Forecasting with Judgment

This chapter explores the roles that human judgment plays in forecasting in organisations. It focuses on the latest research findings to examine why, despite advances in predictive analytics, the rise of machine learning and the availability of Big Data, forecasts still often rely heavily on judgment. We identify the circumstances where judgment brings benefits to forecasts, as well as the dangers that motivational and cognitive biases bring, leading to inaccurate forecasts. Strategies for improving judgment in forecasting are then evaluated. These include providing feedback, restricting interventions, decomposition, correcting forecasts to remove biases, manipulating the time available to produce forecasts, structuring group forecasting processes and integrating judgment with statistical methods. We conclude that, despite advances in predictive analytics, judgment is likely to continue to have a major role in forecasting. There is therefore a need to develop more advanced software systems that provide enhanced support for judgmental inputs to forecasting processes.