Mastering Forecasting: Insights from Overseeing over 100 Demand Planning Initiatives

Join me, Kevin Boake, Global CEO at b2wise, as I share 6 essential lessons from my extensive experience in implementing demand planning and forecasting processes.

Introduction: In the realm of supply chain management, the significance of forecasting cannot be understated. While the debate around Demand-Driven Material Requirements Planning (DDMRP) has previously suggested it operates without a forecast, I can assure you that a forecast is indispensable, albeit in a unique manner that enables a much more stable planning process. This article explores the six key mindsets to adopt when trying to implement a forecasting process.

1. One Forecast, Many Voices: The Power of Consensus: The cardinal rule when designing an effective forecasting process lies in the unity of vision. Implementing a single forecast, derived through a consensus process that accommodates the views of the different divisions of your business, is paramount. Allowing each department to have its own forecast breeds mistrust and hinders collaboration. By fostering agreement and synthesizing insights, a unified forecast becomes the backbone of a stable and transparent system, even if it is not that accurate to start with.

2. AI in Forecasting: Trust, But Verify: Artificial Intelligence (AI) has revolutionized forecasting capabilities, yet blind reliance can lead to unforeseen challenges. The lesson here is to surround AI forecasting with checks and balances. High-level validations should be integrated to identify and rectify exceptions, ensuring the technology augments human intelligence rather than replacing it entirely.

3. Hierarchy of Accuracy: From Annual to Weekly Forecasts: The accuracy of forecasts is directly proportional to the level at which they are conducted. Forecasting at a higher level, such as annual sales for the entire company, yields more reliable outcomes than attempting to predict weekly sales for a specific product in a particular region. Employing hierarchical techniques, where forecasts begin at a broad level and are subsequently refined, enables a balance between accuracy and practicality.

4. Not Every Item is Forecastable: Recognizing that not every item is easily forecastable is a crucial insight. Generating a time series forecast for particularly slow or erratic movers can be near impossible. In my experience, it is often more effective to forecast in annual buckets and use stocking policies to set a 'bin' level that automatically replenishes when the sale finally occurs. This adaptive approach accommodates the inherent unpredictability of certain items, ensuring a more pragmatic and responsive supply chain strategy.

5. Formalizing the Forecasting Process: CFO's Seal of Approval: Formal processes lend credence to forecasting endeavors. Seeking approval from top executives, especially the Chief Financial Officer (CFO), elevates the importance of the monthly forecast. The act of formalization instills accountability, prompting teams to invest more effort in the meticulous review and refinement of numbers.

6. Performance Measurement as the Compass for Improvement: Improvement necessitates measurement, and this holds true for forecasting. Establishing a robust performance measurement system enables organizations to conduct root cause analyses when discrepancies arise. Identifying why certain forecasts deviate from actuals empowers teams to iterate and refine their forecasting methodologies continuously.

Conclusion: In the dynamic landscape of supply chain management, forecasting remains a cornerstone for effective decision-making. Embracing a unified forecast, cautiously leveraging AI, recognizing the hierarchy of accuracy, acknowledging the inherent challenges in forecasting certain items, formalizing processes, and measuring performance are indispensable steps toward achieving excellence in forecasting. As organizations embark on the journey of continuous improvement, these key insights serve as guiding principles for navigating the complex terrain of supply chain forecasting.