Abstract. The paper presents the results of an analysis of the practicability of recruiting personnel through social networks using the Vkontakte social networking site as an example. To that end a study of career behaviour as conditioned by demographic factors, with predicted length of employment for employees drawn from social networks, compared with labour market demand, was conducted. The paper also features a comparative analysis of labour market demand patterns and university specialist supply patterns.

Keywords: social networks, labour market, recruitment.
JEL Classification: J21


1. Introduction

Efficient use of the workforce is a vital part of a firm’s sustainable development, and recruiting the talent most suitable to an organisation’s tasks and corporate culture is the key factor in making the best use of its human resources. Recruitment errors result in increased losses from employee turnover: a mid-size US firm [1] loses more than 12% of its pre-tax profit to employee turnover costs. Direct costs incurred by replacing resignees amount to 50% to 60% an the employee’s annual salary, with total costs reaching up to 90% to 200%, depending on the position level.

Recruitment and workplace adaptation costs can be cut by using less resource-expensive recruitment techniques, like early candidate selection and carrying out active recruitment through social networking sites. The choice of a social network must be guided by its popularity and the degree to which its audience meets an organisation’s requirements.

These considerations call for a study of a typical non-specialised social network’s audience, made with a view to the subsequent recruiting of its users. The aim of the present study is to analyse the extent to which users of social networks could satisfy labour market demand in terms of job specialisation and predicted career behaviour demand.

2. Prior studies

The financial implications of recruiting personnel through social networks and the current trends in recruitment practices is the subject of interest for a wide range of scholars and business professionals, with a large number of studies devoted to it, [2, 3, 4]. [2] Investigates the possibility of using social networks for recruiting university graduates and increasing an organisation’s candidate pool; the paper’s key observation is that using social networks proves useful specifically for recruiting university students and young professionals. [3, 4] study the way social networks’ evolution and rise in popularity influences recruitment and talent attraction. All the studies agree in general that the rise of social networks makes attracting talent through greater employer brand development, previously a cost-intensive technique, increasingly useful.

From the point of view of recruitment, social network content remains largely unresearched due to its being poorly structured, manually entered by the users, and lacking classification. Because of the complexities of human language, automated processing of such content requires a sophisticated linguistic software tool, developing which to solve a one-off task is economically impractical; whereas manual analysis is time-intensive and prone to human error.

Nevertheless, there have been studies on analysing structured social network content, including those in the Russian-language Internet sector. Given the fact that Vkontakte is the largest social network in Russia (more than 220 mln users), its content is of greatest interest to researchers.

In terms of their technical implementation, the papers most interesting to the present research were studies of Vkontakte audience [5] and their priorities in life [6]. [5] employed search queries for the relevant parameters, registering the number of results. [6] made use of the network’s API to automate raw data retrieval. Both studies also include accounts of problems encountered by the authors, namely access to user data restricted by privacy settings, accounts featuring unconfirmed data, and the ways in which the latter can be identified. This research formed the basis for a specialised data retrieval software solution.

3. Methodology

To analyse the prospects of using social networks for recruitment purposes a two-part study was conducted, with a qualitative study of active Internet users’ career expectations done through polling, and a quantitative study of Vkontakte users’ job specialisation and career behaviour-affecting characteristics, as well as an evaluation of the degree to which social networks’ audience satisfies the labour market demand.

4.Qualitative study

Projected career behaviour was analysed through a qualitative opinion survey done via an online polling service [8]. The service featured automated direct mailing of potential respondents providing the kind of random sampling required to be able to extrapolate the results onto social network users in general.
The questionnaire contained questions about the respondent’s demographic characteristics, education, job specialisation, job experience, area of employment, expected length of employment, and vertical or horizontal career focus.

The main results were as follows:

The factors having the most significant influence on the respondents’ career behaviour were job experience, job level, age.

The relationship between length of employment and job experience is shown in Figure 1. The level character of the blue curve shows that for men job experience does not have a significant impact on their preferred length of employment. The red curve shows the impact of job experience on the dynamics of career behaviour among women: having gained job experience, women in general become less likely to change jobs often; however after 10 years in one job position the average preferred length of employment for women begins to fall, most probably owing to the number of executive posts coming to be occupied by women with job experience of over 10 years.

Figure 1. The relationship between job experience and preferred work duration for men and women 

Figure 1. The relationship between job experience and preferred work duration for men and womenFigure 1. The relationship between job experience and preferred work duration for men and women

Figure 2 shows the relationship between preferred length of employment and job level. For both men and women, lack of jobs is the factor that makes an average prospective employee’s career expectations closer to the expectations of specialist-level employees. Men in specialist positions are more disposed towards longer employment, owing to the horizontal career path, characteristic of this group; men in managerial positions are more oriented towards a vertical career path, which shortens their average preferred length of employment.

Female respondents in general showed little inclination towards a vertical career path, which explains the relationship between preferred length of employment and job level: most respondents viewed their career path culminating in a mid-level executive post, which, once obtained, they preferred to keep. All the same, some women aspire to further career advancement, with employee turnover among senior executives significantly higher than among mid-level executives. Low impact of job experience on low-level managerial positions can be explained by employee’s desire to move through the ranks, regarding their current positions as temporary posts they are forced to take up. 

