Strategic Planning: How to Design an Optimal Distribution Network

Want More?

Download the PDF

Want to stay up to date on the trends and issues impacting your supply chain?

I understand that Tompkins will only use this information to contact me about business opportunities. By completing this form I am confirming that I have read and accept the Privacy Policy.

This field is for validation purposes and should be left unchanged.

Contact Us

We would love to hear from you.
Call Us: 561-994-0012

Want to stay up to date on the trends and issues impacting your supply chain?

I understand that Tompkins will only use this information to contact me about business opportunities. By completing this form I am confirming that I have read and accept the Privacy Policy.

Published June 29, 2021

Share on:

Distribution network planning is one of the main areas in which strategic planning is applied. A company will develop a strategic distribution network plan to implement business objectives over a given planning horizon. A good plan will determine the optimal network to provide the customer with the right goods in the right quantity at the right place and the right time to minimize the total distribution cost. As the number of warehouses increases, delivery costs decrease but warehouse costs increase. This is shown in a simplified manner in Figure 6.1. The opposite is also true-as the number of warehouses decreases, delivery costs increase but warehouse costs decrease. To minimize total distribution costs, it is important to find the best balance between warehouse and transportation costs.

Figure 6.1: simplified distribution cost model
Figure 6.1: Simplified distribution cost model

The objective of strategic distribution network planning is to determine the most economical way to ship and receive products while maintaining or increasing customer satisfaction requirements; simply put, a plan to maximize profits and optimize service. Strategic distribution network planning typically answers the following nine questions:

  1. How many distribution centers should exist?
  2. Where should they be?
  3. How much inventory should be stocked at each center?
  4. Which customers should be serviced by each center?
  5. How should the customers order from the center?
  6. How should distribution centers order from vendors?
  7. How frequently should shipments be made to each customer?
  8. What should the service levels be?
  9. Which transportation methods should be used?

Factors in Strategic Distribution Network Planning

Distribution networks can range from direct shipments from the source to demand points for job-shop items to complex multisite networks. The design of a distribution network depends on factors such as type, range and volume of products; the service area’s geographic spread; level of service required; and number and type of customers. Furthermore, since distribution is a dynamic environment, it is also affected by:

  • Geographic shifts in production and consumption (population shifts)
  • Market segmentation, new markets and new customer service requirements
  • Cost increases in energy, plant and equipment maintenance and labor
  • Government regulation or deregulation
  • Fuel and other accessorial charges
  • Negotiated rate contracts
  • Service requirements
  • Product proliferation and life cycle
  • Competition
  • Economy

Internal organizational areas such as marketing, production and finance also affect the makeup of a distribution network. To understand how, it is important to understand their goals. Marketing seeks to maximize sales by being accessible to the customer and having plenty of inventory to minimize back orders. Production wants to minimize cost by running long lots and pushing the product out the door, so both marketing and production request more inventory and more locations to hold it. Finance wants to minimize cost and conserve cash and credit, thereby lowering inventory and the number of stocking points. With these conflicting needs, the distribution network planner must find the lowest cost distribution network and inventory management technique that satisfies both the customer and the company’s objectives.

The Planning Process

Planning a distribution network is a sequential process that continually needs updating, and includes the seven steps outlined below. The pitfall some companies run into is performing steps 3 through 7 before accomplishing-and understanding-steps 1 and 2, which are the most important. The planning process is only as good as the data put into its analysis, and it should:

  1. Document the distribution network
  2. Identify the delivery requirements
  3. Determine the network operating requirements
  4. Develop alternative networks
  5. Model annual operating costs of the existing network and the alternatives
  6. Evaluate alternatives
  7. Specify the plan

Documenting the Distribution Network

The first three steps can be simultaneous-their main goal is to understand the current system and define the requirements of the future system. To document the existing systems, the planner must collect information on the distribution centers and transportation system. It is critical to collect information from all sites being considered because the study could result in recommendations for closing, moving or expanding them. The following information needs to be collected for each site:

