Build Resilience in Supply Chains by Modeling Multiple Demand Futures

Nari Viswanathan
Nari Viswanathan
Sr. Director, Product Segment Marketing, Coupa

Nari is currently Sr. Director of Product Segment Marketing at Coupa, where he helps bring products to markets in the areas of Supply Chain Design and Planning. Over the past 20 years, Nari has held VP and Director of Product Management, Research and Marketing roles at Aberdeen Group, River Logic, Steelwedge and E2open. He has significant experience building products from the ground up and managing the P&L for a product suite. He is a proven B2B marketer with expertise in content marketing, competitive intelligence, and positioning. He has published numerous thought leadership articles, whitepapers, blogs and delivered dozens of webinars during his career. Nari Viswanathan is a five times SDCExec Supply Chain Pro to Know award winner. Nari holds a master’s degree in Manufacturing Systems Engineering at the University of Wisconsin-Madison and a bachelor’s degree in Mechanical Engineering at the Indian Institute of Technology, Chennai.

Read time: 3 mins
Build Resilience in Supply Chains by Modeling Multiple Demand Futures

Being demand driven and focusing on demand forecast accuracy is a key tenet of supply chain success. However, one mistake that companies make is to be hyper focused on a single number forecast that is often determined by a combination of statistical forecasting and collaborative inputs. This process is also referred to as demand planning. We believe that for companies to be resilient, they must augment their existing demand planning approaches with a demand modeling approach that simulates multiple futures.

What is demand planning?

Demand planning is a process used in supply chain management to forecast demand. When done correctly, it helps ensure products are delivered on time by anticipating customer wants and needs. In so doing, demand planning results in more customer satisfaction, better revenue forecasting, and inventory management that aligns stock levels with the peaks and valleys in demand.

Demand planning requires staff to examine internal and external factors that could have an impact on demand in either direction. Such factors may include increased customer interest in a product or service, natural disasters, shifts in weather patterns, political or regulatory issues, and global crises. For planning to be effective, organizations must strike a balance between having enough inventory available to meet expected demand, vs risking stockout or inventory writeoffs due to excess.

Limitations of demand planning

Traditional time-series demand forecasting typically means taking an inside-out approach, often relying on an organization’s historical, internal data to predict demand. The area that companies make a mistaken assumption though is about the medium-long-term period demand. The assumption is that long-term demand plans can be just generated by extending the time period to three or four years, whichever may be the required horizon. The past is not reflective of the future especially in an era where so many disruptions and large scale changes are happening throughout the world.

The danger is that demand planning doesn’t take into account external factors that continually impact the market. Change is accelerating faster than inside the four walls of today’s enterprises. Data is growing exponentially, technology continues to disrupt, customers demand an omnichannel buying experience, and businesses are constantly grappling with macroeconomic conditions and industry complexities. While historical performance may offer helpful insights toward what the future holds, change is happening far too rapidly to rely on internal rear-view mirror metrics.

Role of demand modeling

Companies need demand modeling to augment their existing demand planning processes especially for the operational and strategic horizons. As the name indicates, the first key requirement is to build a realistic demand model from the ground up. The demand needs to be broken down into a series of internal and external causal factors, and the impact of each of these factors on the demand needs to be modeled. Machine learning methods are utilized to leverage external factors such as weather, Consumer Price Indices, housing starts, demographic trends, and Gross Domestic Product to model the demand. Uncertainty data is then utilized to come up with a probabilistic forecast that can be used to feed the supply chain design digital twin and run a host of scenarios to identify points of resiliency challenges and optionality.

The probabilistic demand forecast represents the multiple futures that an organization could face if supply were unconstrained. However, since most companies face supply constraints (especially in the new normal) supply chain design plays the role of aligning the long-term demand with supply. Where gaps are identified, optionality of supply sources is identified through the supply chain design along with cost/service/revenue implications. This process must be repeated on a continuous basis to keep long-term demand and supply alignment while factoring in potential risks as part of demand modeling.

Demand modeling and supply chain design are a potent combination of capabilities that will accelerate the ability to manage risk and resiliency related challenges. Look forward to more details on this innovative offering from Coupa in the near future.