When many customers are considering a network modeling exercise to save money, they often ask this question first: “Where should my warehouses be, and how many do I need?” Like so many other things in life that are worth doing, it’s not as simple as that. Because if you don’t follow a process that analyzes the tradeoffs, you could end up increasing your costs instead of decreasing them.
Many of our network analysis customers expect costs to be reduced as they remove warehouses from their network. So when they see the optimized scenarios, they are surprised to find that as the number of warehouses decreases, transportation costs may actually increase. That’s because as the number of warehouse locations in the network decreases, the average distance to end customers may increase. As a result, it could cost more to transport your products to the customer.
At the same time, however, decreasing warehouse locations and increasing transportation costs isn’t the end of the story. Those increased transportation costs may still be offset by the need for less inventory in the network—a consequence of risk aggregation and lowered costs as you manage fewer locations.
The question is, do the increased transportation costs exceed the savings you expect to get by having fewer warehouses? This fits with Gartner’s report, A User Guide to Network Design and Inventory Optimization Solutions, which concludes: “The decision to optimize the network is typically decoupled from the inventory positioning decisions, even though the two decisions are clearly codependent. Tools to solve this problem are called network design or network optimization tools.”
There’s only one way to find out for sure: by conducting a network modeling exercise that starts at the beginning. A network modeling team can gather existing, reliable data to identify what is happening in the network now, and build a baseline model of the network. The team can talk with the customer about the cost buckets that have been found, and try to identify the source of unexplained variations between actual and modeled expenditures. There may be one-time exceptions at play, for instance. Once the model comes within approximately a 10% variation between actual and budgeted costs, there is an acceptable model of the current network.
Next, the network modeling team runs “what if” scenarios on the existing network to identify ways to optimize it. In many cases customers are surprised to learn that their biggest savings potential is not from changing their supply chain, but from setting up and following better procedures. Things like warehouse to customer alignment—and where and how much inventory to carry—can show the most opportunities for savings. For some customers, making structural changes to their supply chain shows only incremental savings, compared to improved processes for the existing structure.
As the scenarios show opportunities, a deeper dive may be required as the high amount of variable can cause a margin of error. Then, the company must make a small investment to reap the rewards, discovering their many potential options for optimizing the network and driving down costs. As consultant Noha Tohamy stated in, Bringing Supply Chain Modeling and Design to the Masses: “Modeling will not generate one clear answer for what to do. Instead, with various scenario and sensitivity analyses, companies can understand the range of possible strategies and evaluate their pros and cons, and can make realistic, value-based trade-offs.” And don’t forget: the opportunities to improve the inbound network are every bit as powerful to a company as an outbound exercise.
The focus here has been on warehousing and inventory. But the same principles may apply to where suppliers and manufacturing facilities should be. It’s the balance and tradeoffs throughout the entire supply chain that matter when you’re looking to increase efficiencies—and lower costs.
*To gain access to any Gartner reports, a login and password are required.