Predicting the Future of Consumer Demand

Improving demand forecasting is a priority for manufacturers—61% of respondents to a recent AIMMS report say that improving forecast accuracy is either important or extremely important for their organizations. However, only 2% of respondents are extremely satisfied with their forecasts. Many bet on new technologies such as machine learning and demand sensing to improve forecast accuracy (see chart). According to the survey:

Demand-forecasting technology makes a difference in manufacturers’ satisfaction:

• Only 2% are extremely satisfied with the tools they use for demand forecasting
• 39% use spreadsheets for demand forecasting, and nearly 45% of them are dissatisfied
• 30% use a specialist forecasting package, and 61% of them are satisfied
• 23% use bundled enterprise resource planning, and most feel indifferent about the tool

Organizations are looking into alternative techniques to improve demand forecasting:

• 56% are exploring statistical modeling of historical demand at their organizations
• 41% are looking into machine learning
• 41% are investigating demand sensing
• 60% are curious about new techniques such as demand-driven material requirements planning