Beruflich Dokumente
Kultur Dokumente
Although there are many different forecasting tools and methods, they can be divided into four general categories. 1. Judgemental methods- involve the collection of expert opinions 2. Market research methods - involve qualitative studies of consumer behaviour 3. Time-series methods mathematical methods in which future projection is extrapolated from past performance 4. Causal methods are mathematical methods in which forecasts are generated based on a variety of systems variables
A sales force composite can be assembled that combines each persons estimate in a logical way.
Market surveys involve gathering data from a variety of potential customers, typically through interviews, telephone based surveys, and written surveys.
Time-Series Methods
Time-series methods use a variety of past data to estimate the future data. There are a number of techniques, each of which has advantages and disadvantages. Below are some of the common time series methods: Moving Average Each forecast is the average of some numbers of past data points. The key is to select the number of points in the moving average so that the effect of irregularities is minimized. Exponential Smoothing Each forecast is a weighted average of the previous forecast and the last demand point. Thus this method is similar to the moving average, except that it is the weighted average of all past data points, with more recent points receiving more weight.
Causal Methods
In time-series methods, forecasts are based entirely on previous values of the data being predicted. In contrast, Causal methods generate forecasts based on the data, other than the data being predicted. In this, the forecast is a function of some other pieces of data. e.g the forecast for the next quarter may be a function of inflation, GNP, the unemployment rate, the interest rates, weather or anything besides the sales in previous quarters.