Compare and contrast forecasting methods in operations management and explain how an organization uses these.
Forecasting is a prediction of what will occur in the future, but it is an uncertain process. The purpose of good forecasting, like the purpose of all commodity operations methods applied to service industry, is to control or reduce cost. Good forecasting, more often, identifies costs or cost-saving opportunities that might otherwise be overlooked. It permits anticipatory cost control. It must be competently practiced in all the qualitative and quantitative ways possible in service industries, and will be particularly important in managing commodity operations aspects of service ( 1992).
There are two basic categories for the methods of forecasting – qualitative methods and quantitative methods. Qualitative forecasting methods are mainly subjective, opinion-based and judgmental – they are none the less just as valid, and can often be combined with mathematical models (2002). Quantitative forecasting methods are objective methods that attempt to model the relationship between demand and other independent variables (2002).
Qualitative methods are used for long-term strategic planning. In general, long-range forecasting is undertaken by the top levels of management with the objective of predicting long-range trends and important turning points in the life cycle of our products. These demand forecasts are necessarily rather inexact and look only at very general, highly aggregated demands (1998). The method makes use of many techniques. This includes management judgment; expertise and opinion; market research based on surveys, demographics and focus groups; management, marketing, purchasing, and engineering knowledge of trends in their disciplines.
One of the most effective methods of qualitative forecasting is asking the people responsible for sales to estimate likely demand. Often, people such as sales staff, retail shop floor staff, service providers and the like are the first line of contact with the customer. They will have an excellent understanding of short-term demand (volumes, mix, type, etc.) and the likely future trends. Aggregated, such data is extremely valuable and is often overlooked in many organizations ( 2002).
Research concerning the needs and purchasing plans of both customers and potential customers is also a valuable source of information for qualitative forecasting. Information as to product and service advantages and disadvantages as well as possible modifications should be constantly monitored. In addition, this approach can also be used to compare the firm’s offerings with that of their competitors ( 2002).
The Delphi method is another method of qualitative forecasting: these are a group of methods that seek the contribution of a panel of experts ( 2002).A series of questionnaires is developed and sent to several anonymous experts. Questions are developed for experts to specifically to assess predictions of future events. The answers are compiled and sent to all the experts along with a new set of questions to refine their responses. The process goes through several rounds until some of level of agreement is reached about the likely events of the future. The results of the Delphi technique are generally mixed. The technique has the advantage of utilizing expertise and knowledge as to future trends and changes. They are frequently used by organizations attempting to construct futuristic scenarios that extrapolate the implications of major changes in society ( 2002).
Quantitative forecasting methods include time series methods and causal methods. This involves short-run decisions where there is a need for even more accurate and detailed forecasts of demand in order to implement production scheduling on a daily basis. Here lower-level operations managers are involved, and their forecasts are often based on actual orders already received. These managers must be able to revise their forecasts quickly and cheaply in order to act quickly in scheduling production, ordering materials, and so on (1998).
Time series methods are statistical techniques that make use of historical data accumulated over a period of time. Time series methods assume that what has occurred in the past will continue to occur in the future. Time series methods include moving averages, exponential smoothing, and linear trend lines. They are among the most popular methods used for short-range forecasting.
In the time series models, time is the independent variable and demand patterns are projected against this variable in an attempt to predict future trends. Time series involves the collection of historical data such as sales patterns. This data can be broken down into its main constituent components: trends, seasonal, cyclical and random ( 2002). For forecasting purposes, trend and seasonal demand data is the most frequently modeled and projected in order to predict demand (although there is always a large element of uncertainty when trying to accurately understand the future).
Causal models are another type of quantitative forecasting methods. These are also known as explanatory or associative models. The forecaster attempts to use another independent variable (apart from, or in addition to time) with which demand, the dependent variable, has shown a strong past relationship (2002). This model offers us leading indicators of future trends. For example, over time (independent variable) the levels of childbirth (independent variable) may be correlated or related to the demand for baby food (dependent variable).
In choosing a forecasting model a prime consideration is the accuracy of the model. However, other very important considerations are the cost of preparing the forecast and the ease of using the forecasting model ( 1998). Only for long-range forecasts that are made infrequently is a very costly forecasting justified. A general principle of using any quantitative technique is that the cost of using the technique must be less than the savings that can be expected to result from the presumably better decision made using that technique. Thus, as the time frame gets shorter and expected cost savings get smaller, only cheaper forecasting techniques make sense. Also, the frequent updating of shorter-range forecasts increases the necessity for quick, cost-efficient forecasting.
Credit:ivythesis.typepad.com
0 comments:
Post a Comment