Beyond the forecast: Rethinking demand-driven planning

Editor’s note: This article first appeared on the University of Tennessee, Knoxville’s Global Supply Chain Institute’s blog. It is being reprinted with permission. You can read the original post here. It is the second in a two-part series. You can read the first article here.

Author’s note: The white paper, “Love to Hate the Forecast: Segmenting Planning Demand Triggers to Drive Total Value,” co-authored by Haslam College of Business faculty members Mike Burnette and Lance Saunders, and edited by Ted Stank and Dan Pellathy, marks the 40th white paper released by the UT Global Supply Chain Institute. The paper was released during the Spring Supply Chain Forum and is available for digital download.

While inaccurate forecasts create service and inventory defects, the deeper issue is misalignment between demand signals and supply system capability. The solution is not eliminating forecasting altogether. Rather, it is segmenting demand triggers and applying forecasting only where it creates the highest total value.

The Global Supply Chain Institute white paper, “Love to Hate the Forecast: Segmenting Planning Demand Triggers to Drive Total Value,” argues that the answer is not simply mathematical—it is cultural.

Right-to-Left (RtL) planning

Leading-edge companies are adopting Right-to-Left (RtL) supply chain planning. RtL synchronizes activity as closely as possible to actual consumption. Instead of defaulting to forecast-based planning, organizations segment SKUs and apply demand triggers that best match variation patterns.

The vision is ambitious: provide consumers with the exact product they desire at the moment of next consumption, maximizing total value across the enterprise.

Understanding variation

Three sources of demand variation must be assessed:

• Consumer demand variation
• Customer demand variation
• Company demand variation

In many systems, company-driven variation exceeds consumer-driven variation—an indicator that internal behaviors are distorting the demand signal.

When forecasting does not make sense

Forecasting begins to lose its value when the data reveals persistent structural bias. If monthly forecast results are biased high or low more than 60% of the time, the issue is not random error—it is systemic distortion.

Another warning sign appears when customer service performance and inventory days on hand consistently lag competitors. In these environments, improving forecast accuracy alone rarely fixes the problem. Instead, the underlying demand signal is being influenced by internal behaviors.

Cultural signals are equally revealing. When sales rewards are tied to beating the forecast, when general manager compensation depends on forecast comparisons, or when ownership of revenue and inventory is split across functions without integrated accountability, forecasting becomes political rather than operational. Under these conditions, statistical refinement simply sharpens a distorted signal.

Additionally, when shipment patterns fail to follow clear regression-based slopes or when variation is driven primarily by internal promotions, pricing decisions, or end-of-quarter revenue management, the forecast becomes reactive to internal noise rather than consumer behavior.

In short, forecasting is least effective when internal variation exceeds external demand variation.

When forecasting does make sense

Forecasting remains valuable in specific contexts.

It plays an essential role in long-range business planning, particularly in six-month to five-year horizons where strategic investment decisions must be informed by directional demand expectations.

It is also indispensable for new-to-the-world product introductions, where historical shipment data does not yet exist. In these cases, statistical modeling, combined with structured assumption testing, provides necessary guidance.

Forecasting is most effective when consumer variation is the dominant source of demand fluctuation and when shipment patterns follow identifiable statistical slopes. In these environments, regression-based forecasting tools can meaningfully improve planning precision.

Equally important, forecasting works best in organizations that rigorously measure bias, detect minimal distortion, and operate without financial incentives tied to beating the forecast. When the organization embraces the forecast as a value-added tool, rather than a scoreboard, it regains its strategic relevance.

Case study results

A Fortune 500 CPG company reduced finished goods inventory from 180 days on hand to below 100 after implementing a produce-to-shipment RtL trigger, while improving service performance for 48 consecutive months. A regional food manufacturer identified unused capacity and reduced inventory by 25–30% without sacrificing service. An electronics company implemented a rate-based trigger and reduced safety stock by approximately 20% while maintaining service levels.

These examples demonstrate that forecasting should be applied intentionally, not universally.

Supply chain leaders must segment SKUs, assess variation patterns, eliminate cultural distortions, and build capabilities that move planning closer to actual consumption.

Written by UT professors in collaboration with GSCI partners, white papers translate rigorous research into practical insights for business leaders. The institute’s applied research has been featured in Forbes, Harvard Business Review, Supply Chain Management Review, and The Wall Street Journal. To learn more about how your company can partner to explore advanced supply chain management concepts, visit ASCC.

About the author

Michael H. Burnette is a Global Supply Chain Institute fellow at the University of Tennessee, Knoxville. Burnette came to UT after a 33-year career as a supply chain executive at Procter and Gamble. Most recently, he was the P&G Global Supply Chain leader for Skin Care ($2 billion+ Olay brand) and P&G Global Supply Chain Leader for Hair Care ($4 billion Pantene and Herbal Essence brands). His supply chain leadership and expertise include supply strategy/design, manufacturing, logistics, innovation, PLCM, acquisitions and human resources.

Burnette teaches supply chain courses and manages multiple GSCI projects, including coordinating and publishing white papers based on research conducted between UT faculty and industry leaders. He is a consultant, speaker and co-author of the book Supply Chain Game Changers.

SC
MR