Time-Series Forecasting Essay
Company demand describes estimated company sales at alternative levels of company marketing effort. It remains for management to choose one of the levels. The chosen level of marketing effort will produce an expected level of sales, called the company sales forecast (Goleman, 2001). Too often the sequential relationship between the company forecast and the company marketing plan is confused. One frequently hears that the company should develop its marketing plan on the basis of its sales forecast.
The forecast-to-plan sequence is valid if forecast means an estimate of national economic activity or if company demand is expansible, or where forecast means an estimate of company sales. The company sales forecast does not establish a basis for deciding what to spend on marketing; quite the contrary, the sales forecast is the result of an assumed marketing expenditure plan (Kuratko, et al. , 2001). Management sets sales quotas on the basis of the company forecast and the psychology of stimulating its achievement. Generally, sales quotas are set slightly higher than estimated sales to stretch the salesforce’s effort.
The sales budget considers the sales forecast and the need to avoid excessive risk. Sales budgets are generally set slightly lower than the company forecast (Sullivan, 2000).
Need essay sample on "Time-Series Forecasting"? We will write a custom essay sample specifically for you for only $ 13.90/page
It is found by fitting a straight or curved line through past sales (Sullivan, 2000). The second component, cycle, captures the wavelike movement of sales. Many sales are affected by swings in general economic activity, which tends to be somewhat periodic. The cyclical component can be useful in intermediate-range forecasting (Sullivan, 2000). The third component, season, refers to a consistent pattern of sales movements within the year. The term season broadly describes any recurrent hourly, weekly, monthly, or quarterly sales pattern.
The seasonal component may be related to weather factors, holidays, and trade customs. The season pattern provides a norm for forecasting short-range sales (Sullivan, 2000). The fourth component, erratic events, includes strikes, blizzards, fads, riots, fires, war scares, and other disturbances. These erratic components are by definition unpredictable and should be removed from past data to discern the more normal behavior of sales (Sullivan, 2000). Time-series analysis consists of decomposing the original series into these four components (trend, cycle, season, and erratic events).
Then these components are recombined to produce the sales forecast (Sullivan, 2000). For a company that has hundreds of items in its product line and wants to produce efficient and economical short-run forecasts, a newer time-series technique called exponential smoothing is available. In its simplest form, exponential smoothing requires only three pieces of information: this period’s actual sales, this period’s smoothed sales, and a smoothing parameter (Kuratko, et al. , 2001). The sales forecast is always between current sales and smoothed sales.
The relative influence of current and smoothed sales depends on the smoothing constant. Thus the sales forecast “tracks” actual sales (Sullivan, 2000). For each product, the company determines an initial level of smoothed sales and a smoothing constant. The initial level of smoothed sales can simply be average sales for the last few periods. The smoothing constant is derived by trial-and-error testing of different smoothing constants between 0 and 1 to find the constant that produces the best fit of past sales. The method can be refined to reflect seasonal and trend factors by adding two more constants (Sullivan, 2000).