FORESIGHT Issue 68

Description

2023:Q1

Special Feature ~ Does Forecast Accuracy Even Matter?

• Johann Robette leads off our special feature section by utilizing a value-oriented simulation to investigate the question Does Improved Forecast Accuracy Translate to Business Value? His experiment demonstrates that better accuracy does not guarantee monetary benefits, and that much forecast improvement effort is wasted.
• In Using Simulation to Determine When Forecast Accuracy Matters, Stephan Kolassa notes that whether accuracy matters must be assessed in the context of the business decision being addressed.
• In Increased Bullwhip in Retail: A Side Effect of Improving Forecast Accuracy with More Data, Arnoud Wellens, Robert Boute, and Maximiliano Udenio studied a Belgian food retailer and found that adding explanatory variables to statistical and machine-learning models increased forecast accuracy.
• Steve Morlidge explores the formidable complexities involved in Measuring the Cost of Forecast Error, proposing a proxy for cost based on features of forecast error such as bias and avoidable error.
• Longtime research collaborators John Boylan and Aris Syntetos address the fundamental question of Why Does Forecast Accuracy Matter?, noting that accuracy may be irrelevant if the forecast is not being used to inform decisions and actions.
• In Better Forecasts or More Appropriate Stock Control Policies?, Evangelos Theodorou and Evangelos Spiliotis question whether we should focus our efforts on forecast accuracy or improving inventory policies.
• Sinisa Vukovic concludes our special feature with a personal take on the difference between Accuracy and Usefulness in Applied Forecasting.

Articles

PRACTITIONER’S CORNER Patrick Bower joins Foresight to provide a regular Practitioner’s Corner column with his perspective on the issues of the day, drawn from a lengthy career in the forecasting and supply chain planning arenas. Patrick’s first column asks, Why Do We Talk about Forecast Accuracy Measures (So Much)?

INTEGRATED BUSINESS PLANNING Foresight’s Integrated Business Planning editor Niels van Hove envisions A New Approach to Business Planning during Crises, suggesting that traditional IBP is too cumbersome during times of disruption. He proposes the formation of a decision squad with the skills and resources to turn decisions around in a timely manner. A Commentary by Vaishal Patel, George Monokroussos, and Jason Chen describes the actions taken by online retailer Wayfair during COVID, many of which are akin to van Hove’s recommendations.

SPOTLIGHT The Spotlight section shares interviews of Foresight authors, Foresight staff, and IIF members. In this issue we meet Oyebimpe Adeniji, a PhD student at the University of Ibadan in Nigeria, and President of the West Africa Young Statisticians Association, along with Anne-Flore Elard of Kinaxis – our new column editor covering Forecasting Support Systems.

OPINION-EDITORIAL SECTION There is generally much to learn from the perspectives of others, and the more diverse the perspectives the better. Just as combinations of forecasts tend to perform better than an individual model, so do varied perspectives provide a collective value greater than the individual parts. To help gather and share those perspectives – especially the most informed and provocative ones – this issue introduces a new Opinion-Editorial section, where we invite readers to exercise their voice. For our first Op-Ed, Associate Editor Simon Clarke discusses Business Forecasting: Issues, Current State, and Future Direction.

MINITUTORIALS Foresight has a history of printing full-length tutorials, and even offers a Guidebook on Forecasting Method Tutorials in the IIF Bookstore. We now introduce the Minitutorial, a concise (two pages maximum) and visual (when possible) explication of topics all practicing forecasters should be aware of. In this issue, Stephan Kolassa instructs us on the “pinball loss” – a metric commonly used in research papers to assess quantile forecasts – and I provide a mildly infographic presentation of Forecast Value Added.