Consider the two sets of data displayed in the upper part, Figure 1 represents the monthly demand for a replacement of high-cost. Note as the historical demand is composed of numbers with values which vary from 1 to 10 units in any given month. Figure 2 represents the precipitation in a city determined for the period from 2005 to 2008. Note that historically it has rained between 50 to 300 inches per year and the amount of rainfall can vary considerably from year to year. None of the data sets has trend or seasonality, for this reason these predictions will be horizontal lines.
This does not necessarily mean that the future demands in both cases will be a horizontal line; in both cases surely these amounts continue fluctuating in the future, the line itself only indicates that the quantities fluctuate the prognosis it cannot rely solely on the history of the data, in this case the confidence limits show the range within which these values will fluctuate. We review more in detail the behavior of the replacement of high-cost, the forecast corresponds to 3 units per month and the confidence limit superior is 97.5%, the actual value will be fluctuating in whole numbers between 0 and 7 in any given month. In this case the problem is that the timing and size of orders not be deduced from the past. Do then that tells us the forecast and how it can be used? In the statistical sense (according to model) the forecast corresponds to a constant value for the coming periods, however it is equally real considered that the value will fall between the upper limit and the limit lower if we try to predict our spare high-cost, therefore if we wish to estimate our income the more advisable is to take the prognosis and multiply it by the average selling price, however if what we want is to know the proper amount to have in inventory probably will be healthier using the upper confidence limit. Note that even when our data are extremely variable and our forecast is a horizontal line flat, the accuracy of the forecast and the confidence limits have a strong impact on our revenue planning and inventory policy, for this reason it is very important to predict what more aptly as possible, even when the prognosis is a flat horizontal line. Related articles: When and how they should adjust the forecast quantitative qualitative methods. Better as an input to the process of quantitative prognosis. Whats the importance of forecasting and planning of demand in times of crisis? Improving their forecasts with the modeling Top Down refining the Sales Operating Planning (SOP) author original and source of the article