Nestle DSD Makes Optimal Production Sourcing Decisions

Nestlé USA has a Direct Store Delivery (DSD) division focused on several of its frozen food brands, including Edy’s and Dreyer’s (ice cream) and Digiorno, Tombstone, and Jack’s (frozen pizza). The company’s move to DSD began in 2006 when it acquired Dreyer’s Grand Ice Cream and the scope of its operations has grown since acquiring the pizza brands.

DSD can be a competitive route to market. Food companies often pay to have their products placed on retailer shelves. But if a food company has DSD capabilities supported by an efficient supply chain organization, it doesn’t have to “pay to play” (or at least not to the same extent). This is particularly true for smaller stores and chains.

Brad Alford, Chairman & CEO of Nestle USA, gave a presentation last September where he provided some facts about Nestle’s DSD operations. For its ice cream brands, Nestle pre-sells and pre-picks items before delivering them to freezers at the back of the store. Later, merchandisers arrive at the store, pull the product out of the freezer, and bring it to the front of the store. Merchandisers visit the stores multiple times a week. This business generates $2.2 billion in annual revenue and is supported by 2,200 full-time employees.

In contrast, pizza DSD is based on a “rolling inventory model,” which I assume means the drivers arrive at the stores, check the inventory, and then deliver what is necessary. This business generates $1.9 billion in annual revenue and is supported by 1,500 full-time employees.

Nestle has the fourth largest, company-owned DSD system in the US food industry and it is by far the largest in frozen foods. To support this business, the company has 30 distribution centers and about 200 cross-docks. This is a large number of warehouses, but in DSD you need warehouses close to the stores.

At JDA’s user conference last month, two Nestle DSD supply chain managers gave a presentation on how the company is using a strategic network design solution from JDA—Production & Sourcing Optimization—on a quarterly basis to determine optimal sourcing by lane. In the ice cream DSD supply chain, transportation and frozen storage costs are high. Consequently, Nestle has redundant plants and production lines across the US. The network is made more complex by the fact that demand for ice cream is seasonal (greater demand during the warmer months). Thus, the decision about how far ahead to forward load production and where production should be done is complex.

The JDA solution answers the question, “Where should we make a product based on storage, handling, manufacturing, transportation, service, and inventory carrying costs?” Service costs are based on a penalty logic that is parameterized into the solution. The thinking is that retailers don’t like it when their shelves are empty, and some retailers even impose fines. The solution can also look at the profitability by case of the different SKUs. In instances where demand is outstripping supply, the solution can determine which cases should be cut.

Running these plans quarterly is a bit unusual; strategic planning is typically done on an annual basis. In Nestle’s case, however, demand changes are significant enough that doing it quarterly makes sense.

In Nestle’s first year using this solution, the company saved $5.5 million!

To get there, the company had to move to centralized production planning and it had to create a centralized strategic sourcing group to manage these long range plans (the group contains some math-savvy planners – i.e., industrial engineers rather than just business undergrads). The company also had to clearly explain to factory managers, at its annual planning meeting, the reasons behind its sourcing decisions.

You can’t argue with success. Over five million dollars in annual savings—now that’s how I spell “ROI”!

(Note: JDA Software is an ARC client)