A Perfect Marriage: Demand Modeling and Supply Chain Design

The supply chain industry is moving so quickly that even the most seasoned professional is likely to face a “well, I’ve never seen that before” moment.  With the unprecedented pace of change in regulations, political and socioeconomic shifts, and rapidly evolving customer preferences, it’s increasingly critical for supply chains to be able to rapidly respond, or risk reduced profit and market share. Designing a more responsive supply chain is more than just calculating optimal cost and service levels. You need to understand the key drivers of your demand and have the ability to accurately predict and test future demand scenarios for more confidence in critical supply chain decisions, while also responding to them quickly.


Demand modeling has emerged as a key decisioning tool in supply chain design. Standard forecasting tools can fall short if they can’t take into consideration external causal factors that drive demand like weather and economic and political indicators, or help predict demand 5-10 years out when making long-term strategic decisions.

The marriage of demand modeling and supply chain modeling technology offers a better demand signal that drives a better supply chain design for the future. Here are three examples of how leading organizations are benefiting from this new approach.

  • Manufacturer leverages demand modeling to better understand key demand drivers of engine demand

Previously, long-term demand modeling for engine sales in the oil and gas sector was done using simple time series algorithms in spreadsheets. Since business and external factors were not considered, there was a lack of understanding of key demand drivers, and very little confidence in the long-term business decision-making.

The manufacturer utilizes demand modeling with built-in causal data to identify 13 potential macroeconomic factors that influences demand, including GDP construction/mining and gas prices. This insight not only improved the accuracy of the model, but also provided more statistical evidence to the business on what is driving the demand to support long-term strategy, potentially freeing up significant additional working capital.

  • Chemical company uses demand modeling to improve long-term capacity planning

This company had been applying simple growth factors of 5, 10, 15 percent to historical demand across its primary product lines to understand capacity needs in the network to meet future demand for the five years. This solution lacked true visibility into product life cycle and key drivers of demand.

Using demand modeling, the company can now build demand models to determine the optimal growth strategies for each product line considering business factors and external causal factors. They can understand key drivers of your demand for each product line and quickly perform what-if analysis on demand drivers. This means it’s much simpler to create alternate growth strategies to test potential supply chain changes and capacity requirements.

  • Home appliance company applies demand modeling to enable holistic inventory optimization

This major home appliance manufacturer had no good way to de-seasonalize and de-trend data to remove known sources of variability and isolate true demand variability. It lacked full understanding of product life cycle (which products are growing and which are declining compared to mature products) and the impact to inventory. The company also needed to better understand the various types of inventory such as prebuild, safety stock, and promotional inventory.

Today, the company is incorporating demand modeling as part of its supply chain design process, allowing analysts to explore demand and extract out demand patterns such as seasonality, trend, lifecycle, and promotions. Forecast error is used as a better signal to drive optimal safety stock targets, and the company has been able to develop a holistic inventory strategy by more accurately modeling the various components of demand as part of inventory optimization.

Now is the Time to Adopt a Demand Modeling Practice

The wealth of uncertainty in the market isn’t slowing down, but with a more accurate view of your demand, you can be better prepared to manage the pace of change.

Vikram Srinivasan is LLamasoft Product Group Manager for all Data and Analytics products, including Data Guru, Demand Guru, and Data Services. Vikram is primarily responsible for the overall product strategy of these products, and managing a business plan to expand customer usage, increase customer penetration and grow product revenue. He also engages with market analysts and industry thought leaders to raise awareness of the Data and Analytics solutions. Prior to his role in Product Management, Vikram worked as a Senior Consultant at LLamasoft on multiple projects solving various supply chain design problems. He received his M.S in Industrial Engineering majoring in Operations Research at The Ohio State University.


Leave a Reply

Your email address will not be published. Required fields are marked *