  • Space utilization. Determine the use of the distribution center. This will allow you to determine the amount of inventory space that will be required if this facility is to be closed when the analysis is complete. It also identifies how much more inventory could be consolidated to this location.
  • Layout and equipment. List the equipment and layout of each facility. If you have a list available during the planning stage, it will be easier to determine the investment requirements of a new or expanded facility.
  • Warehouse operating procedures. Understand the order picking and shipping procedures. If there are two product lines in one location, are they picked and shipped together? Understand the differences in operating methods between facilities. This may tell you why one facility achieves a higher throughput efficiency per person. Understand how replenishment orders are pulled or pushed to the distribution center.
  • Staffing levels. Document levels by position. Understand which jobs could be consolidated. Collect labor rates by level, including fringe benefits.
  • Receiving and shipping volumes. Understand the number of incoming and outgoing trucks and the number of docks. This will be important if the facility is required to increase throughput.
  • Building characteristics. Collect square footage, clear height, column spacing, lighting levels, etc. for the same reason as layout information, and remember to review expansion capabilities.
  • Access to location. Review the access to main highways. Determine if this affects freight cost.
  • Annual operating cost. Collect lease, taxes, insurance, maintenance, energy and other facility costs.
  • Inventory. Collect information on inventory turns and levels, fill rates, safety stock levels and ABC analysis. With this information, you can determine the savings that can be gained by consolidating facilities. Also collect information on which and how much stock is slow moving or seasonal to help determine if it should be centralized in one location or stored in public warehouse space. Get future inventory goals.
  • Performance reporting. Understand the performance measures for service requirements, order completeness and shipping accuracy.

The following information should be collected for the transportation system:

  • Freight classes and discounts. Collect the freight classes and rates used. In addition to freight classes, determine the discounts by carrier or location. It is also important to understand where, and under which parameters, the discounts apply (e.g., routes, minimum weights).
  • Transportation operating procedures. Understand how a certain mode of transportation and a carrier is selected.
  • Delivery requirements. What are the delivery requirements (days of delivery) to the customer, and how is carrier performance measured? Is order completeness measured?
  • Replenishment weight/cube. At what weight is a trailer cubed out? Get this information from each replenishment point and for a typical load of general merchandise.

At the end of the data collection, which would benefit from site visits, a project team meeting should be held to summarize the data collected and assess each site. This assessment will give the team insight into its operation. In addition, it will discover information unknown to management that will be useful in developing alternatives.

Documenting the future distribution network requirements requires an understanding of marketing strategies and sales forecasts. Here are some questions that marketing and sales should answer:

  • Are any new products coming out? Where are they sourced from? What is the target market (geographical) area?
  • What are the current ordering parameters? For example, what is the minimum ordering size? Are any terms of order changing (e.g., charging for expedited service)?
  • What is the direction of the market (packaging changes, wholesalers, mass merchants having more volume)?
  • What are the sales increases by year?
  • Are there noticeable customer shifts? Are fewer customers handling more volume?
  • Are geographic shifts emerging? Have sales increased by geographic region?

Identifying Delivery Requirements

One of the key elements to analyzing a distribution network is delivery requirements, or the time from order placement to receipt of the shipment. If these requirements are not identifiable, it is important to conduct a customer service gap analysis. The gap analysis is a series of questions directed at internal staff and customers. It aims to identify discrepancies between the level of service that will satisfy the customer and current performance. The gap analysis attempts to determine the customer’s balance between service and cost. In general, is it more important to have the goods faster or at a lower price?

Figure 6.2 shows distribution cost as it relates to days of delivery, and its effect on profits. The figure indicates that as delivery requirements are reduced, there is a point where the cost and distribution of the product will outweigh the income achieved from sales. The figure also indicates that the longer the delivery time, the higher the profits. At some point, sales sharply decline because competition exceeds both your delivery time and cost (assuming equal product quality). The key is to determine the best service that also maximizes profits.

Figure 6.2: Gap analysis
Figure 6.2: Gap analysis seeks to determine the optimum combination of production and distribution costs that will deliver maximum profit from the price the customer is willing to pay.

