Supply chain network optimization is the use of mathematical models to decide where facilities should be, which products flow through them, which customers each serves and which transport modes carry the goods. The goal is to balance total cost against service level under real constraints such as capacity, lead times and policy.
It is often called network design. It informs infrequent, high-impact decisions, for example opening a distribution centre, closing a plant or reassigning customers to a different warehouse.
Decisions it supports
- How many warehouses or plants the network needs and where they should be.
- Which sites serve which customer regions.
- What to produce where, and what to hold in stock at each location.
- Which transport modes and carriers to use on each lane.
- Whether to own, lease or outsource facilities.
- How the network behaves when demand grows, shifts or a node is lost.
Inputs a model needs
| Input | Examples |
|---|---|
| Demand | Volume by product, customer or region, and expected growth |
| Facilities | Existing and candidate sites, capacities, fixed and variable operating costs |
| Transportation | Lanes, modes, rates, transit times, minimum shipment sizes |
| Service rules | Delivery time targets, order cut-offs, customer commitments |
| Supply | Supplier locations, material costs, production constraints |
| Policy and trade | Tariffs, duties, border crossing time, sourcing restrictions |
The process, step by step
- Define the question and the scope, such as one region or the entire network.
- Collect and clean the data, and build a baseline model of the current network.
- Check that the baseline reproduces known costs and flows closely enough to be trusted.
- Define scenarios: new sites, demand shifts, changes in carrier rates, disruption cases.
- Run the optimization and compare results on cost, service and risk.
- Test sensitivity. Which assumptions, if wrong, would change the recommendation?
- Present options to decision makers with the trade-offs, then plan the transition.
Tools and methods
Most models use mixed-integer or linear programming to find the lowest-cost arrangement that meets constraints. Simulation is added to test variability. Specialist network design software exists, and many companies also use spreadsheets for small cases. For tool selection, check how the product handles data import, scenario management, solver transparency and the size of network it can model.
AI features, such as automated data preparation or faster scenario generation, can shorten setup work. The core decisions still come from an optimization model and from human judgment about strategy and risk.
Cost and service trade-offs
Adding warehouses generally shortens delivery distance and improves service but raises fixed cost and total inventory. Consolidating sites lowers fixed cost and may reduce stock needs, but lengthens transit times and increases dependence on fewer locations. Good analysis shows these curves, rather than a single answer.
Common pitfalls
- Using the lowest cost as the only goal and ignoring resilience.
- Relying on one forecast instead of testing several demand cases.
- Dirty or incomplete data, especially freight rates and lead times.
- Leaving out practical costs such as moving inventory, lease exit terms and workforce changes.
- Treating the output as final. Models simplify reality and need review by operations.
- Designing once and not revisiting after demand, products or trade rules change.
Common questions
How often should a network be reviewed?
Many companies run a full study every few years or after major changes such as acquisitions, new markets or large shifts in demand. Smaller adjustments can be tested more often.
Is network optimization the same as route optimization?
No. Network optimization is a strategic or tactical decision about facilities and flows. Route optimization sequences stops for individual vehicles day to day.
Sources and further reading
- MIT Center for Transportation and Logistics (MIT CTL)
- Association for Supply Chain Management (ASCM)
- Harvard Business Review: Supply Chain Management topic (Harvard Business Review)