Question

You are asked to provide sales forecasts of several products for a large biscuit manufacturing company. Define the steps to perform an effective forecasting in the context of this company.

Answer 1

The steps for effective forecasting are as follows:

1. Identify the Problem

The first step is to identify the problem regarding what has to be estimated and why. You could simply ask how the market is reacting for your biscuit or how the company's sales would look like in a few months to conduct the forecasting. You should have the knowledge about how the market works effectively, who are your competitors and what is your customer base for the product offered.

2. Collect Information

Information is collected as per the past data available with us to conduct the forecasting method. This data which has been effectively collected by the knowledge and research experts of the company is converted into information by the judgement and analysis by the experts. If the judgement is based on numbers, quantative forecasting would be performed, if not, qualitative would be performed. In our case, a quantative data forecasting for the sales forecast of biscuit would be performed as per the past data records.

3. Perform a Preliminary Analysis

A preliminary or early analysis would be performed to make sure if the data weare using is usable or not. Patters and trends would be effectiveky displayed for the past data records that can help for the future estimation for example one can also evaluate the data and also make rightful assumptions to reduce the amount of data to be more accurate for the forecast made.

4. Choose the Forecasting Model

Once the data is evaluated and the information is collected and treated, you can choose the model you think would work the best for the forecasting method. There are a no. of methods like regression method, exponential smoothing method which could be most effectively used in this case to forecast the no. of biscuits for the sales for the company as a quantitative forecasting.

5. Data analysis

This is a simple where after you choose the most suitable model for forecasting, the same method is implemented and the data is run through it.

6. Verify Model Performance

Finally, the forecasted data would then be compared with the actual data like the forecasted sales volume of the biscuit would be compared with the actual one when the right time comes to make sure if the method we used for forecasting was effective or not and how we can improve out forecasting.