Market research

Pigeons YE.P.

The determination of the instantaneous and forecast values of the demand of different form and indices of market portion for the concrete markets is the most important directions of market researches (market segments). In order to make data of questions sufficiently to object to expediently conduct the classification of markets and different it is specific market demand.

1. Classification of the markets

In the general economic plan by market is understood the place, where are gathered for the accomplishment of the report of purchase - sale both the salesmen and the buyers, all subjects of buying and selling the specific goods. In marketing usually by market is understood the totality of all potential users, who experience the need and having the capability it to satisfy for the goods of the specific branch.

Market is created around different objects, which present any value. In this plan they speak about the market for consumer goods, the labor market, the market for securities, the market for capital and so forth depending on the form of users are distinguished the following types of markets: consumer market and the markets for organizations or organizational markets. The latter are subdivided into the markets for the production of production-engineering designation, the markets for resales and the markets for offices of state.

Consumer market - totality of individuals and families, which buy goods and services for the personal consumption. The markets for consumer goods are characterized by mass user, diverse competition, decentralized structure.

The market for the production of production-engineering designation - totality of organizations and private individuals, that acquire goods and services, which are used with the production of other products. Key strategy of marketing of the production of production designation is system sale, in the implementation of which the buyer accomplishes system purchase.

System purchase - purchase of the package solution of problem for the purpose to avoid the purchases of separate components of this problem. For example, purchase by the government of weapon systems through the general contractor instead of the independent purchases of the separate components of data of systems separately. The collection of services usually also enters into system purchase.

Market for resales - totality of the organizations and individual persons, who acquire goods for the purpose of their resale or leasing.

Market for offices of state - offices of state of all levels (from the state to the local), the buying or leasing goods and service for fulfilling its functions.

In contrast to the consumer market the market for the production of production-engineering designation is characterized by a smaller number of buyers; however, buying production in a larger quantity. For example, the purchase of tire covers by auto-building companies.

Furthermore, the value of the purchases of the production of production-engineering designation is determined by demand for the end product, for example, to the passenger automobiles.

It is possible to isolate the following special features of organizational markets in comparison with the markets for the consumer goods:
They are more professional, especially relative to buyers.
In making of a decision about purchase, as a rule, participate several men.
Salesman and buyer to the larger degree depend on each other.
They attempt to establish long-term contacts.
Straight purchases much more frequently are used.
The much smaller role emotional factors play with the selection of purchase.

Many organizational markets are characterized by inelastic demand, i.e., demand weakly reacts to a price change. Hardly the factories of finished clothing greater will buy material with reduction of it in the price. In this case the volume of similar purchases is rather dictated by the value of demand for the finished production.

Depending on that, who prevails on the market, the latter is subdivided into the market for salesman and the market for buyer.

The market for salesman is characterized by the stronger position on it of salesmen in comparison with the buyers.

The market for buyer is characterized by the stronger position on it of buyers in comparison with the salesmen.

Depending on the degree of the involvement of user into the process of sales they separate: potential market; accessible market; the qualified accessible market; purposeful market; the mastered market.

Potential market - totality of users, who manifest a certain interest in the specific product.

Accessible market - group of users, who have interest, means and access to the specific product.

Qualified, accessible market - totality of users, who have interest, means, access to markets, and also satisfying legislative requirements, for example, to age limits for the driving of automobile.

Mastered market - totality of users, who already purchased some product.

It is obvious that with conducting of the segmentation of market it is necessary to allow these or other special features of it, to consider the uniqueness of the products, realized on the different markets.

2. Forms of market demand and its determination

At the basis of the process of the selection of purposeful markets lies the study of this base index as market demand. Market demand - this is the total volume of sales on the specific market (quotient or joint) for the specific stamp of goods or totality of the stamps of goods during the specific period of time.

On the value of demand have an effect both the uncontrollable factors of environment and the marketingovye factors, which are the totality of the marketingovykh efforts, applied on the market by competitor firms.

Depending on the level of marketingovykh efforts are distinguished primary demand, market potential and current market demand.

Primary or unstimulated demand - summary demand for all stamps of this product, realized without the use of marketing (Fig. 1).

Fig. 1. Dependence of demand on the expenditures for marketing

This is the demand, which "rots" on the market even in the absence of marketingovoy activity. From the point of view of the influence of marketingovoy activity to the value of demand are separated two extreme types of the market: the expanding market and the nonexpanding market; the first - reacts to the application of tools of marketing, the second - does not react.

Market potential - this is the limit, which is approached market demand with the approximation of expenditures for marketing in the branch to such value, that their further increase no longer leads to an increase in the demand under specific conditions of environment. C by the specific assumptions as the market potential it is possible to examine the demand, which corresponds to its maximum value in the curve of the life cycle of some product for the stable market. In this case it is assumed that the competitor firms for maintaining the demand make maximally possible marketingovye efforts. The factors of environment have an essential effect on market potential. For example, the market potential of passenger automobiles in the period of the decrease in the economy is considerably less than in the period of its prosperity.

Furthermore, is separated the absolute potential of the market, which should be understood as the limit of market potential with the zero price. The usefulness of this concept in the fact that it makes it possible to estimate an order of magnitude of the economic possibilities, which open this market. So the absolute potential of the market for passenger automobiles can be determined by total population, beginning from the age of obtaining driver rights. It is obvious that there is a large break between the absolute potential of market and the market potential. The evolution of absolute market potential is caused by such environmental factors as the income level and prices, the habit of users, cultural values, government control and by the like the given factors, to which the enterprise does not have real action, they can have the decisive effect on the development of market. Sometimes enterprises can have indirect impact on these environmental factors (for example, via the lobbying of reduction in the age of obtaining driver rights), but these possibilities are limited. Therefore the basic efforts of enterprises are directed toward the foresight of a change in the environment.

