5 BUSINESS FORECASTING

  • These steps are a systematic way of initiating, designing, and implementing a forecasting system

  • When used regularly over time, data is collected routinely and calculations performed automatically

  • There is seldom one superior forecasting system

  • Different organizations may use different techniques

  • Whatever tool works best for a firm is the one they should use

## Warning: package 'DiagrammeR' was built under R version 4.0.5
%0 rec1 1. Identify the purpose of forecast rec2 2. Collect historical data rec1->rec2 rec3 3. Plot data and  identify patterns rec2->rec3 rec4 4. Select a forecast model that  seems appropriate for data rec3->rec4 rec5 5. Compute forecast for period of historical data rec4->rec5 rec6 6. Check forecast accuracy with one or more measures rec5->rec6 rec8 8b. Select new model or adjust parameters of existing model rec5->rec8 rec7 7. Is accuracy of forecast acceptable? rec6->rec7 rec9 8a. Forecast over planning horizon rec10 9. Adjust forecast based on additional qualitative information rec9->rec10 rec11 10. Monitor results rec10->rec11 rec7->rec8 rec7->rec9

FIGURE 5.1: Forecasting process steps

%0 rec1 Qualitative models rec2 Time-series models rec1->rec2 rec4 Delphi  method rec1->rec4 rec3 Causal  methods rec2->rec3 rec5 Moving averages rec2->rec5 rec6 Simple regression rec3->rec6 rec8 Executive opinion rec4->rec8 rec9 Exponential smoothing rec5->rec9 rec10 Multiple regression rec6->rec10 rec11 Sales force composite rec8->rec11 rec12 Trend projections rec9->rec12 rec13 Consumer market survey rec11->rec13 rec14 Decomposition rec12->rec14

FIGURE 5.2: Forecasting techniques

  • Time-series models and causal methods are both quantitative forecasting techniques