
Moreover, Barilla’s pasta plants are specialized by the type of pasta produced in the plant where especially the composition of the pasta plays an important role. 2. Solutions As a first step, Barilla and its supply chain partners should start gathering and analyzing demand information to forecast the expected orders of the retailers based on past time series data. With the combined statistical information of the fluctuations in demand on the one hand and the specific trade promotions (canvasses) on the other hand, Barilla and its supply chain partners will in time be able to forecast the effects of the trade promotions.
In order to enable precise demand forecasts, capturing the peaks and troughs associated with trade promotions should be modeled explicitly across time. As soon as the effects of the trade promotions are known, Barilla can anticipate on the effects and efficiently plan its production while avoiding batch ordering by distributors as well as reducing the risk of stock outs. 2. 1Promotions and quantity discounts A solution to deal with the promotion and discount problem would be to apply an ‘every-day low pricing’ strategy, offering a single consistent price instead of offering a regular price with periodic trade promotions.
By eliminating trade promotions, Barilla can eliminate many of the dramatic shifts in demand that currently occur due to the trade promotions. 2. 2Lack of sophistication at the retailer and distributor level Additionally, the sales representatives that spent 90 percent of their time at the store level could for instance use their portable computers to gather demand information. In order to increase the sophistication of the retailers and the distributors, the demand information should be shared with both the retailers and all the distributors.
This way, each stage of the supply chain can use actual customer demand data to create more accurate sales forecasts rather than relying on the orders received from the previous stage, which vary significantly more than the actual customer demand. 2. 3Large number of SKU’s The large number of SKU’s that Barilla currently offers fits into the differentiation strategy that Barilla is pursuing. However, Barilla should rethink its focus on differentiation as most of the end customers clearly notice and act accordingly when pasta is offered at a discount.
A product assortment optimization resulting in a reduction of SKU’s could reduce the logistical issues Barilla is facing and decrease the costs while at the same not hampering the total sales turnover too much. 2. 4Production process Boute and Lambrecht (2009) clearly indicate that when production is inflexible and significant costs are incurred by frequently switching production levels up and down, standard order-up-to policies are not desirable. Instead, because of the huge expenses, it may be important to avoid variance amplification (the bullwhip effect) or even to reduce variability of customer demand.
Given its production process, this is exactly what Barilla should be aiming for. 3. Proposed Strategy for decreasing bullwhip effect The new strategy of Barilla will be: “Instead of a logistic project, Just-In-Time-Delivery (JITD) is a cross function and cross companies change project aiming for win-win situation in integration and optimization of processes”. The goal for Barilla, while implementing this new strategy is to implement JITD within Barilla and at its customers. This strategy should be implemented in four steps: * Step 1 – Involve top management to realize changes within the company * Step 2 – Run a pilot with JITD at a GD Step 3 – Run a pilot with JITD at a DO * Step 4 – Implementing JITD across the supply chain, using best practices These steps will now be further detailed. After that, a number of other improvements will be suggested. 3. 1 Step 1 – Involve top management to realize changes within the company The starting point for Barilla to succeed in the JITD project is to gain the commitment of top management represented by the different functional areas in the organization (general management, logistics, sales, etc. ).
Next to involving top management, having a steering committee for reviewing milestones and decision-making is important, as well as ensuring the endorsement and commitment of top management during the whole implementation. Next to that, Barilla should gain ‘buy-in’ by showing joint interests for all stakeholders and involving them from the early stage onwards. Showing the joint benefits will increase stakeholders appeal to cooperate. For the sales personnel, benefits of the JITD project will be less hassle in ordering and in operation which will lead to happier customers.
For the marketing department, the JITD project will lead to a redirection of its resources to key areas, such as new product launches, building brand equity, customer segmentation and differentiation. From the financial perspective, Barilla will eventually reduce its operating working capital (OWC), allowing it to operate more efficiently and financially sustainable. The JITD project will should smoothen the production planning and scheduling in order to produce more efficiently.
The upside for the purchasing department is that fluctuations will smoothen and as a result negotiation power will be strengthened. Finally, top management will be contemplated with a higher EBIT, which will eventually impact their compensation and benefits. External stakeholders can also benefit from the JITD project – distributors and retailers will need to manage lower levels of stocks and fewer events of stock-outs will occur. By reducing stock levels, the storage costs will decrease and distributors and retailers will also be able to operate with lower levels of OWC.
As mentioned previously, higher efficiency and more focus on the customer will lead to more customer satisfaction and therefore retention. 3. 2 Step 2 – Run a pilot with JITD at a GD Once all the stakeholders are informed and onboard, the first thing that Barilla should do is to implement the Vendor Managed Inventory (VMI) method at one of its own GDs. This will help Barilla to build up a successful example to present as a winner to prove the added value of the program. It is important to create a sense of urgency within Barilla from the start and to present the current issues.
