Forecasting (OM)

Planning is a fundamental activity of management. Forecasting forms the basis of planning. Be it planning for sales and marketing or production planning or manpower planning forecasts are extremely important.

What is forecasting?

Forecasting is a scientifically calculated guess. It is basic to all planning activity –

(1) whether it is national, regional, organizational, or functional planning; and
(2) whether it is long range plan or a short range plan.

My salesman looks out of the window and gives me the sales forecast for the next year said one senior manager. This salesman may be quite effective in his job, but he is only predicting and not forecasting. Forecasting is a little more scientific than looking into a crystal ball. The scientific basis of forecasting lies in studying past, present and future trends, present and future actions and their effects. What happened in the past is relevant to what is happening now and what could happen in the future. Thus, forecasting takes into account al the three dimensions of time — past, present and future. In spite of all the calculations, forecasting remains a calculate guess. Errors are bound to be there, but it remains the foundation of management planning.

One point needs a little emphasis. Many tend to think that forecasting is important only for marketing planning and not for production because the figures for production planning are received from marketing planning any way. This is erroneous view. Production planning need not necessarily follow marketing planning. There are many situations where production planning and marketing planning have to be done together and many other situations where production planning may be done separately from marketing planning. Therefore, forecasting is a very important activity for production planning be it strategic or tactical.

Elements of Forecasting:

Forecasting consists basically of analysis of the following elements:

Internal Factors

1. Past
2. Present
3. Proposed or future

External factors:

1.Controllable: (1) Past (2) Present (3) Future

2. Non-controllable: (1) Past (2) Present (3) Future

Forecasting is essentially the study of internal and external forces that shape demand and supply. The shape of the things to come will depend partially upon how one shapes the controllable factors. With different strategies, the forecasting will be different offering multiple scenarios for management decision making.

If previous data is all linear, there is no problem in extrapolation. But if the previous data shows upward swings as well as downward swings then how is one to extrapolate? Many a time the upwards and downwards swings quite random or one time effects. For instance, in one particular year, a company may have received a substantial government order which may not be repeated again for many years to come. It is, therefore better not to consider such one time or random phenomena for forecasting purposes. And that is why an average taken over a number of past years or time periods is more reliable as a forecast for the future year. This is what is done in the averaging techniques. One can have a continuous average or an average only over a certain number of years or periods in time. Both continuous and discrete averaging methods are used, which we discuss below.

Moving Averages Method:

The moving average is a discrete averaging method, where periods in the past beyond a certain number are considered irrelevant for the analysts. Suppose a company wants to use a 10 week moving average for forecasting sales of a particular item; they will add sales for the last 10 weeks and divide by 10 to get the average. A week later, they would add the newest weeks’ sales and discard the oldest so that once again they have a current total of the past 10 weeks of sales. Again this needs to be divided by 10 to get the new moving average.