Further is separated the current market demand, which characterizes the volume of sales during the specific period of the time under certain conditions of environment with the specific level of the use of tools of marketing by enterprises in the field.

By selective demand is understood the demand for the specific stamp of any goods; appearance and development of this demand is stimulated via the concentration of marketingovykh efforts in the sufficiently narrow direction.

Another important index, whose value must be determined and forecast, is the index of market portion. Market portion - this ratio of the volume of sales of the specific goods of this organization to the summary volume of sales of this goods, realized by all organizations, which act on this market. This index is key during the estimation of the competitiveness of organization. This position escapes from the following: if in organization above index of market portion, then it more greatly realizes product on this market, therefore, it more greatly produces this product, since the volume of output must correspond to the value of potential realization. If organization lets out more than product, then the prime cost of the unit of product in this organization, as a result of the action of the scale economic factor, according to which, the higher the volume of output, the lower the value of prime cost, it will be below in comparison with other competitors. Consequently, the positions of this organization in the competitive activity will be more preferable.

The indices of demand for a number of goods, whose markets are characterized by the limited number of suppliers (in the first turn oligopolisticheskiye markets, it yields to statistical analysis, since the information about the volumes of the sold production and the shown services in the most varied aspects is assembled and is published: for the international markets, the markets for the individual countries and regions, in the section of separate branches and enterprises. However, it is specific goods detailed, reliable statistical information it is absent for many. Therefore for determination and predicting the values of the demand and other market characteristics it is necessary to conduct the special marketingovye studies, whose content will be described below.

3. Approaches to the market exploration

Supplementing and developing that state aboved, it is possible to pridti to the conclusion that the markets for consumer goods and production of production-engineering designation are studied mainly on the basis of the use of three approaches: 1. by means of analysis of second information; 2. by the experiment of motivation and behavior of users; 3. via the analysis and realizable output.

Within the framework of the first approach are studied all documents, which are of interest for the enterprises, published by the statistical organs, different by ministries, by the Chamber of Commerces, by regional control elements, and being also been the result of special nemarketingovykh studies. This information is sufficiently cheap, even free, relatively easily attainable. It is obvious that the information can appear by that closed, incomplete, is insufficient detailed, for example, not to be represented in the sufficiently detailed standardly designated section. Therefore for obtaining its reliable results usually it is clearly insufficient.

Let us note taking into account the made reservations that as a result of studying the documents it is possible to obtain the common picture of import and export, structure of production, and also the structure of consumption on I see productions and to branches.

But the study of documents cannot give the operational information of the required quality; it gives only certain statistics about production and sale of the studied production. However, the nomenclature of production usually is too amalgamated; therefore this information now and then cannot be used for the needs of concrete enterprise.

The second approach to the market exploration assumes a study of motivation and behavior of users by conducting the special inspections: interviewing, interviews, filling of forms, i.e., here adapts both the methods of sociological of studies and the expert estimations. In the latter case as those interrogated the professional- experts, who deeply examine the problem, come out: the users, that influence the selection of salesman, the buyers, for whom the product has no secrets. Therefore the specialist, who conducts a similar study it must know well the appropriate market and tested product, possess information about the company- producers and the intermediary organizations.

It is here necessary to bear in mind, that the motivation of the purchases of the production of production-engineering designation appears rational, and to the substantially smaller degree emotional, in comparison with the purchase of consumer goods.

Sample in the case of studying the production of production-engineering designation covers not individual people, but enterprises. When questioning in the industrial sphere is carried out, then sample, as a rule, is small (exception compose the atomized markets, where the potential users are represented by a large quantity of enterprises), replacements are not possible: there are companies, which must be compulsorily interrogated, for example, important firms - leaders in their branch.

In the case of the atomized markets there is a great possibility of the formation of the sample of enterprises on the representative basis, after including in it the enterprises of different sizes and level of market activity.

With the inspection of the production of production-engineering designation it is necessary to consider that in the process of preparation and making of a decision about purchase participate different specialists and the leaders of enterprises, which form part of his purchasing center.

Besides the study of the opinions of users is studied also the opinion of leaders and specialists of intermediary, distrib'yutorskikh organizations, and also producer enterprises.

It is here important to consider one important circumstance. Demand for the production of production-engineering designation is derivative of demand for the final consumer goods, with production of which it is used. Therefore for forecast purposes it is necessary to study not only the existing and potential users of the production of production-engineering designation, but also also the market for the corresponding final goods.

In the industrial medium is very extended the distrust to the questioning, especially among those, activity of which can influence the obtained results. On the producer enterprises - these are workers, the workers of design bureau, possibly, the management of enterprise. This is explained by a whole series of the reasons: in the industrial medium often the preference is given to production (quantity, quality, prime cost), technical specialists, as a rule, do not have economic formation, they are not familiar with the specific character of marketing of the production of production-engineering designation. Technical specialists assume that they know everything about their production, its high quality is presented by them by the second-order condition in order to find buyer. Conducted studies do not give the such accurate quantitative results, for which they calculate. On the basis of data of studies it is sufficiently difficult to estimate the effectiveness of the planned actions on the market.

Study of the opinions of leaders and specialists of intermediary commercial organizations, obtaining they have the necessary information often even more complex problem than conducting studies on the manufacturing concerns. Here appear the problems of the retention of commercial secret and ethical plan, connected with the assignment to information about the production (attitude of users, the volumes of sales and the like.) firm- competitors. Much depends on the skill of marketologa "to acquire" the necessary information even under the unfavorable conditions.

