Introduction
In the modern world, businesses have discovered the power in understanding their market. As such, they have leveraged all the available tools in order to guide their steps into the future. The desired goal can only be achieved by combining knowledge of the market and application of the combination of the right forecasting tools (Green & Armstrong, 2012). Therefore, forecasting is defined as a decision-making tool that is used by business to project its future. Consequently, demand forecasting will be the decision tool that an organization will use to foretell the amount of their product that will be demanded. Broadly, there are two forecasting categories that organization will apply namely judgmental forecasting and quantitative forecasting.
Categories of forecasting
Judgmental forecasting is a form of forecasting that is entirely based on the estimators experience and intuition. In this mode of forecasting, the forecaster is expected to be highly experienced and has an extensive understanding of the market since the forecast is solely based on value judgement. In addition, advocates of this form of forecasting argue that the human brain has the capability to view data in a manner that no computer can. As such, the advocate sites the superiority of human intelligence over artificial intelligence as the underlying reason. As such, they believe that using experienced and knowledgeable personnel will always yield better results compared to the use of computers (Kavanah & William, 2014: Lawrence, Goodwin, O'connor & Önkal, 2011). However, this method is criticized since human brain, despite the superiority of human intelligence over artificial intelligence, is prone to bias that will affect the quality of prediction. Other challenges include difficulties in forecasting long term values, undefined assumptions and identifying the right experience and knowledge sufficient enough to make the prediction. Nonetheless, the method is highly useful when there in an absence of historical data since the method is not solely dependent on data.
Quantitative data analysis is the second form of forecasting. Quantitative forecasting uses analytics to analyze historical data to determine the future trends of the demand. Therefore, unlike judgmental forecasting, this form of forecasting cannot function without historical data therefore; this is its key weakness. However, this method does not suffer from bias, make a projection for longer periods and only requires technical skills to make the projections (Hyndman & Athanasopoulos, 2014). In addition, the approach is effective in the identification of patterns and also supports data manipulations thus creating improved understanding of the trend. The method also supports sensitivity analysis and scenario testing thus improving the scope of understanding of the firm’s demand trajectory.
The two methods have their strengths and their weakness. Therefore, in order to leverage on the strengths and reduce the weakness of each of the category, it is highly encouraged that a forecaster applies the combination of the two categories (Mahendra, 2012). The combination yields the best results. By making reasonable forecast, a business will manage to optimize on its resources by keeping optimal inventory, keeping the right number of workers, reduce holding costs of finished goods, and minimize spoilt goods due to expiry among other benefits
Types of Forecasting Methods
Judgement Category
The category does not have any predetermined methods that can be applied. However, all that is called is critical reasoning and objectivity in order to make meaningful projections.
Quantitative Category
Under this category, there are numerous methods that have been developed, tested and applied and have been found to be functioning. Several select methods are outlined. First, Naive forecasting is a form of forecasting that assumes that the last period observed demand. Second, Seasonal Naïve Method is similar to the naïve method. However, in this case, the last observation is made from a similar period in the previous year. The third method is Drift Method. The method follows the Naïve method approach in making its projections. However, this method will differ from the naive method since it will include a drift. A drift is a mathematical term that will allow the forecast to vary over time where the change (Drift) is the average change over time observed in the historical data available (Mahendra, 2012). The fourth method is Regression analysis which is a method that seeks to determine the line of best fit based on historical values. After the regression line has been derived, it is possible to make a prediction to the future since the line can be extrapolated into the future. Monte Carlo Simulation is the last example. The method examines the historical occurrences to project the future by considering the probability of occurrence of an event while the future projection (simulation) is based on a random number.
The case applies exponential smoothing to predict the values in the third year. In order to make the projections using exponential smoothing, the following formula is applied. y^T+1|T=αyT+α(1−α)yT−1+α(1−α)2yT−2+⋯
Where YT is the observed actual demand in time T
α is the smoothing parameter: In this case it was taken to be 0.5
Exponential smoothing, commonly referred to simple exponential smoothing is one of the best quantitative tools to use when forecasting data that does not have a trend or seasonal patterns. In this case, the data being used is not large enough to observe neither the seasonality nor the trend thus making it suitable for the forecast.
The paper uses the component form of the method. Under this form, the initiation value needs to be set. In this case, the first forecast value is the same as the actual value of demand that corresponds to the first forecast. Therefore, this value will have a bearing in all future forecast. The stated formula is applied to obtain the following table.
Examining the actual sales and the forecasts, it is evident that only service A exceeds its forecasted values in all quarters. In this case, it shows that the firm promotion activities are not driving sales at a higher rate. However, there is a general trend of increase in actual sales and forecasted values. Therefore, the firm should not lay-off workers since the demand is gradually rising. Also, they should not recruit since the growth in the third year will not be large to warrant additional workers. Nonetheless, considering the firm is struggling to meet its targets, Freddie should be concerned about his promotion activities. In addition, Freddie should be concerned about his company ability to handle competition. These are the two fundamental reasons that would make a firm fail to meet its target demand assuming the services are up to customers’ expectations.
Conclusion
The essay has demonstrated that demand forecasting is an extremely important element in business operation management. The paper has further showed the two major categories that guide forecasting approaches and the methods that spring out of each category. Further, the paper has applied one method of forecasting to assist Freddie in making a business decision. These have demonstrated the manner in which forecasting can be used to add value to business decisions.
References
Green, K. C., & Armstrong, S. (2012, October). Demand forecasting; Evidence-Based Methods. Retrieved from http://www.kestencgreen.com/demandfor.pdf
Hyndman, R. J., & Athanasopoulos, G. (2014). Forecasting principles and practice. S.l.: OTexts.
Kavanah, S., & William, D. (2014, December). Making the Best Use of Judgmental Forecasting. Retrieved from http://www.gfoa.org/sites/default/files/GFR61508.pdf
Lawrence, M., Goodwin, P., O'connor, M., & Önkal, D. (2011). Judgmental forecasting: A review of progress over the last 25years. International Journal of Forecasting, 22(3), 493-518.
Mahendra, R. G. (2012). Industrial statistics and operational management: Forecast Techniques. Retrieved from http://nsdl.niscair.res.in/jspui/bitstream/123456789/829/1/CHAPTER-6 FORECASTING TECHNIQUES- Formatted.pdf
