Optimization of Demand Forecasts

Forecasting is a fundamental element of business procurement and sales planning. Demand information, which are based on market and industry knowledge of individual employees, and historic demand figures serve as an information basis to forecast future demand. Different mathematical forecasting methods have been established which are supported by APS and SCM systems (e.g. moving average, linear regression or exponential smoothing). Our investigations showed that the selection and parametrization of forecasting methods have a significant influence on the attainable forecasting quality. When choosing a forecasting method the demand needs to be considered differentiated on article location level: Therefore individual demand variations of different articles (fast-movers, slow-movers, seasonal items) require the use of diverse forecasting methods. The temporal granularity of the demand forecast (e.g. weekly, monthly or quarterly forecasting values) also has an influence on the attainable forecasting quality and the selection of a forecasting method.

To optimize your demand forecast we offer:

  • Classification of product range (e.g. ABC analysis, XYZ analysis) to derive suitable product classes for the forecast
  • Performance of ex post analyses to evaluate the suitability of different forecasting methods in order to select and individually parameterize forecasting methods
  • Development and implementation support of business processes for forecasting