Under the favorable conditions the procedure of similar inspections escapes from general systematic recommendations, articles it is previously the examined in this cycle.

With conducting of marketingovykh studies, especially when there is no possibility to obtain reliable quantitative information on basis of one of the approaches examined, should be used parallelel'no all three approaches. Eventual results (for example, the value of demand, the index of market portion) can present the average (weighted mean) estimations, obtained by different ways, also, from the different sources.

Questionnaire for studying the tendencies of the evolution of market is given below.

Questionnaire for evaluating the tendency of the evolution of the market
Which the capacity of global market (in the physical and monetary term)?
Which the capacity of separate market segments?
Which market portion on the global market and on the separate market segments?
What tendency of a change in the market and market segments does occur (increase, stagnation, the decrease)?
How average consumption per capita of population, family, client?
Is how level value of saturation of market?
Which the degree of the extent of outfitting of family or firm by goods of this form?
Which average life of goods?
What portion of sales is caused by demand for the replacement?
They do have sale seasonal structure?
Are such the goods- substitutes, which fulfill the same function?
Are such main innovations in the users, who can change their relation to this goods?

This, it goes without saying, incomplete enumeration of questions only indicates, information of what type should be found. If the discussion deals with the production of production-engineering designation, questions must concern not only direct users, but also end users.

4. Estimation of the current demand

The greatest practical value has a determination and a prognostication of the current market demand. In the general case the determination of the current market demand in the money calculation (Q) is accomplished according to the formula:

Q = n * q * p,

where n - number of buyers of this form of goods on the market as a whole or on the market for concrete region;
q - number of purchases of buyer during the period of time being investigated;
p - mean price of this goods.

In this base formula with its concrete definition under the concrete forms of goods one must take into account the complementary factors, which determine demand for these goods. Thus, for instance, with the determination of demand for the goods of prolonged use as a result of conducting marketingovykh studies it is necessary to estimate demand for the replacement. Following data are necessary for this:
the volume of the available in users park of the goods of prolonged use being investigated;
the distribution of this park with respect to the service life (taking into account the factors of physical, economic and psychological aging);
the rate of the replacement of goods;
the possibility of the appearance of new alternatives of replacement.

Demand for the replacement find in direct dependence on the size of park and period of the service of goods of prolonged use. The rate of replacement not necessarily coincides with the rate of curtailment of the period of the service, by which is understood the share of goods of prolonged use, that cease existence. Goods can become obsolete because its economic indices became unsatisfactory or because it simply left the mode.

In the general case the rate of curtailment of the period of service is connected with inverse dependence with the duration of this period. For example, if average life is equal to 12 years, the average rate of curtailment of this period comprises: (1:12) 100 = 8,3 %.

For example, in France [ 8 ] the real average life of the automobiles of ten years was ago 10-11 years. If one assumes that it will reach 12, years, then the rate of curtailment of the period of use will be equal to approximately 8%, which corresponds to demand for replacement on the order of 1,7 million machines. If we, on the contrary, consider that the average life will not exceed 9 years, then the rate of curtailment of the period of use will be of about 11,1%, which corresponds to demand for the replacement in the size of 2,1 million machines.

Some datum, necessary for evaluating the demand, for example, the data about the existing park of goods and its distribution dependent on age, can be obtained from the analysis of past sales. The necessary estimation of distribution with respect to the service lives can be found on the basis of a selective study of the owners of goods, for example those, which are occupied with the replacement of the article available to them. Can be also studied the reasons (factors) for replacement. The substantial part of sales of goods of prolonged use corresponds to demand for the replacement, especially in the economically developed countries, where the level of the extent of outfitting of housework by such goods is already sufficiently high, and increase in the population is insignificant.

Initial data and the methods of calculation of demand for the production of production-engineering designation, with exception of some differences, with which the discussion will deal below, in the essence are the same as and for the consumer goods.

The current market demand frequently they are determined on the basis of normative method. This method assumes the sequential decomposition of the potential of market up to the presence of the estimation of demand for the concrete goods or stamp on the basis of the use of a number of norms and longitudinal indices. As an example let us examine the case of firm, which sells the additive (production of production-engineering designation), intended for the application together with the reagents for the softening of water to boiler rooms. Since many enterprises do not thus far yet use this additive, it is necessary to estimate the current and possible potential of market, and also the real level of demand in the specific natural region. Calculation is produced as follows:

On the basis of current, normative and statistical information it was determined [ ya]ob"em of the consumption of water by all firms of the specific region, which have boiler rooms - 7 shch00000gl;
the standard of the expenditure of the resource of softening per the liter of water 1%;
the portion of firms, which use this means: 72%;
the standard of the expenditure of additive per the liter of the means: 9%;

The possible potential of the market is calculated:

7500000 main X of 0,01 X of 0,72 X of 0,09 = 486000 l;

Studies showed, the portion of firms, which already use additive, was equal to 54%;

The summary current market demand is determined on the basis of these data:

7500000 main X of 0,01 X of 0,72 X of 0,09 X of 0,54 = 262000 l.

If the purpose of firm in attain the market share 40%, sale of goods in this region (current market demand for the firm) they must be brought to 105000l.

The difficulty of this method, obviously, consists of the presence of the corresponding norms and longitudinal indices. Their obtaining usually requires conducting special studies. It is at the same time evident that the error in each coefficient is transferred to each following level and to the total result. In order to avoid this danger, should be used several probable values, i.e., to obtain not one estimation, but their range. In any event this method should be used together with other analytical methods.

The more deep analysis of demand is directed toward the detection of the most important real factors, which influence the volume of sales, and toward the determination of their relative influence; such factors, as prices, income level, structure of users and influence of different methods of the advance of product most frequently are analyzed. With conducting of this analysis the methods of mathematical statistics widely are used.

