Demand Forecasting – Meaning, Scope, Types and Importance – Managerial Economics

Posted on Jun 4 2020 - 1:41pm by Preeti

2. Opinion Poll Methods

The opinion poll methods aim at collecting opinions of those who are supposed to possess knowledge of the market, such as sales representatives, sales executives, professional marketing experts and consultants. The opinion poll methods include:

(a) Expert-opinion method,
(b) Market studies experiments.

(a) Expert-Opinion Method
In this method, the opinion of experts is taken to forecast the demand for the product.

i. Single expert
The experts may be from the firm itself, for example, the sales managers who possess the sufficient knowledge about the market. Sales representatives, being in close touch with the consumers or users of goods, are supposed to know the future purchase plans of their customers, their reaction to the market changes, their response to the introduction of a new product and the demand for competing products. They are, therefore in a position to provide at least an approximate, if not accurate, estimate of likely demand for their firms product in their region or area. The estimates of demand thus obtained from different regions are added up to get the overall probable demand for a product. Alternatively, the experts may be from outside the firm, for example, some marketing experts and consultant firms who are specially trained for such tasks.

ii. Delphi Technique
Another method of collection opinion is by “Delphi technique”, originally developed by the and Corporation at the beginning of the Cold War, to forecast the impact of technology on warfare. It is an extension of the simple expert opinion poll method. Delhphi is a way of getting the opinion of experts without teir face to face interaction.

This method tries to arrive at a consensus by questioning a group of experts, repeatedly, until the responses appear to converge on one line. It is also possible to get disagreement but the causes of disagreement are supplied with responses of others to previous questionnaires coming from different respondents. The leader of the group, or co-ordinator, gives these responses of others alongwith reasons. The experts to whom these previous questionnaire-responses (with reasons) are given are requested to give their reactions. In exchanging opinions among experts, the names of earlier respondents are not exposed. This avoids bandwagon effect’ or ‘ego involvements’ in publicly announced opinions. The co-ordinator then arrives at a consensus or common opinion of the group or panel of experts, to serve as a forecast.

Advantages

(i) Since anonymity is throughout maintained, experts can express their views candidly.

(ii) It enables to obtain the opinions of several experts simultaneously without actually inviting them together for a conference. This saves time as well as money.

(iii) Another advantage is that a large number (at times, a few hundreds) of experts with diverse specialisations in areas like industrial policy, science and technology, administration, economics, accountancy, financial management, marketing management, etc. can be simultaneously approached. This brings in a balanced view.

Limitations

(i) The method assumes that the experts identified are rich in their own expertise, experience and knowledge. This may not always be possible for all firms, big and small.

(ii) The method also relies heavily on the group leader or co-ordinator who is expected to be objective in judgement, able to conceptualise issues for discussion and competent to analyse and draw inferences. Such co-ordinators may not be easily available or to get them may be a huge expenses.

(iii) In case of external experts are invited for opinion, the firm may be exposed to the risk of loss of confidential information to rival firms.

(b) Market studies experiments

An alternative method of collecting necessary information regarding demand is to carry c market studies and experiments on consumers behaviour under actual, though controlled market conditions.

i. Market Tests
This method is known in common parlance as market experiment method Under this method, firms first select some areas of the representative markets – three or four cities having similar features, that is, population, income levels, cultural and social background, occupational distribution, choices and preferences of consumers, then they carry out market experiments by changing prices, advertisement expenditure and other controllable variables in the demand function under the assumption that other thin remain the same. The controlled variables may be changed over time cither simultaneous in all the markets or in the selected markets. After such changes are introduced in the market, the consequent changes in the demand over a period of time (a week, a fortnight, or month) are recorded. On the basis of data collected, elasticity coefficients are computed. These coefficients are then used along with the variables of the demand function to assess the demand for the product.

ii. Laboratory Tests
Alternatively, market experiments can be replaced by consumer clinics or controlled laboratory experiment also known as ‘artificial market’ . Under this method, consumers are given some money to buy, in a stipulated store, goods with varying prices, packages, displays, etc. The experiment reveals the consumers’ responsiveness to the changes made in prices, packages and displays. Thus, the laboratory experiments also yield the same information as the market experiments. But the former has an advantage over the latter because of greater control over extraneous factors and its somewhat lower cost.
The Grabor- Granger test (named after the economists who invented it in 1960s) is popular technique of market simulation. Half of the members of a group of consumers are shown the new product and responses are sought on whether they would actually buy it at various prices on a random price list. Thereafter, they are shown the existing product. The other half is shown the existing product first and then the new product. This test can also be used to ascertain if a product would be bought at different prices; form the results the optimum price for each individual consumer can be estimated.

Limitations

i. They are very expensive. Therefore, small firms cannot afford experimental methods.
ii. Experiments may take considerable amount of time an dmoney to be conducted.
iii. It has been established that people behave differently when they are being observed hence their behavior may not be very reliable.
iv. A big disadvantage of experimental methods is that tinkering with price increases may cause a permanent loss of customers to competitive brands that might have been tried.
v. Changes in socio-economic conditions during the field experiments, such as local strikes or lay-offs, advertising programmes by competitors, political changes and natural calamities may invalidate the results.
vi. Experimental methods are based on short-term and controlled conditions that may not exist in an uncontrolled market. Hence, the results may not be applicable to the uncontrolled long-term conditions of the market.
vii. Being a costly affair, experiments are usually carried out on a scale too small to permit generalisation with a high degree of reliability.

