A key component of network optimization/modeling revolves around data collection and analysis. I’ve been a network modeler for over 18 years and I’ve found that data analysis with respect to validation and model building lends itself to various “outside the model” opportunities. Potential uses of network optimization data (without building a model) are discussed below.
Customer Dispersion
It is often interesting to view customer shipments (weight or units shipped) through use of a heat map. Viewing the map reveals where top customers are located from a sales or units perspective, but it is interesting and often eye-opening to see where that differs when viewed by weight shipped. Not only does it validate where customers are concentrated, but it may highlight untapped markets. Aside from obvious gaps in customer penetration (no shipments to a given state or region) you may find existing markets that are underserved. Comparing customer demand to census data is a useful step to project where there is a potential sales uplift.
Plant Direct Potential
Whether or not plants are used currently to ship directly to customers, another use of customer shipment data may be to analyze plant direct potential. Start by aggregating weekly shipments by customer, which consist of items produced at a given plant. Assuming plants would ship only truckloads you could further refine the analysis to eliminate smaller shipments. Finally, some consideration must be given to the plant’s ability to hold inventory and build customer shipments vs moving product out of the plant to a distribution center once its manufactured. Once plant direct volume is identified you could apply linehaul rate assumptions to derive a cost to serve from the plant vs the existing cost to serve from a warehouse. Keep in mind that you eliminate the costs to ship product from the plant to the warehouse and additional handling costs at the warehouse by shipping directly to the customer from the plant.
Potential Use of a Wholesaler
Smaller customer shipments (LTL) often suggest more warehouses to gain transportation savings which may or may not be feasible or desirable. One way to gain density on trucks and maintain the current network footprint is through the use of wholesalers to reach end customers. Depending on the industry you operate in, there is potential to consolidate smaller LTL shipments to customers within a region into larger LTLs or even TL shipments to a wholesaler. For example, you could look at opportunities to consolidate shipments in a region with large customer concentration but no local warehouse. Aggregating shipments by week would allow you to calculate the truck density improvement. Assuming you convert the smaller LTL shipments to individual customers into TL shipments to a wholesaler you could calculate the anticipated linehaul cost and compare it to the existing cost to serve those customers. There may be some loss of margin on the sale to the wholesaler which would have to be weighed against the transportation savings.
I have worked on many projects where data analysis “outside the model” is requested by the client. In some cases, they have never looked at the data in the manner that I do or they don’t have the bandwidth to perform additional analyses. Regardless, once network optimization data is collected and validated there may be additional uses of the data without even building a network model.
—Dan Gunter, St. Onge Company