So with the determination of demand the method of the leading indicators can be used. The leading indicators - indices or their time series, that are changed in the same direction as the index being investigated, but anticipating it on the time, for example, an increase in the indices of standard of living anticipates the index of an increase in the demand. Thus, studying the dynamics of a change in the indices of standard of living, it is possible to make conclusions about a possible change in the index of demand for the specific production.

During the estimation of the market potential of the regions or countries the indicators of purchasing power frequently are used. Purpose in this case consists of attractiveness measurement of market in terms of the weighted mean value of three key components of any potential of market, i.e.,:

- quantity of consuming units,

- the purchasing power of these consuming units,

- the readiness of these consuming units for the expenditures.

The statistical indicators of these three variables are determined for the selected territorial base (country, region, region, city), after which weighted mean index for each zone is determined. There are two approaches to its definition: to use the standard purchasing power index (IPS), which propose the firms for market exploration, or to build index is special for the analyzed sector or the gamma of goods.

The approaches presented are used also during the prognostication of demand. For these purposes in the model of demand the forecast estimations of its separate parameters are included.

5. General characteristic of the methods of the prognostication

The problem of prognostication, as a result of the rapid, now and then badly predictable changes in the environment, in the last decade became especially complex. Taking into account of these difficulties and criticality of errors in forecasts some specialists were forced to address about the futility of prognostication. In reality prognostication - this is the responsibility, which in the explicit or implicit form unavoidably they must carry out all organizations.

Besides obtaining of the possible future estimations of various parameters being investigated, the purpose of prognostication is also the motive to the reflection about that which can occur in the environment and this it will give to what consequences for the firm. Prognostication increases the vigilance of managers and, therefore, their ability to react to the changes. This effect is achieved even when plan is not fulfilled in connection with the fact that some hypotheses, assumed as the basis of forecast scenario, did not materialize.

The methods of prognostication as all methods, utilized with conducting of the marketingovykh studies, it is possible to classify to the heuristic, during application of which predominate subjective beginnings, also, to the economic and mathematical methods, during application of which predominate the objective beginnings, number of which includes statistical methods.

Heuristic methods assume that the approaches, utilized for the formation of forecast, are not presented in the explicit form and are not separated from the person, who makes the forecast, with development of which prevail the intuition, previous experience, creation and imagination. The methods of sociological of studies and expert methods are placed in this category of methods. That interrogated, giving their estimations, can base their judgments both on the naked intuition and using the specific cause-effect connections, data of statistics and calculations.

So during the prognostication of demand the preferences of users are studied; as the experts it can be examined commercial personnel, which operates the specific territories, dealers, distrib'yutery, marketing advisors, etc.

With the use of economic and mathematical methods the approaches to the prognostication are clearly formulated and can be reproduced by the others, who unavoidably will arrive at obtaining of the same forecast.

If during the application of expert methods the structure of cause-effect connections, utilized by different experts, can be different, then with the use of economic and mathematical methods the structure of models will be established and is checked experimentally, under the conditions, which are yielded to objective observation and measurement.

Determination of the system of factors and cause-effect (kazual'noy) structure of the phenomenon - starting point of economic and mathematical simulation being investigated.

In reality all these methods are mutually complementing. Effective forecast system must ensure the possibility of using any of these methods.

An example of complex forecasting problem, which is not solved with the aid of some method, is the prognostication of the volume of sales of new goods. With conducting of marketingovykh studies are evaluated the volumes of sales of new goods for the first years (let us say three) after release. For this purpose can be used expert methods, methods of interrogations, conducting sales on the control market.

The expert estimations, formulated by specialists in marketing, be based on the information, assembled at the stage of preliminary analysis and of the considering the data about sales competitors, size of potential market, overall demand, fractions of the products of different stamps on the market, accessibility of marketable networks, etc.

The missing information is assembled via the straight interrogations of potential users, merchants, suppliers and, if this is possible, competitors.

Checking market, or control sale, in the course of which is observed the real market behavior of buyers, makes it possible to estimate the level of trial and repeated purchases and the volume of potential sales of new goods. It is possible to also conduct trial sales through the place of residence or experiments in the special laboratory- stores.

Given methods are commonly used together. Using any of that enumerated or any other approach, the service of marketing must establish the promising volume of sales of new goods, on basis of which are developed strategies of the starting of goods.

It is clear that under the conditions of strongly variable environment the intuition and imagination are capable of becoming the important tools of the perception of reality, supplementing the quantitative approaches, which, regarding, rest only on the observed factors. From other side, it is understandable that the significant errors are also inherent in the purely qualitative method of determination and that the intuition must to the largest possible degree to be checked with the aid of the accessible facts and the knowledge. Thus, should be ensured the joint use of these two approaches.

As far as the prognostication of demand is concerned, in the methodological correct setting - this skill of the estimation of future demand under the assumption about the specific behavior of buyers under the assigned conditions. The prognostication of demand in this case must is accomplished into three stages. The forecast of environment at first is developed, then - forecast of the development of this branch finally the forecast of the value of demand for the goods of concrete company is developed. Such complex, that more analytical models, to develop and to realize is extremely complicated; therefore in practice simpler statistical models obtained application.

All forecasts of the volume of sales are constructed on use of three it is specific information, obtained on the basis of the study: what the people say that people make and that the people made. Obtaining the first form of information is based on the study of the opinion of users and buyers, commercial agents and mediators. The methods of sociological of studies and expert methods here are used. That people make the study of that it assumes conducting the testing market. The study of the fact that people made, assumes the analysis of the statistical data about the purchases made by them.