Network Operating Requirements

Once customer service requirements are understood, the next step is to determine network operating requirements. This involves determining the baseline cost and the service and performance characteristics of the current network. Key elements to identify include:

  • Current facility locations, capacity, throughput, cost, performance, flexibility, effectiveness and efficiency
  • Inbound transportation costs from plants and suppliers
  • Outbound transportation costs to customers and intra-company facilities
  • Current inventory levels, in-stock percentages and inventory carrying costs
  • Delivery time to customers
  • Current supply points for vendors and production facilities
  • Distribution of customer demand

This information provides a baseline for comparing alternative scenarios. Without the baseline, it is difficult to evaluate the costs and benefits of each alternative versus the status quo.

As the baseline information is gathered, it is important to analyze and validate it against information available from alternate and independent sources within the company. It is not uncommon for databases or database inquiries to yield incomplete results that would potentially skew the analysis. Cost information should be compared with source documents, the general ledger or profit-and-loss statements. Volume information from production or distribution should be compared with volume information from purchasing or sales. Stakeholders who would be affected by any changes in the distribution network should also review the baseline information to make sure that it represents the world as they know it.

Developing Alternatives

Once the data has been collected and validated, the next step is to develop alternative site locations and operating methods. The inputs used to determine alternatives are site visits, future requirements, database analyses and customer service surveys. The methods used for the selection of each site will vary. The main factors influencing site location are listed in Table 6.2.

Table 6.2: Bases for Site-Location Decisions
Supplier and MarketHow quickly suppliers reach you and you reach markets (delivery days); volume, certainty and variability of supply or demand; international border issues
TransportationHighway, water and rail access; weather restrictions; congestion; road limitations; transportation penalties, premiums or benefits
Government & UtilitiesTaxes, incentive programs, planning and zoning, energy cost
LaborUnions, right-to-work laws, wages, skills available, holiday observances
Real Estate

Availability, cost per square foot, site restrictions, proximity to markets

Typically, site selection is based on the factors shown in Table 6.2, and methods are available for weighing alternative sites based on customer volumes and distance. These methods range from experience to determining the total road miles to running a network model. Network models fall into three general categories:

  • Centroid (center of gravity) analysis
  • Optimization models
  • Simulation models

A centroid analysis calculates the weighted center of customer demand by using map coordinates and customer volume. It was one of the first methods used to determine site location, but it is inadequate when compared with today’s modeling techniques.

Optimization models come in a wide variety of complexity and sophistication, with prices to match. They are typically linear or mixed-integer programs that are capable of determining an “optimal” distribution network based on the data, assumptions and parameters provided. Changes to any of the assumptions, parameters or data will cause the model to yield a different result. They are therefore very dependent on the quality of the data and parameters and the experience of the individual performing the modeling analysis.

Simulation models, like optimization models, come in a variety of shapes and sizes. Unlike the optimization model, which starts with a set of data and gives a single answer, a simulation model will start with a single answer-a network alternative or scenario-and examine the impact of a variety of data sets on the scenario, over time. Simulation models are very useful for determining the impact of supply or demand variability, network constraints and bottlenecks on the efficient operation of the network. Like optimization models, they are very dependent on the quality of the inputs and the skills of the modeler.

The various techniques differ largely in their ability to analyze complex networks. The planner needs to determine how important it is to include complex variables or if assumptions and averages can provide sufficient grounds for decision-making. Centroid analysis, for example, can be done on paper and assumes that transportation costs are proportional to distance, that the straight-line distance between two points is representative of actual distance and that there is a uniform shipment size from or to each location. It ignores capacity constraints, service requirements and differences in transportation and facility processing costs. An optimization-modeling program is more sophisticated, but it is limited to evaluating a static range of variables. If a network can be described by summarized data, or by looking individually at one or more slides in time, then an optimization model is very effective. For example, it may use an annual average when looking at the shipping volume of a facility even though there may be wide seasonal variances. Alternatively, such sensitivity analysis could be done “outside the model” to gain an understanding of the impacts of supply or demand variability. Simulation modeling is able to use actual data that reflects such variances or include randomness in sales or supply patterns. This may better represent the volatility a company faces in the real world. The tradeoff for the increasingly complex capabilities is increasing cost.

