Forecasting Methodology

DemandCaster includes robust forecasting capabilities to produce best fit or user selected statistical forecasts using historical shipment, order, or point of sale (POS) data. Forecasting may be performed at any level of aggregation or within the context of customers, channels, or regions in the Sales & Operations Planning (S&OP) add-on.


Data sources for the forecast include historical order or ship demand, firm and planned sales orders, blanket orders, customer stocking commitments, inter-company orders, and point of sale data. The Forecasting algorithms form the basis of the the DemandCaster Demand Planning process which may be further supplemented by demand sensing and demand shaping data.

Demand Forecasting comprises of 4 key components:

  1. Import historical sales data: Automatically via Data Integration, or manually via a tab-delimited text file uploaded in the upload portal.
  2. Create statistical forecasts: Forecasts can be generated as often as required (most clients run new forecasts weekly or monthly).
  3. Import customer forecasts: Data Integration can connect to external data sources where such data is shared. Customer or external sales forecasts can also be uploaded using Excel spreadsheets converted to text files.
  4. Review and edit forecasts: Numerous metrics are included to test and evaluate results. DemandCaster includes confidence indices, numerous error measures, trend indices, hold-back testing, and other analytics to test the reliability and continually improve the forecasting process.

Forecast Algorithms

Forecasting models accommodate seasonal demand, product hierarchies, product promotions, slow-moving items, causal variables, outliers and much more:

  • Expert Selection: The built-in expert system analyzes your data, selects the appropriate forecasting technique, builds the model and calculates the forecasts.
  • Exponential Smoothing: Twelve different Holt-Winters exponential smoothing models are provided to accommodate a wide range of data characteristics. The robustness of exponential smoothing makes it ideal when there are no leading indicators.
  • Croston's: Croston's intermittent demand model is provided to accommodate low volume and "sparse" data (i.e., data where the demand is often zero and volumes are low).
  • DemandCaster Intermittent Demand Model: A model that calculates an items reorder point by sampling historical demand. The traditional approach to calculating intermittent demand order points is by using Croston's method with a calculated safety stock. In this case, there is no safety stock applied. In certain situations, this methodology provides a more reliable result.
  • Event Models: Event models extend exponential smoothing by providing adjustments for special events such as promotions, strikes or other irregular occurrences. You can adjust for events of several different types such as promotions of varying sizes or types, or movable holidays like Easter and Rosh Hashanah.
  • Multiple-level Aggregate Models: With the S&OP add-on, multiple-level models allow you to aggregate data into groups that can be reconciled using a top-down, middle-up, or bottom-up approach to produce consistent forecasts at all levels of aggregation. Seasonal and event indexes can be extracted from the higher-level aggregates and applied to lower-level data.
  • New Product Forecasting: Includes several methods for forecasting new products, including forecasting by analogy and item supercession.
  • Curve Fitting: Curve fitting provides a quick and easy way to identify the general form of the curve which your data are following..
  • Simple Methods: Moving average, same as last year, percentage growth, and fixed forecast value models are included.
Was this article helpful?
0 out of 0 found this helpful
Have more questions? Submit a request