DEMAND FORECASTING

Contents
• Meaning & Definition.
• Purpose of Demand Forecasting.
• Steps Involved in Demand Forecasting.
• Determining Scope of Demand Forecasting
• Methods of Demand Forecasting.
• Statistical Method.
• Criteria Of A Good Demand Forecasting.
 Meaning & Definition
• Meaning of Demand Forecasting - Demand
forecasting for a product is the technique of
estimating its demand in the immediate or
distant future.
• Definition - According to Evan J. Douglas
“Demand forecasting will be taken to mean
the process finding the values for demand in
future time periods”.
 Purpose of Demand Forecasting
 
• Purpose of short term forecasting.
– To avoid overproduction and underproduction.
– Proper management of inventory.
– To setup reasonable sales target
– Formulating a suitable sales strategy – (changing pattern of demand)
– Forecasting financial requirements for the short period.
• Purposes of Long term Forecasting.
– Planning a new project, expansion and modernization of an existing unit.
– Assessing long-term financial needs.
– Arranging suitable manpower, (trained and skilled labor and business
executives.)
– Evolving a suitable strategy for changing pattern of consumption.
(Industrialisation, urbanisation, education.
 Steps Involved in Demand Forecasting.
• Setting the Objective. The firm should be clear as to the
purpose of demand forecasting in determining the rise of
output, fixation of price, allocation of funds for sales
promotion.
• Selection of Goods. Consumer goods and capital goods,
existing goods and new goods. Method of demand
forecasting will differ according to the nature of goods.
• Selection of Method. The scope and success of a
particular method depends upon the area of investigation,
resources – monetary, time available, availability of data,
trained personnel.
• Interpreting the Results. The results should be carefully
analysed before any inference is drawn.
 Determining Scope of Demand Forecasting
• Period Covered Under forecasting.
– Short-term forecasting. (Tastes and Preferences)
– Medium-term forecasting. (Quality improvement)
– Long-term forecasting. (Government’s fiscal policy)
• Levels of Forecasting.
– Macro-economic level. (Industrial Production, Wholesale
price index)
– Industrial level. (Trend of electronic industry - 3D)
– Firm level. (Position of competitors)
• Nature of Product. Non durable, durable consumer
goods, Capital goods
• Forecast of established or new product
 METHODS OF DEMAND FORECASTING.
• An efficient demand forecasting should strike a balance
between mathematical techniques and sound judgment.
• Opinion Poll Method.
– Consumer’s Survey Method. The representatives approach buyers
personally to know their intention for the likely purchases of product.
• Complete Enumeration Survey. The find probable demands of all the
consumers for the entire forecast period.
Advantage, first hand unbiased information.
Disadvantage, a costly proposition, because of personal privacy, they hesitate
to divulge.
• Sample survey. The find probable demand expressed by each selected unit. is
especially useful for forecasting sales of new products. But it is exceptionally
costly in both time and money. It is also difficult to select test area.
• End-use (or Input-Output) Survey. As the number of end-uses of a product
increases it becomes more and more inconvenient to use this method. Useful
for industries which are largely producers’ goods (aluminum).
– Collective Opinion Method/Sales force Opinion
Method.
Merit: collective wisdom of salesmen, (forecasting sales
of new products). Demerits: Personal bias, not foresee
the influence of several unknown factors, (Useful for
Short term forecasting).
– Experts’ Opinion Method/Delphi Technique. Views of
the specialist or experts is sought. Their interaction
continues until a best solution is found. Best suited in
situations where intractable changes are occurring.
(Advantages of speed and cheapness)
 STATISTICAL METHOD
In order to maintain objectivity and precision in demand forecasting, statistical techniques are used.

•  
– Trend Projection Method. Here we attempt to measure the past trend
and then on its basis, forecast the future trend. Trend can be measured
by using a number of techniques
• Graphic method. All values of output or sales for different years are plotted on a
graph and a smooth freehand curve is drawn in upward or downward trend.
Disadvantages: chances of bias on the part of investigator.
• Least Squares Method. This technique is most widely used. With its help, a trend
line is fitted to the data. It is also known as ‘the line of best fit’, extrapolating the
trend line for future gives corresponding figures of forecasted sales.
– Linear Trend. Straight line.
Y = a+b(X) where a and b are intercept and slope. For finding a & b, two normal equations
need to be solved. ∑Y = Na +b ∑ X and ∑XY = a ∑ X + b ∑ X2. N is time(Yrs, Mths).
• Time Series Analysis. Time series is composed of trend, seasonal fluctuations,
trade cycles, and natural calamities. If data is quarter-wise/month-wise, it is
possible to identify the seasonal effect. If data is available for long period, the
trend and cyclical effects can be found.
 • Method of Moving Averages. It consists of a series of arithmetic
means calculated and overlapping groups of successive values of a
time series.
• Exponential Smoothing. Used for short term forecasting,
determines values by computing exponentially weighted system.
More recent values are assigned greater weight than previous
period values. Weights (w) are so assigned that w lies between zero
and unity (0 < w < 1) i.e., St = wYt + (1-w) St-1
– Barometric Techniques. Based on the presumption that a
relationship can exist among various economic time series.
Three kinds of relationships.
• Leading series. Birth rate of children is the leading series for
demand of seats in schools.
• Coincident series. A series of data on national income coincident
with the series of employment.
• Lagging series. Industrial wages is lagging when compared to price
index for industrial workers.
– Regression Method. In regression a quantitative relationship
is established between dependent variable and independent
variables, i.e., determinants of demand (income & Price).

D = f(determinants of demand),

In particular Dx = a + b1Px + b2Y + b3T + b4Pr …

– Simultaneous Equation Method/Complete Systems

Approach. It involves simultaneous considerations of all
variables. The set of equations is made equal to the number
of dependent (controllable) variables. Once the model is
specified, it is then solved to find regression of sales on all
the pre-determined variables in the system. The reason for
unpopularity of this method is mainly, its complexity.
– Econometric Method – The Science of economic
measurement. It Study past economic activities by
driving mathematical equations.

– Controlled Market Experiment – Manipulated elements

study separately in controlled environment.

– Computer Assisted Forecasting –Well programmed

method used to predict future and keep track of
forecasting errors.
CRITERIA OF A GOOD DEMAND FORECASTING.
• Accuracy. Past data provides basis for present
performance and on the basis of the accuracy of present
results forecasting can be undertaken
• Economy. A firm has to strike a balance between the
benefits from increased accuracy and the extra cost of
providing the improved forecasting.
• Availability. Results should be readily available and well
understood.
• Timeliness. There is always a lead time gap between
occurrence of an event and its forecast. Known as ‘lead’
time Longer the lead, greater would be its usefulness.
• Simplicity. It should be simple rather than cumbersome
and intricate technique.