Usually in this case the discussion deals with the prognostication on the basis of statistical data by the volume of sales for the concrete company or the concrete market for the value of the current market demand for the specific goods. In the literature, in which are given the results of using various statistical models, very frequently is not done the differences between different forms of demand, and by its straight means they identify with the volume of sales.

5. Prognostication of demand, based on the methods of the mathematical statistics

It is possible to isolate two methods of developing the forecasts, based on the methods of the mathematical statistics,: extrapolation and simulation.

In the first case as the forecasting base is used the past experience, which prolongs to the future. Is done the assumption that the system is developed evolutionarily under the sufficiently stable conditions. The larger the system, all the more is probable the retention of its parameters without the change, of course, for the period of not too great. Usually it is recommended so that the period of forecast would not exceed one third of duration of initial temporary base.

In the second case is constructed the forecast model, which characterizes the dependence of the studied parameter on the number of factors, it influencing. It connects the conditions, which, as it is expected, will occur and the nature of their influence on the studied parameter.

Data of model do not use functional dependences; they are based only on the statistical interrelations.

Paired and plural regression analysis most frequently is used during the construction of forecast models; at the basis of extrapolation methods lies the time-series analysis.

Paired regression analysis is based on the use of an equation of the straight line:

y = a + bx,

where y - evaluated or forecasted dependent variable (successful sign);
a - absolute term of equation;
x - independent variable (factor sign), utilized for determining the dependent variable.
b - regression coefficient, which measures the average ratio of the deviation of successful sign from its average value to the deviation of factor sign from its average value of one unit of its measurement - variation y, which falls per unit of variation x.

Coefficients a and b are calculated on the basis of the observations of values y and x with the aid of the method of least squares [ 3 ].

Let us assume that the commercial agent sells children's toys, attending apartments randomly. The absence of visit some of apartment indicates the absence of sale or a = 0. If on the average each tenth visit is accompanied by sale for 62 dollars, then the cost of sale for one visit will compose 6,2 dollars or b = 6,2.

Then y = 0 + 6.2x.

Thus, it is possible to expect that with 100 visits the income will be 620 dollars. It is necessary to remember that this estimation is not required, but bears probabilistic nature.

Analysis on the basis of multiple regression is based on the use more than one independent variable in the equation regression. This complicates analysis, making with its multidimensional. However, regression model more fully reflects reality, since in the reality the parameter being investigated, as a rule, depends on many factors.

Thus, for instance, during the prognostication of demand are identified the factors, which determine demand, are determined interrelations existing between them, and their probable future values are forecast; from them with the condition of the realization of the conditions, for which the equation of multiple regression remains valid, the forecast value of demand is derived.

All concerning multiple regression it is conceptually identical to paired regression, with exception of the fact that is used more than one variable. The terminology and statistical calculations slightly changes from this point of view.

The multifactor equation of multiple regression takes the following form:

y = a + b1 x 1 + b2x2 + b3 x3 +.... + bm of the xm,

where y - dependent or forecasted variable;
the xi - independent variable;
a - absolute term of equation;
bi - coefficient of conditional net regression;
i = 1, m;
m - number of independent variables (factor it is indicative).

Term "coefficient of conditional net regression" means that each of the values b measures mean on the totality of the dependent variable (successful sign) from its average value with the deviation of dependent variable (factor) x from their average value of the unit of its measurement and when all other factors, entering the equation of regression, are fixed on the average values, they do not change, they do not vary.

By the limitation of prognostication on the basis of regression equation, as those more paired, serves the condition of stability or at least small changeability of other factors and conditions for the studied process, not connected with them. If sharply changes the "environment" of the elapsing process, the previous equation of the regression of successful sign to the factor will lose its value.

One should soblyudat' one additional limitation: it is not possible to substitute the values of factor sign, which considerably differ from entering the basic information, according to which is calculated the equation of regression. With qualitatively different levels of factor, if they even are possible in principle, would be different the parameters of equation. It is possible to recommend with the definition of the values of factors not to exceed limits of third of spread of variation both for the minimum and for the maximum value of sign- factor, existing in the initial information.

The forecast, obtained by substitution into the equation of the regression of the expected value of factor, is called point forecast. The probability of the precise realization of this forecast is extremely small. It is necessary to accompany its value by the mean error of forecast or by the confidence interval of the forecast, into which with the sufficiently great probability fall forecast estimations. Mean error is the modulus of precision of forecast on the basis of the equation of regression. There are usovershenstvovannye methods of paired regression, into some of degree its preodolevayushchiye deficiencies [ 2 ], [ 3 ].

The simplest methods of predicting the demand on the basis of statistical marketingovoy information are the extrapolation methods, based at the time-series analysis.

Many data of marketingovykh of studies are represented for different time intervals, for example, on the yearly, monthly and other basis. Such data are called time series. Time-series analysis it is directed toward development of three it is specific laws governing given change: trends, cyclic recurrence and seasonality, the development of the reasons for a change of the demand in the past with the subsequent transfer of the obtained regularities to the future.

Trend characterizes the general tendency in changes in the indices of a number. These or other qualitative properties of development express different equations of the trends: linear, parabolic, exponential, logarithmic, logistic, etc. after theoretical studies of the special features of the different forms of trend it is necessary to turn to the actual time series, especially as far from always it is possible reliably to establish, what the form of trend from the purely theoretical considerations must be. On an actual dynamic number the type of trend is established on basis of graphic representation, via the averaging of the indices of dynamics, on the basis of statistical testing of hypothesis about the constancy of the parameter of trend.

In tabl.y are cited the data of the volume of sales of the bicycles of the specific company in 17 years.

