Learn Live Jobs Discussions Nilendra Bajpai Navigate Q&A Forecast Error While it has already been established that demand forecasting is an extremely important step in the entire supply chain planning, it is equally vital to measure the accuracy of forecasting and to clearly understand when and where the forecasting has gone wrong. Measuring the forecast accuracy can be done in a number of ways. Some of these include the following: Forecast error Forecasting errors are calculated by taking the summation of the deviations between the forecast figures and actual figures. Forecasting error = (Actual demand - forecast demand) * 100 / actual demand Standard deviation Standard deviation is also a very critical measure of the accuracy of a demand forecast. Forecasting is usually performed by a team of forecasters who provide their own take on what the forecast number should be. Based on the numbers provided by each team member, the mean and standard deviations are calculated. Standard deviation signifies the spread of observations from the average. The higher the standard deviation, the farther away from the mean the forecast number is. If an organisation decides to proceed with a demand forecast that has a high standard deviation, then the risk associated with such a decision is high as the actual demand could be far away from the mean of the forecast number. The lower the standard deviation, the higher the probability that the forecast number is closer to the mean. If an organisation decides to proceed with a demand forecast that has a low standard deviation, then the risk associated with such a decision is also low as the actual demand will not be far away from the mean of the forecast number. Question 1/1 Mandatory Consensus vs Average Techniques of Forecasting To accurately predict the demand for its products for the next year, HUL (Hindustan Unilever) follows a couple of forecasting techniques, both qualitative and quantitative in nature. Consider the scenario of the Cornetto brand of ice-cream from HUL. The company has a team of five forecasters who predict the demand for Cornetto. In the first stage, every forecaster predicts the demand using the quantitative demand forecasting method of their choice. The following table provides the forecasted figures from each of the forecasters. Forecaster Forecast Demand (in million units) 1 11 2 18 3 14 4 12 5 15 Once the individual forecast figures are ready, the forecasters use two techniques to arrive at a single figure. The first technique is consensus-based forecasting, wherein all the five experts give their opinions on why their predicted figures are correct and then arrive at a consensual figure (a single demand forecast agreed by all teams). In this scenario, the team arrived at a consensual figure of 15 million units. In the second technique, the forecasting team takes an average of all the forecast figures. For the given scenario, the average figure comes out to be 14 million units. The company then proceeded with these two figures in mind. After the end of the year, the actual demand was found to be 16 million units. Determine the forecasting error for both the techniques. Hint: Forecasting error = (Actual demand - Forecast demand) * 100 / Actual demand Consensus-based technique = 10%; average taking technique = 4% Consensus-based technique = 1.25%; average taking technique = 2.5% Consensus-based technique = 16.5%; average taking technique = 22.5% Consensus-based technique = 6.25%; average taking technique = 12.5%

Consensus-based technique = 10%; average taking technique = 4%

Consensus-based technique = 1.25%; average taking technique = 2.5%

Consensus-based technique = 16.5%; average taking technique = 22.5%

Consensus-based technique = 6.25%; average taking technique = 12.5%