Supply chain network design is a project undertaken to optimize the entire supply chain system (from supply or manufacturers through warehouses or distribution centers and finally to wholesalers, retailers, stores and ecommerce customers.) Sometimes you will hear this process called network modeling, as a mathematical model is required to be built in order to perform this type of analysis. The output of network design is typically the optimal number and location(s) of facilities as well as the flows of product into and out of those facilities. Many people ask what the network design project and process look like. To help answer some of these questions, I want to give an overview of each of the 5 milestones in a supply chain network design project.
At the onset of a network optimization project, a large amount of historical data is collected. This data is summarized and reviewed with the project team in the first milestone meeting – Data Review and Validation. The summaries shown at this meeting describe the existing customer geography, DC size and capacity, inventory history, transportation modes and costs, and much more. The goal of this first milestone is to ensure that the data collected accurately represents the business (specifically focused on the supply chain activities) during the data collection time period (typically 12 months).
The second milestone in the network optimization project is the baseline model review. The data from the data review/validation milestone is taken and loaded into the network optimization software. This model replicates the existing distribution network flows and associated costs as they occurred during the 12-month data time period. Additionally, this model serves to validate the input data and modeling methodology and ensures that the model costs calculated are calibrated to actual company financials. The second milestone is used to review the modeling assumptions used to build the baseline model and review the associated output.
After the baseline model is constructed and calibrated, the model is carried forward to a design year (typical 3-5 years in the future but sometimes as long as 10 years). This model incorporates not only business growth, but any significant changes that occurred during the data collection time period or any planned future network changes (openings/closings, sourcing changes, etc.). No new facilities are added into the model at this time. In summary, this model optimizes the network under future state conditions using current facilities. This model will serve as a “benchmark” model against which all other alternatives can be compared.
Now to the exciting part. In the scenario review milestone, various scenarios are tested based on the objective of the study. In these scenarios, new facilities may be considered. These scenarios are compared back to the benchmark in both cost and service. This process is iterative and after each round of scenarios the team will evaluate and develop additional scenarios to test based on the learnings of the scenarios already ran. After several rounds of scenarios (and sensitivity testing), typically 2-3 leading scenarios are identified.
In the last milestone meeting, the leading scenarios are summarized with pros and cons, quantitative results, and qualitative considerations. Recommendations are be based not only on the quantitative model results, but on various qualitative considerations such as service, ease of implementation, investment cost, flexibility, business friendly locations, and risk management. The leading scenario is then recommended and a high-level transition plan is developed. A final presentation is reviewed which summarizes the input data and assumptions, describes the various scenarios tested, presents the model results, and explains the associated evaluation and resulting recommendations.
In summary, there are typically 5 milestones in a network optimization project. These milestones are held jointly with the core project team. This ensures the entire team is aligned prior to moving to the next milestone and helps the team stay engaged. Each milestone is important to complete before moving onto the subsequent step in the process to ensure a successful and achievable network optimization is defined.
—Brad Barry, St. Onge Company