Table 1

Volume of sale of the bicycles

Yr

Annual volume of sales (into 1000 dollars)

1

1340

2

1221

3

909

4

1501

5

1350

6

1253

7

1561

8

1435

9

1114

10

1239

11

1453

12

1890

13

2220

14

2450

15

2790

16

3450

17

3759

18

????

It is necessary to determine the forecast estimation of the volume of sales for eighteenth yr.

After representing in the graphic form data of table 1, it is possible with the aid of the method of least squares to select the straight line, to the greatest degree which corresponds to obtained data (ris.2) and to determine the forecast value of the volume of sales.

At the same time the more careful examination of ris.2 makes it possible to make the conclusion that not all points are closely spaced to the straight line. Especially these divergences are great for the last years, and to believe to last data, apparently, follows with the large base.

In this case it is possible to use the method of exponential smoothing, assigning the different weight coefficients (large for the last years) by data for the different years [ 3], [' ]. In the latter case forecast estimation to the larger degree corresponds to the tendencies of the last years.

The cyclic nature of the fluctuations of statistical indices is characterized by the continuance (solar activity, the yield of separate crops, economic activity). Such phenomena are not usually the object of the experiment of marketologov, which the dynamics of problem in the relatively short time interval usually interests.

Seasonal variations of indices have regular nature and are observed during each year. They are the object of the study of marketologov (demand for the lawn mowers, for leisure in the health resort places during the year, to the telephone services during the twenty-four hours, etc.). Since the revealed regularities bear regular nature, completely soundly it is possible to use them for forecast purposes.

In contrast to the forecast on the basis of regression equation the forecast for the trend considers the factors of development only implicitly, and this does not make it possible "to lose" different alternative forecasts with the different possible values of the factors, which influence the studied sign. Then forecast on the trend covers all factors, while into the regression model at best it cannot be included explicitly more than 10-20 factors.

Time series besides the simple extrapolation can be used also for purposes of the deeper forecast analysis, for example, of the volume of sales. The purpose of analysis in this case is the decomposition of time series of sales for the main components, the measurement of the evolution of each component in the past and its extrapolation to the future. At the basis of method lies the idea of the stability of cause-effect connections and the regularity of the evolution of the factors of environment, which makes the use of extrapolation possible. Method consists of the decomposition of time series into five components:
structural component, or the long-run trend, usually connected with the life cycle of goods on the market being investigated;
the cyclic component, which corresponds to fluctuations relative to the long-run trend under the action of the intermediate-term fluctuations of economic activity;
seasonal component, or the short term periodic fluctuations, caused by different reasons (climate, sociopsychological factors, the structure of holidays so forth.);
the marketingovaya component, connected with the actions on the advance of goods, by temporary reductions in the prices, etc;
the random component, which reflects the joint action of the badly studied processes, not predstavimykh in the quantitative form.

The parameter, based on the observed regularities, is calculated for each component: the long-term rate of increase in sales, conjuncture fluctuations, seasonal coefficients, specific factors (demonstration, measure for the stimulation of sale, etc.). Then these parameters are used for the composition of forecast.

It is understandable that this forecast makes sense as short term, for the period, with respect to which it is possible to consider that the characteristics of the studied phenomenon substantially do not change. This requirement frequently proves to be realistic as a result of a sufficient inertness of environment.

To the number of main limitations of extrapolation methods should be carried the following.

The majority of forecast errors is connected with the fact that it was implied at the moment of the formulation of forecast in the more or less explicit form that the existing tendencies will be preserved in the future, which rarely is justified in the real economic and public life. Extrapolation methods do not make it possible to actually "predict" the evolution of demand, since are incapable to foresee any "rotary points". At best they are capable of rapidly considering the already occurred change. Therefore they are called "adaptive forecast model". Nevertheless for many problems of control this "a posteriori" forecast proves to be useful when there is sufficiently time for the adaptation and the factors, which determine the level of sales, are immune to abrupt changes.

So in the 40's of our century American specialists predicted: the production of passenger automobiles in THE USA will reach saturation and it will compose 300 000 pieces per month. But as long ago as their 1969 in THE USA were produced more than 550 000 pieces. At present this number grew 1,2 more - 1,3.

In 1983-1984 for the American market 67 new models of personal computers were introduced, and the majority of firms relied on an explosive increase in this market. According to the forecasts, which gave at that time marketingovye firms, the number of established computers in 1988 it had to be from 27 to 28 millions. However, toward the end of 1986 15 million were set only, since the conditions of using the computers radically changed, but this no one foresaw.

These errors in forecasts bore not the mathematical, but purely logical nature: indeed during the prognostication were used the time series, which sufficiently well reflect statistical material available at that time.

The development of society is determined by the very large number of factors. These factors are strongly connect together and far from all they yield to direct measurement. Furthermore, in proportion to the development of society now and then unexpectedly begin to enter into the action newer and newer factors, which earlier were not considered.

Time series can become unreliable basis for developing the forecasts as the economy acquires ever more international nature and increasingly it undergoes large technological reconstruction. In connection with this it is necessary first of all to develop the abilities of foresight, which implies a good knowledge of key factors and the evaluation of the sensitivity of organization to the external threats.

Name aboved not in koye to measure diminishes the significances of extrapolation methods in the prognostication. Like any methods they must be known how to use. First of all extrapolation methods should be used for the relatively short term prognostication of the development of the sufficiently stable, well studied processes. The forecast period of time must not exceed 25-30% of the initial temporary base. With the use of equations of regression forecast calculations should be conducted for the optimistic and pessimistic estimations of the initial parameters (independent variables), obtaining thus the optimistic and pessimistic estimations of the forecasted parameter. Real forecast estimation must be located between them.

