Analyst Perspectives

Sales and operations planning (S&OP) is trending toward becoming more strategic in product-centric companies through the end of the decade. The purpose of S&OP has grown in importance. Since the mid-teens, the trade and economic environment has become less benign and more unpredictable, forcing many enterprises to redesign their supply chains for resiliency while still surmounting the dual challenges of remaining cost competitive and achieving financial targets. Over the past decade, there have been repeated periods of commodity and currency volatility, often the result of exogenous factors in a more unstable world order. These circumstances have therefore made it necessary for those managing supply chains to be able to forecast and plan more nimbly, react to changes faster while making better informed decisions about trade-offs.  

Fortunately, a new trend has emerged to help address these challenges: AI using machine learning (ML) and generative AI (GenAI). Supply chain forecasting, especially in larger enterprises, is complex, while supply chain execution is characterized by having to manage a heavy volume of minutiae accurately in short cycles. AI can assist humans to do both faster, less expensively and more accurately. So much so that Ventana Research asserts that by 2026, one-half of supply chain management organizations will have adopted some form of AI/ML to shorten planning and forecasting cycles to promote agility and resilience. 

S&OP is a supply chain-oriented planning process that aligns demand-, supply- and financial forecasts. Demand planning involves analyzing internal projections to generate a set of detailed unit sales predictions — potentially down to the individual stock-keeping unit (SKU) — while considering optimal inventory levels to satisfy customer service requirements. Supply planning involves assessing how the organization can meet demand given available capacity and other internal considerations, supplier commitments and inventory availability. S&OP systems are used to reconcile and match demand and supply while respecting financial objectives and constraints. 

Five years ago, I first wrote about supply chain management in the new era of trade, which was an observation on the strategic challenges facing organizations resulting from the decade-long dismantling of the liberal global trade environment and the shift to more restrictive regimes. I noted that supply chains designed for the lowest possible cost during a benign era had to be restructured for resiliency in the current unsettled environment. This trend accelerated in 2020 as a series of extraordinary events, especially the pandemic-induced shocks, compounded the tactical challenges facing supply chain managers and increased awareness of the brittle nature of a lowest-cost strategy. The impact of this shift has been accentuated by the recent fracturing of the world’s political order which has been reshaping thinking about where goods are sourced and manufactured. Adding to this complexity is the increased regulatory focus on environmental, social and governance (ESG) issues. Ventana Research asserts that by 2027, over two-thirds of organizations will incorporate sustainability metrics in sourcing and supply chain selection criteria to support ESG efforts. Actions and policies in the political sphere will have an even greater impact on economic decisions, requiring even more adaptability in supply chain management. 

Although dislocations, input cost inflation and financial market volatility have diminished lately, the need to be resilient in the face of rapidly changing events has not. A better planning process — not more accurate planning — is the key to resiliency. We live in an age of uncertainty, not unpredictability. An organization’s ability to rapidly respond to changes in circumstances makes it better able to adapt. Dedicated planning software is a basic requirement to achieve fast cycles. As AI and GenAI capabilities are incorporated into software design, the benefits organizations can achieve from using a dedicated application will increase while lowering the barriers to adoption.  

S&OP software can improve an enterprise’s performance, especially for those that have complex supplier relationships, longer lead times for some parts or materials, short product cycles, relatively high seasonal variability, constrained supplies or some combination of any of these factors. Integrating constrained demand planning with supply chain forecasting can cut working capital requirements and related costs and obsolete inventory charges. By improving the coordination between marketing, sales, manufacturing, purchasing and supply chain management, software can reduce the incidence of stock-outs, potentially boosting revenue and customer satisfaction. Because software can cut the time it takes to execute a planning cycle, companies can react faster to events and implement changes that optimize plans consistent with their specific tactical and strategic objectives. 

The application of AI/ML has potential to significantly improve S&OP and supply chain management. The use of predicative analytics in forecasting can improve its accuracy because the process can suss out a richer set of factors that contribute to a business outcome and quantify their relative importance. Because ML is designed to be always-on, the system is continuously testing to ensure that the model is up to date, accurately reflecting current market and environmental conditions. Understanding the relevant factors to an outcome and their importance allows the system to notify users when trends in outcomes diverge meaningfully from the prediction and highlight the likely reasons why. Early warning of a divergence from forecast allows planners to revise their projections sooner to respond with agility. 

Supply chain execution will also benefit significantly from AI because of the intricate orchestration of processes and the granularity of data involved. Supply chain planning and S&OP occupations traditionally involve the tedious mastering of an almost infinite number of details but also demand experience, insight and cleverness. By automating the many repetitive and unambiguous steps currently handled by people, AI can cut unproductive workloads, shorten cycles and enable supply chain managers and planners to focus on the work that makes best use of their skills, shifting the balance of human effort from the wearisome to the inspired. AI can automate the ingestion of documents such as purchase orders and bills of lading, reducing manual effort and decreasing the incidence of errors and omissions in data used end-to-end in acquisition and manufacturing processes. AI can flag errors or ambiguous signals and then suggest alternate courses of action. GenAI can generate first drafts of routine internal and external communications for review and editing, cutting the time needed to complete these repetitive tasks and helping to ensure consistency. The list of use cases that can easily be put into practice over the next three years is much longer than this.  

Dystopian visions of AI see it destroying work and creating mass unemployment in certain professions such as accounting and supply chain management. I suggest that the evolution and adoption of AI will follow the usual contours of technology innovation. As usual, there will be some dislocations but for society overall, the value of the technology in terms of increasing productivity, reducing costs and waste as well as increasing affordability of goods will far outweigh the negative impact. I recommend that product-centric enterprises that currently do not use S&OP software reassess their supply chain planning and execution processes and consider the use of a dedicated application enhanced with AI to improve their performance. 

Regards,

Robert Kugel