Most of the network studies St. Onge Company performs are what I like to refer to as “event-driven” studies. These are typically classified into short-term events and long-term events, and some of these events might dictate a deeper dive into your network growth strategy. Short-term events can include the following:
Long-term events can include the following:
Now that network growth drivers have been identified, you are ready to think about your network analysis’s design horizon. Short-term studies typically look at a zero to one-year design horizon. These studies are geared to either look for quick wins to reduce costs and improve service within the existing supply chain network, or to help inform network decisions that need to be made in the next twelve months. A longer-term study typically evaluates the network with a three- to five-year design horizon, which usually aligns with the corporate customer demand growth plan already in place.
A design horizon of three to five years is the most common assumption for a supply chain network analysis, as this is usually the farthest that companies are comfortable projecting their business. This is typically fine for modest growth (3% to 5% annually). When growth exceeds 8% annually, you certainly need to consider extending your design horizon, or at least doing sensitivity analysis beyond five years.
Consider the following example. A high-growth company is doing nationwide customer distribution from a single facility in Dallas, TX, and they have outgrown the current facility. They want to investigate two options for their facility growth plan. The first is to keep Dallas and add a new distribution facility in a new market. This would allow them to maintain the current building and workforce by pushing volume to the new facility. Dallas is a decent location for a single distribution center(DC) network, but it is not as good in a two-DC network, as the second facility will more than likely be added in the northeast, which leaves the west coast underserviced. The second option is to move out of Dallas completely and open two new locations (maybe Louisville, KY, and Southern CA). This is an excellent two-DC network to provide great service to both coasts while controlling costs. Makes sense for a five-year plan, right? Fast-forward to years six, seven, and eight. Remember, this is a high-growth company, and they have, again, run out of space and are looking to add a third facility. The west coast facility in Southern CA is adequate, and it will continue to handle the twenty to twenty-five percent of U.S. volume that is typical for a facility in that location.
The problem is the eighty percent of demand that is east of the Rockies. While Louisville is great in a two-DC network, it is not as great as part of a three-DC network. In year seven, the optimal network might have been Central PA, Dallas, and Southern CA. Wait, didn’t they just move out of Dallas? Yes! If they had just looked out a little further for the design horizon, they might have seen the best option was to keep Dallas and build on that network. They may have been sub-optimal for a couple of years, but they would have avoided the mistake of opening Louisville. While this is just an example, it illustrates the importance of making sure you do not sacrifice the long-term for the short-term in your facility planning process.
—John McDermott, St. Onge Company