In a number of cases the forecast estimation, obtained on the basis of extrapolation methods, uses as an indicator of the desirability of obtaining the specific value of the forecasted parameter. Let us assume that the forecast estimate of the magnitude of demand for some goods was obtained. She speaks, that with the same conditions of environment, structure and effective force of initial factors the value of demand up to specific moment of time will reach certain value. The managers, who use results predicted data, one should answer the question: "A does arrange us given the value of demand?" If "yes", then it is necessary to apply the maximum efforts in order everything to preserve without the change. If "no", then it is necessary to use internal possibilities (for example, to conduct additional advertising company) and to try to influence the specific factors of the environment, which are yielded to the indirect effect (for example, to influence the activity of mediators, to prolobbirovat' a change in the specific tariffs, the import duties). This entire activity is directed toward the guarantee of obtaining the desired value of demand.

During the prognostication of demand can be used the method of the leading indicators and the indicators of the purchasing power, described earlier, and also curved of life cycle.

In the latter case as the potential of market the market demand of the goods, which entered the phase of the maturity of its life cycle, is examined. The application of a model of the life cycle of product implies the ability to formulate the forecasts of qualitative or quantitative nature relative to the evolution of initial demand on the market for the specific goods. These data are determined by expert evaluation or on the basis of the study of the statistics of a change in the volume of sales on the time. In the latter case it is possible to use S- descriptive curves. By the most known curves of this type is logistic curve [ 5 ].

7. Brief characteristic of the expert methods

The methods of expert estimations are used for predicting the events of future, if statistical data be absent or it is insufficient them. They also adapt for the quantitative measurement of such events, for which there does not exist other method of measurements, for example, during the estimation of the importance of purposes and preferability of the separate methods of advance. In other words, the methods of expert estimations adapt both for the quantitative measurement of events in the present and for purposes of prognostication.

When the discussion deals with the prognostication of the volume of sales, then usually commercial personnel and personnel of commercial intermediary organizations have sufficiently precise idea about the potential of sales, which can ensure their clients, and, furthermore, has the possibility to give the estimation of the potential of market as a whole, at least in that territory, which he attends. It is simplest to ask commercial workers to give estimations for each goods, but not it is abstract, but on the basis of the concrete hypotheses about the marketingovykh efforts in questions of prices, advertising support and the like after this the managers of the service of sale formulate total estimations, summarizing the estimations of all experts.

To include commercial workers in the process of prognostication usefully, first of all, in order to create the appropriate motivation in them and to contribute to the adoption of the quotas on sales assigned it. Furthermore, they are irreplaceable during the construction of the forecasts of sales in the very small segments, at the level of separate territory or individual client.

It is obvious that it is often difficult because of the absence of statistical and current information (especially being concerned activity of firm- competitors) to obtain the quantitative assessments of such indices, for example, as the indices of market portion and dynamics of a change in the volume of sales. In this case also can be used the expert estimations, which form the purely qualitative values of these indices (in the terms "above, at the same level, below" and so forth).

At the same time expert estimations have deficiencies. From one side, there are no guarantees, that the obtained estimations are in actuality reliable, while with another - are definite difficulties in conducting of the interrogation of experts and processing of obtained data. If the second deficiency relates to the surmountable difficulties, then the first has fundamental value. The existing methods of determining the authenticity of expert estimations are based on the assumption that in the case of the coordination of the actions of experts the authenticity of estimations is guaranteed. This in reality is not always correct, and it is possible to give the cases, when individual experts, not consonants with the opinion of majority, gave correct estimations.

Consequently, the unanimity of the majority of experts is not always the criterion of the authenticity of estimations. The need of the thorough selection of experts hence emerges. The fact is that with the consideration of many questions, especially not standard, for example, the prognostication of market situation under the unstable political and economic conditions, must participate the experts of high qualification. The forecasts, comprised by "average" experts, will be based at best on the traditional, customary estimations, whereas the highly skilled specialists will reveal and will be estimated the concealed factors.

During the presence of estimations by expert evaluation besides the error, vnosimoy by a deficiency in the information about the events and by the insufficient competence of experts, is possible the error in entirely different kind, caused by the interest of experts in the results of estimation, which compulsorily will affect their authenticity. The presence of this type of error can considerably distort estimations, in consequence of which it is necessary to provide the appropriate measures for eliminating the error.

For example, there is always a danger of the systematic understating of the estimations of potential demand from the side of the marketers, who are interested to have easily feasible plan with respect to sales, and at the end of the planned period to attain the significant exceeding of planned indices. Let us note the following versions of the correction of such estimations, the decreasing risk systematic errors.

To ask commercial workers to independently determine the degree of an error in their estimations. These data can be then used for refining the forecast.

It is further possible to correct the estimations of commercial workers with the aid of the regional manager according to sales, who can have wider view.

Possible to introduce the corrective coefficient, based on the calculation of errors in the past forecasts of each commercial worker.

Adapt both individual and group (collective) expert interrogations.

The number of group expert interrogations includes:
open discussion of the presented questions with the subsequent open or secret ballot;
the closed consideration with the subsequent secret ballot or the filling of the forms of expert interrogation;
free statements without the consideration and the voting.

Experience showed that the traditional methods of considering presented before the appraisal group of questions, which relate to the first type of expert interrogations, before reaching of the specific agreement between them or to the production of general position are not the best methods of using the appraisal group. These methods of consideration suffer the number of deficiencies, such, as the mutual influence of the opinions of experts and the unwillingness of participants in the consideration to reject the points of sight, previously expressed publicly. Therefore in practice with the preparation for the solutions by the wide circle of questions the second and third types of group expert estimations find increasing propagation.