Merits
i. Market experiment method is often used to provide an alternative estimate of demand and also “as a check on results obtained from statistical studies.
ii. This method generates elasticity coefficients, which arc necessary for statistical analysis of demand relationships.
iii. Market experiments often provide useful information on consumer behavior regarding a proposed change in any of the determinants of demand.
iv. Experiments are very useful in case of an absolutely new product of which consumers had never heard before.

B. Statistical Methods

Statistical methods of demand forecasting utilize historical (time-series) and cross-section data for preparing long-term demand forecasting. Statistical techniques are considered to be superior one for demand forecasting for the following reasons:

1. In the statistical methods, the element of subjectivity is minimum,
2. Method of estimation is scientific, as it is based on the theoretical relationship between the dependent and independent variables,
3. Estimates are relatively more reliable, and
4. Estimation involves smaller cost.

1. Trend Projection
2. Barometric Methods
3.  Econometric Methods

1.  Trend Projection
Trend is a general pattern of change in the long run. Often referred to as a “classical method”, trend projection is a powerful statistical tool that is frequently used to predict future values of a variable on the basis of time series data. Time series data are arrangement of the values of a variable in chronological order of days, weeks, months, quarters or years. The basic assumption in using time series is that past trend is likely to continue. GDP of a country, sales of m any companies, population of a country usually follow trend.

In most time series data we see phases of fluctuation in the values of the variable over time. This can be explained with the help of the following components of the time series data.

Time series data are composed of:

• Secular trend
• Seasonal trend
• Cyclical trend
• Random events

Secular Trend
This refers to change occurring consistently over a long time and is relatively smooth in its path. Sale of Personal Computers may show an increasing trend over the years, while demand for jute products may show a declining trend.

Seasonal Trend
This refers to seasonal variations of the data within a year, e.g.. demand for woolens, ice cream, raincoats are dependent upon weather and would vary during a period of one year.

Cyclical Trend
This refers to a cyclical movement in the demand for a product that may have a tendency to recur in a few years. These changes may push economic activity in expansion mode or recession.

Random Events
Typical examples of random events may be natural calamities, social unrest, foreign aggression creating war like situation etc., these events have no trend of occurrence hence they create random variation in the series.

These components of time series may be written in additive form or multiplicative form. In additive form it is assumed that each of these components acts independently.

Y = T + S + C + R                                                 (1)

or Y = T.S.C.R                                                       (2)

The multiplicative form can be rewritten as:
Log Y = log T + log S + log C + log R                (3)

where Y = value of time series, S = seasonal variation, T = secular trend value of the variable, C = cyclical variation. R = random variation.

Merits

• Trend projection method is simple to apply.

• It is reliable in forecasting demand.

Demerits

• The accuracy of trend projection method depends on the availability of time seriesdata; the longer the series, the better the results.

• There may be an element of subjectivity in trend projection by geometric method.

• This method cannot be applied to forecasting for new products where past data is not available or products which are launched recently where past data is not sufficient to project trend.

• It may not necessarily be (rue that past pattern would be continued in the future.

Methods under Trend Projection Method

(a) Graphical Method
In graphical method the past values of the variables are shown on the vertical axis and time on the horizontal axis. Th available data are then plotted on a graph and a line is drawn through all the points. Then a free hand line is so drawn that the total distance between the line and the points is minimum. The solid fluctuating line shows the actual trends, while the dotted line show the secular trend. By extending the trend lines, we can forecast an approximate sales in future. Thereafter, the movement of the series is assessed and future values of the variable are forecasted.
Lets Study it with the help of an example.

1999 Demand of steel (million tonnes)
2011 20
2012 22
2013 24
2014 21
2015 23
2016 25
2017 23

The orange line shows the actual data whereas the blue doted line is the trend line.

Merits
i. Simplest method of measuring trend.

Demerits
i. It provides a general indication and fails to predict future value of demand.
ii. It involves subjectivity and personal bias of the analyst.

(b) Least Squares Method
Least squares estimation is a powerful tool to estimate the coefficients of a linear function. It is based on the minimization of squared deviations between the best fitting line and the original observations given. In this method (also regarded as the algebraic method), we fit the data on demand and time in the form of equations and then project the demand for the future period.

(c) Box-Jenkins Methods
This method is also called Auto-regressive moving average (ARIMA) method. Suggested by G E P Box and G M Jenkins, this method of forecasting is used only for short¬term predictions. Besides, this method is suitable for forecasting demand with only stationary time-series sales data. Stationary time-series data is one, which docs not reveal a long-term trend. In other words, Box-Jenkins technique can be used only in those cases in which time-series analysis depicts monthly or seasonal variation recurring with some degree of regularity.
Steps in Box-Jenkins method: As mentioned above, Box-Jenkins method can be applied only to stationary time-series data. Therefore, the first step in Box-Jenkins approach is to eliminate trend from the time series data. Trend is eliminated by taking first differences of time-series data that is, subtracting observed value of one period from the observed value of the preceding year. After trend is eliminated, a stationary time-series is created. The second step in the Box-Jenkins approach is to make sure that there is seasonality in the stationary time-series. If a certain pattern is found to repeat over time, there is seasonality in the stationary time- series created. The third step involves use of models to predict the sales in the intended period.
Any Stationary time-series data can be analysed by the following three models:
(i) Auto-regression model
(ii) Moving average model and
(iii) Auto-regressive moving average model

(Visited 122 times, 2 visits today)

Pages: 1 2 3 4 5
About the Author

B.Tech Biotechnology,MBA(HR and Marketing), UGC/CBSE NET Qualified

Leave A Response