Demand Driven Forecasting

Demand Driven Forecasting

To access the Forecating Report, go to:

Home > Manufacturing > Reports > Forecasting

The forecasts will be computed using historical Sales Order, Delivery Note, and Quotation data and the exponential smoothing method. The following is the formula for the exponential smoothing method:

The system forecasts the data for each period using the exponential smoothing approach, and the same forecast data will be utilized to anticipate the data for the subsequent period.

How does the exponential smoothing method works

We used monthly sales order data over a one-year period for the aforementioned scenario. The average of all total orders will be used to produce the first-month prediction. The difference between the previous month's total order and forecast value will be multiplied by the smoothing constant value starting in the second month (in between 0 to 1). The Smoothing Constant's default value of 0.3 produces accurate forecasting data. The difference between the total order and forecast value from the previous month is referred to as the forecasting error, and this error will be added to the current month's forecast value to determine the forecast for the following month.

How do the filters work

1. Company By using the company filter, the user may forecast for a certain company.
2. From Date and To Date : The forecasting will be done by the system for this time period, with the default from date being the present date and the to date being the date one year in the future.
3. Based On Document By default, the system forecasts based on historical sales order data. To forecast the data, the user can choose an alternative document like a delivery note or a quotation.
4. Based On The forecast data is displayed depending on quantity and amount.
5. Based On Data ( in years ) : This filter enables the algorithm to check historical data for a number of years because forecasting requires historical data.
6. Periodicity : The forecasting data can be viewed on a monthly, quarterly, half-yearly, or annual basis; however, for better viewing, the older prediction data cannot be presented for the monthly period.
7. Smoothing Constant - Data forecasting will be done using a smoothing constant, whose value should fall between 0 and 1.