The second type of group expert estimations can be subdivided into two categories: the expert interrogation, conducted into one stage via single-time filling of forms, and conducted into several stages via repeated filling of forms by experts for the purpose of the sequential refinement of estimations.

It is possible to isolate the following basic stages of the expert interrogations, conducted into one stage,:

A. podbor experts and the formation of expert groups.
B. formirovaniye questions and the composition of forms.
V. formirovaniye the rules of the determination of summary estimations on the basis of the estimations of individual experts.
G. rabot with the experts.
D. analiz working expert estimations.

The content of the separate stages of expert interrogations is opened in the book [ 1 ].

One of the best methods of using the judgments of experts is the method of "Delphi", which foresees conducting expert interrogation into several stages.

The creators of this method named its "Delphi" on the name of the Ancient Greek city of Delphi, which deserved its reputation by oracles, who were being occupied by the prediction of future.

In the method of "Delphi" is made the attempt to improve group approach to the solution of the problem of developing of forecast or estimation via the mutual criticism of the subjective views, voiced by individual specialists without the direct contacts between them and with the retention of the anonymity of opinions or lines of reasoning in defense of these opinions.

In one of the versions of this method direct consideration is substituted by the exchange of information and of opinions with the aid of the thoroughly developed questionnaires. To the participants they are turned with the request not only to express their opinions, but also to give their substantiation, but in each of the subsequent stages of interrogation the new and precise information in the expressed opinions, which is formed as a result of calculating the agreement of points in the previously executed stages of work, reveals by it. This process continues until advance in the direction of an increase in the agreement of the points of view becomes insignificant. The divergent points of view are fixed after this.

For understanding of the essence of the method of "Delphi" let us examine an example of the estimation of demand for some goods, with the conducting by which different points of sight and judgment can exist. We will solve this problem in the following sequence. First, let us ask separately each expert to give our estimation of demand s, then it will arrange the answers in the ascending order of the proposed values and let us determine quartiles Q, M and Q in such a way that each of four intervals, formed by these three points on the line of the values s, would contain one one fourth of the estimations. For 11 participants this will appear as follows:

In - the second, values Q1, M and Q3 (opinion of expert group) we report to participants in the interrogation, also, in such a case, when initial estimation falls outside the mezhdukvartal'noye value (Q1 - Q3), we request to reexamine it, and to also express our considerations, why answer must be below (or above) the values, determined by 75% of participants in the first stage.

C - the third, we transfer the results of the second stage (which, as a rule, have smaller divergence than in the first stage) in the generalized form to all participants in the interrogation, including here new quartiles and median. Furthermore, are fixed the substantiations of decrease or increase in the values, proposed to participants in the interrogation in the second stage (certainly, with their generalization and editing it remains the anonymity of the authors). After this, experts they request to examine new estimations and their substantiations, to express their doubt about their ponderability and to reexamine their previous estimation. In such a case, when the reviewed estimations fall out from the mezhdukvartal'nykh values, then the author they request to briefly present the reasons for his disagreement with the arguments, which would lead his estimation nearer to the value of median.

Finally, in the fourth stage the quartiles of the third distribution of answers and the counter-arguments, expressed in the third stage report to participants in the interrogation and request to again reexamine estimations. The median, obtained as a result of the fourth stage, is considered as the value of the group expert estimate of the magnitude of demand. The often required level of the consensus of experts is reached for two stages.

The method of the collective generation of ideas (according to the American terminology the method of a "brainstorm") relates to the third type of group expert estimations and is directed toward obtaining of a large quantity of ideas, including from those, who, possessing the sufficiently high degree of erudition, usually abstain from the statements.

With conducting of expert interrogation with the aid of the method of the collective generation of ideas the problem must be formulated in the basic terms with the isolation of a central question. Furthermore, are provided for the absence of any form of the criticism, which impedes the formulation of ideas, the free interpretation of ideas within the framework of this question, tendency toward obtaining of a maximum quantity of ideas, takeing into account the principle of an increase in the probability of useful proposals with an increase in their total quantity and finally the encouragement of different combinations of ideas and ways of their improvement.

The process of the advancement of ideas proceeds in a sense avalanche-type: the idea voiced by one of the members of group generates creative reaction in others. Studies of the effectiveness of the method of the collective generation of ideas showed that the group thinking is produced to 70% more valuable new ideas, than the sum of individual thinkings. Most productive are acknowledged groups into 10 - 15 people, although there are examples of use and more numerous groups - to 200 people.

Since the results of this method represent not the totality of the free statements, but the system of ideas, not one proposal personifies. The results of consideration are considered fruit of team work of entire group. This is completely regular. Indeed any idea, expressed at given moment by one of the participants in the interrogation, could already earlier "mentally belong" to its associate, that expects word. Furthermore, specific proposal can be directly prompted by idea, by subject someone by several minutes it is earlier. Taking into account the aspect indicated, to the examination one should not carry the problems, which affect whose- or priority in the region in question.

With the aid of the method of the collective generation of ideas it is possible to successfully solve many problems of marketingovykh studies, for example:
the determination of the possible ways of the development of the forecasted process or object, one of which is optimum; during the solution of similar problems this method adapts for the purpose of the determination of the complete set of the possible ways of development;
the determination of the widest circle of the methods used, if the solution of problem requires the parallel or sequential use of a number of diverse methods;
the development of the circle of the factors, which must be taken into consideration, determining the final version of the solution. For example, the value of demand.

The number of most important deficiencies in the method of the collective generation of ideas includes the significant level of information noise, created by trivial ideas, the spontaneous and spontaneous nature of the generation of ideas.