Forecasting process is the technique of predicting the future based on historical and current data. These can then be compared (resolved) to what really happens. For example, a corporation might forecast sales for the coming year and then compare it to actual outcomes. Forecasting can relate to formal statistical approaches that use time series, cross-sectional, or longitudinal data, as well as less formal judgmental methods or the prediction and resolution process itself. The terms “forecast” and “forecasting” are sometimes reserved for estimations of values at specified future times, whereas the term “prediction” is used for more general projections.
Steps Involved in Forecasting Process
1- Determine the purpose of the forecast – The first step is to determine why we are doing this Forecasting for example to determine the manufacturing or purchasing targets, capacity
2- Determine what will be Forecasted – Specify which particular unit ,product or SKU will be Forecasted
3- Determine the time horizon – Specify that this Forecast will be for short term ,medium term, or Long term forecast
4- Visualize the data – Map any available historical data on a graph to see if they have obvious trends or seasonality . This will help when selecting the forecasting method
5- Choose the forecasting Method – You can choose qualitative or quantitative methods or both .For quantitative method decide if a time series forecast or an associative forecast would work better. if historical data are available and trend appears to be relatively steady, a time series forecast is a good choice. if no data are available or trend changes frequently, it may be best to develop an associative forecast based on elements that appear to be driving the changes in the trend.
Get free Supply chain certification – CLICK HERE
6- Prepare the data – Gather data to be used as forecast inputs. If the visualisation showed strong seasonality (Seasonality or volatile data due to any specific reason) remove that temporarily.
7- Test the forecast using historical data – if the historical data are available, Prepare a forecast for few periods back from the present and compare the forecast results to the actual historical results. Forecast using multiple methods to find the most accurate one.
8- Forecast – After making any necessary adjustments use the model . if the seasonality was removed from the data add it back in . Any qualitative adjustments would also be made at this point.
9- Review and improve models for accuracy – Monitor and control error levels and continually models. Forecasters should eb providing error statics with their Forecasts and part of your job is to understand this reliability level so you know how much reliance to place on forecasts.
Time Series Forecasting process
A Forecasting method that projects historical data patterns into the future. it involves the assumption that the near -term future will be like the recent past. Time Series forecasting process is more commonly used because the methods are less complex mathematically and thus easier to explain the decision makers . time series forecasting process assume that the factors that influenced the past will continue on into future. when that trend is unlikely to be stable associative forecasting may be needed.
There are number of types of time series of forecasting ranging from the very simple to the realtively complex. forecasting simply assume the last period demand will be this period ‘s forecast. it can be cost effective but does-not account for trends and any random spike or trough in demand would be carried forward.