Location is not the only decision to make. Companies must consider operating methods, as well as choices like consolidating vendor shipments, centralizing slow-moving items in one place and minimizing order delivery times, which could impact future revenue. Once alternative sites have been determined, data must be collected on freight rates, warehouse cost and labor cost for those sites.

Modeling the Annual Operating Cost

Modeling software does not guarantee the right answer. Modeling should be used only as a tool to aid in the decision process. Sometimes interpretation should be pursued. The real value in distribution planning is the knowledge gained from understanding the workings of a company’s distribution system and applying imagination to the model in ways that will truly benefit the distribution network. Facility alternatives can be close in cost but range widely in other factors. That makes it important to have other criteria by which to judge the modeled costs. For instance:

  • Central administrative and order processing costs. Typically, these costs increase with the number of warehouses. It takes more effort to coordinate and manage a larger network of facilities.
  • Cycle and safety stock carrying costs. More warehouses means more total system inventory. Inventory theory supports the notion that safety stocks will increase with the number of facilities.
  • Customer order-size effects. Customers who operate close to a warehouse generally tend to order more frequently and in smaller quantities than customers who are farther away. This implies that delivery costs tend to increase on a dollars-per-hundredweight basis as the number of facilities increases.
  • Inter-warehouse transfer costs. The more distribution centers there are, the greater the coordination problems and the more likely the need to transfer inventory between facilities due to imbalances.
  • Negotiated reduction in warehousing and delivery costs. The fewer the facilities, the greater their individual volume and the more opportunity there is to negotiate more favorable arrangements for warehousing and delivery service.

Because this article is concerned with the overall approach to distribution planning, it does not discuss the various techniques available to model annual operating cost. However, no matter which modeling method is used, the overall approach should closely resemble the following steps:

  • Validate the existing network. Run a computer model to simulate the existing cost. Compare this cost with actual cost.
  • Run alternative networks. Once the model is validated, run alternative networks for present and forecasted volumes.
  • Summarize runs and rank. Create a table to summarize cost by alternative. The table should list individual distribution center costs.
  • Summarize all annual costs and service factors. Create a table that shows, by alternative, all cost and service factors.
  • Perform a sensitivity analysis. Sensitivity analysis is based on the idea of setting up runs that fluctuate some components of the data. One might be a cost that is uncertain or has the potential to change. By modifying this one parameter, you can determine the effect on the run.
  • Determine all investment costs associated with each alternative. Look, for instance, at the costs of new warehouse equipment required to save space, expansion and construction or any building modifications such as adding dock doors.

Evaluating Alternatives

An economic analysis compares the benefits of a recommended network plan with the implementation cost. To perform this analysis, you must determine all the investments and savings associated with each alternative. Costs such as new warehouse equipment, construction or any building modification should be included. Additionally, the following information must be identified: personnel relocation, severance, stock relocation, computer relocation, taxes, equipment relocation and income from the sale of existing land and buildings.

The result of this evaluation should be the return on investment of each alternative compared with the baseline. Once this step has been completed, perform a sensitivity analysis that fluctuates various costs and savings to see which alternatives are the most stable. To round out the analysis, perform a qualitative analysis looking at factors such as customer service and ease of implementation. Once a conclusion is reached, draw up a time-phased implementation schedule that lists the major steps involved in transferring the distribution network from the existing system to the future system.

While designing an optimal distribution network in an unpredictable world may seem overwhelming, Tompkins is here to help. Contact us today to learn how to learn how to develop a successful network strategy during these uncertain times.

Newsletter Signup


Sign up for our latest Insights and News.
Join over 50,000 others, it’s completely free!

I understand that Tompkins will only use this information to contact me about business opportunities. By completing this form I am confirming that I have read and accept the Privacy Policy.