Because Barilla owns the GDs and top management is committed to this strategy, there are no obstacles to do so. Benefits of this strategy are: 1) Barilla will collect a lot of data which can be used to create forecasts; and 2) the pilot and its lessons learned can be used as an example towards the DOs (independent/external distributors) and the supermarkets. The VMI method should be implemented step by step. So first one GD, and after a while when the first GD shows decreasing inventory levels (without stock-outs), the next GD can implement VMI.
As result, Barilla will have a large database which can be used for to create a solid forecasting model. With these results, Barilla can approach the DOs and the supermarkets to convince them that JITD/VMI lowers the inventory which results in lower costs. 3. 3 Step 3 – Run a pilot with JITD at a DO The next step is to convince a DO to join the pilot. This could be achieved based on the positive outcome of the internal pilot, a real case to convince that JITD and/or VMI creates value. This should lead to cooperation of the supply chain partners to and sharing of their data.
It will be essential for Barilla to prove the potential savings of the JITD program to its external distributors, as to make the JITD endeavor more credible and to facilitate information sharing. Alternatively, a 3rd party who has the trust of Barilla and the DO could be used to facilitate the cooperation. The information sharing could be done through Electronic Data Interchange (EDI) or via the Internet. Alternatively, an independent 3rd party could be used for the information gathering. Building up the necessary database will enable Barilla to forecast the expected demand and adjust its production to this demand.
Barilla will need an adequate IT infrastructure to operate JITD successfully. By installing the right ICT system and having adequate master data management in place Barilla can align with customer on SKU coding, forecast time, frequency, MoU, etc. EDI will be a useful tool in communicating information promptly across the supply chain channels and enable responsive reactions in the event of fluctuations in the chain. The sales representatives need to be equipped with the proper ICT tools (e. g. portable notebook, iPad, etc. ) to be connected to the information flow.
Also this pilot (at the DO) will be used to be presented as a winner and a sign of gained efficiency and added value. 3. 4 Step 4 – Implementing JITD across the supply chain, using best practices Create creditability by building successful stories, such as starting the experiment within Barilla’s depots. Barilla will select more distributors to implement JITD plan and thereafter it will expand to the other distributors once success can be demonstrated and trust built. There will be a champion to function as ambassador of the JITD.
Trust will be built by having an open and regular dialogue across the supply chain, with transparency in the information sharing. 3. 5 Changes to the sales organization As discussed previously, Barilla should adopt an “every-day low pricing” strategy to reduce variability within the supply chain. To do so the sales organization of Barilla must be convinced that this is a good idea. After all, Barilla’s customers are accustomed to promotions and discounts and they expect them to be offered to them periodically.
Barilla needs to make use of its strong marketing position to spread the new strategy that will replace the promotion-based strategy and to retain customer satisfaction. Barilla’s new strategy of “every-day low pricing” should be implemented together with other targets or KPI’s for the sales force. In order to optimize its product portfolio, Barilla needs to rationalize the tail SKUs with low margin and allocate marketing resources to value-added activities. Training for sales representatives on forecasting will equip them with the right competences to adhere to the program in a successful manner.
Also, Barilla will provide planning trainings to educate the supply chain partners, including customer, on how to run the business more efficiently and smoothly. An incentive system which encompasses the right target and KPI’s is vital to align interests across the company. For instance, Barilla can use sales turnover, profit margin (EBIT, EBITDA), OWC, sales forecast accuracy, supply reliability, etc. Finally, the HR department needs to safeguard alignment by implementing an adequate recruitment plan that can respond to fluctuations in the market force, such as high turnover of sales representatives. . 6 Improvements at the retailers and distributor level As discussed in chapter 1, there is almost no automated process in replenishing the supermarkets. Most of the ordering is done by means physically counting the actual inventory. This problem could be solved by using barcode scanners which are directly attached to the ordering system of Barilla, or to the cash registers at the supermarkets. This way, the inventory will automatically be monitored and the system can determine by itself if the safety stock level is reached and a new order must be placed.
By using this method, less orders will have to be placed that fluctuate less in quantity. This will help the distributors in forecasting their inventory better and as a result, the costs will be lower and lead times shorter for all parties involved. 3. 7 Upgrading Barilla’s production process To avoid production changeovers, Barilla could think of making use of the push-pull boundary. Because dry products have a quite long perishable date they could think of making bulk products, store them as bulk and on the basis of demand they can package on the demanded SKU.
The warehouse will have its floor redesigned to facilitate cross docking. Ultimately, if Barilla wants to reduce its packaging and SKU’s, the production department needs to team up with marketing to redesign packaging to optimize transport, handling and optimize storage costs. 4. Final considerations Implementing JITD will tackle the uneven distribution workload Barilla currently encounters with its distributors. In order to achieve this, better demand forecasts are required that can only be obtained with the support of the distributors.
The use of JITD will enable Barilla to: 1) more effectively meet the end consumer’s needs by distributing the workload on manufacturing and logistics systems more evenly; 2) operate more efficiently and; 3) be more customer-focused. Communication and transparency here is the critical success factor. The supplier (Barilla) has to collect information from the distributor, and the distributor has to be able to collect information from the wholesaler, and then the wholesaler has to pass it on. IT infrastructure and cooperation is vital to enable the program to succeed.
