Chapter 5 Demand Forecasting

Learning Objectives

• Explain the role of demand forecasting in a supply chain.

• Identify the components of a forecast.

• Compare and contrast qualitative and quantitative forecasting techniques.

• Assess the accuracy of forecasts.

• Explain collaborative planning, forecasting, and replenishment.

Introduction

• Organizations are transitioning to a demand-driven supply chain model.

• Suppliers must align supply with demand more effectively.

• Accurate forecasts benefit trading partners and mitigate supply-demand mismatches.

• Strategies for better supply-demand alignment:

  • Maintain stock availability for timely delivery.

  • Implement flexible pricing to influence demand.

  • Use short-term methods for supply increase like overtime and subcontracting.

The Importance of Demand Forecasting

• A forecast estimates future demand and guides planning decisions.

• The primary goal is to minimize the deviation between actual and forecasted demand.

• Factors influencing demand must be incorporated into forecasts.

• Collaborative sharing of information is crucial for a consensus forecast.

• Benefits of good forecasting:

  • Lower inventory levels

  • Fewer stockouts

  • Smoother production plans

  • Reduced costs

  • Enhanced customer service

Demand Forecasting Techniques

Qualitative Forecasting Methods

• Based on intuition and expert opinion, useful when data is scarce.

• Types include:

  • Jury of Executive Opinion

  • Delphi Method

  • Sales Force Composite

  • Customer Surveys

Quantitative Forecasting Methods

• Use mathematical models with historical data.

• Methods include:

  • Time Series Forecasting: Assumes the future extends past trends (e.g., naive, moving averages, exponential smoothing).

  • Cause-and-Effect Models: Predictive models using independent variables (e.g., linear regression).

Components of Time Series Forecasting

• Trend variations: Long-term directional changes.

• Cyclical variations: Long-term periodic movements exceeding one year.

• Seasonal variations: Regular patterns over time intervals.

• Random variations: Unpredictable events affecting demand.

Performance Assessment of Forecasts

Forecast Error

• Definition: The difference between actual and forecasted quantities.

• Measurement methods include:

  • Mean Absolute Deviation (MAD)

  • Mean Absolute Percentage Error (MAPE)

  • Mean Square Error (MSE)

  • Running Sum of Forecast Errors (RSFE)

Demand Sensing

• Involves identifying short-term trends for more responsive forecasting.

• Should integrate:

  • Real-time inputs from diverse sources.

  • AI and ML for pattern recognition.

  • Automated models with self-adjusting algorithms for daily forecasting.

Demand Planning Post-COVID-19

• Recommended strategies include:

  • Altering forecasting models to respond to new dynamics.

  • Enhancing customer communication.

  • Streamlining Sales and Operations Planning processes.

Useful Forecasting Websites

• Institute for Business Forecasting & Planning: https://ibf.org/

• International Institute of Forecasters: www.forecasters.org

• Business Forecasting Blog: www.businessforcastingblog.com

Forecasting Software

• Tools to simplify forecasting processes:

  • Business Forecast Systems, Inc: www.forecastpro.com

  • John Galt: https://johngalt.com/forecastx-new/

  • Mi9 Retail: https://mi9retail.com/retail-demand-management-software/

  • SAS: https://www.sas.com/cs_cz/software/supply-chain/demand-driven-forecasting.html

Artificial Intelligence and Machine Learning in Demand Forecasting

• AI drives self-learning capabilities in machines.

• ML optimizes predictive accuracy through algorithmic improvements.

• Focus areas include data governance and innovation enhancement through analytics culture.

Cloud-Based Forecasting

• SaaS applications for advanced forecasting with subscription benefits:

  • Enhanced data storage and analysis.

  • Improvement in forecast accuracy and productivity.

  • Reduction in stockout risks and IT costs.

Summary

• Effective demand forecasting is essential for aligning supply chain management, influencing inventory practices, and optimizing decision-making processes.