Figure 2. The relationship between length of employment and job level for men and womenFigure 2. The relationship between length of employment and job level for men and women

Figure 3 shows the relationship between preferred length of employment and employee age. For women it is nearly linear: the older a female employee is, the less likely she is to change jobs. For men the relationship is less uniform: men under 25 tend to change jobs infrequently, owing to their need for gaining job experience and professional growth; 26 to 35 year-olds show the greatest degree of active vertical career advancement, with men over 36 feeling the need for greater stability because of greater family commitments.

Figure 3. The relationship between length of employment and employee age for men and womenFigure 3. The relationship between length of employment and employee age for men and women

5. Quantitative study

Since Vkontakte has the largest audience among Russian social networking sites, it was chosen as the subject of automated data retrieval for the quantitative study. A list of required data was prepared prior to retrieval, including user demographics (date of birth and sex), details of students’ or graduates’ university (name, faculty, department, teaching method, date of graduation). Selected user profiles were mined in accordance with the relevant data and ‘friend’ lists using the social network’s API.

Although Vkontakte offers developers a dedicated way of accessing its full user list, a different way of drawing the sample was developed to make the technique applicable to any social network. A new, universal algorithm was used to crawl user profiles checking their ‘friend’ lists for users it has not yet indexed by comparing them with the ones already present in the database. To make the algorithm less time-intensive and to get a more varied mix of users, only those with over 100 ‘friends’ were processed.

Mathematical statistics requires a study to be conducted using a sample of more than 2 000 people. However, drawing the sample from users’ social links meant that the selection would be inherently restricted, requiring a larger sample size to cover the more general user audience. In addition, informal interaction-oriented social networks have poor standards of personal data, with many of the required fields missing, made private, or containing unreliable data. Given the amount of data inevitably to be rejected, in order to produce a representative sample the initial list of users had to be made even larger, reaching 298 271 user profiles.

As the present paper’s prime focus lay on data relating to users’ higher education, profiles with empty university and job specialisation fields were dismissed. This reduced the size of the sample by more than 300%, with 83 691 profiles remaining.

The sample was then further rectified by, firstly, dismissing users making evidently false claims about their higher education, and, secondly, searching for profiles with a private or obviously misleading date of birth (like 1845), but a correct date of graduation, that helped to obtain the projected user age as close to reality as possible.

The projected user age was calculated to be the current year minus the year of graduation, corrected by a value dependant on method of education and degree type. I.e., 22 for full-time Specialist degree graduates, 21 for Bachelors, and 23 for Masters. The formula was proofed against factual data present in other profiles judged to be reliable, showing mean absolute deviation of projected and factual figures to be just 1.6 year, that is, within the margin of error permissible in social research.

The resulting sample fit to be studied was 1/8 the size of the initial user list, 37 457 profiles. The profiles included into the sample then had to be grouped accordind to the users’ qualifications. This required a rough list of subjects the students and graduates specialised in, which were then assigned keywords for easier identification.

Where a user’s stated department and faculty titles contained one of such keywords, they were automatically assigned to the respective specialisation group, allowing, of course, for multiple specialisations. Users with multiple specialisations or the ones not established automatically were categorised manually.

Figure 4. Sex-age structure of Vkontakte audience 

Figure 4. Sex-age structure of Vkontakte audienceFigure 4. Sex-age structure of Vkontakte audience

 Figure 4 shows the sex-age structure of Vkontakte audience. Most of the users fall within the 18 to 30 years old bracket, that is, either young specialists in need of work experience or employees focused on vertical career advancement.

Figure 5 shows specialisation preferences in order of attractiveness for Vkontakte users. With users showing a significantly higher interest in economics, various job types in that group have been researched separately in more detail. Second preference was given to engineering jobs, IT came third, law fourth, and fine arts fifth. Preferences for these subjects were analysed for demographic factors.

Economics, engineering, IT, law, fine arts, philology, psychology, mathematics, medicine, physics, journalism, history, geosciences, chemistry, biology, design, cybernetics, pedagogics, physical education and sport, architecture, literature, philosophy, miscellaneous.

Figure 5. Specialisation preferences of Vkontakte users 

Figure 5. Specialisation preferences of Vkontakte usersFigure 5. Specialisation preferences of Vkontakte users

When broken down by sex, the differences in male and female preferences are, generally, self-evident: women, in general, tend to choose social sciences and the humanities, while men prefer engineering, exact, and natural sciences. Philology, psychology, sociology, journalism, design, pedagogics, and literature can be classified as female specialisations. Engineering, IT, mathematics, physics, geosciences (primarily geology), cybernetics, and sport as male ones.

Still, a research into specialisation preferences would be incomplete without correlating it with labour market demand. As is shown below, the bulk of Vkontakte audience is young adults aged 19 to 30. Therefore jobs supply was gauged by vacancies posted on jobs advertising web sites, whose target audience largely falls within this age bracket.

Chosen specialisation, vacancies, engineering, IT, law, fine arts, philology, psychology, mathematics, medicine, physics, journalism, history, geosciences, chemistry, biology, design, cybernetics, pedagogics, physical education and sport, architecture, literature, philosophy.

Figure 6 shows a comparative diagram of subject preferences of Vkontakte users (excluding economics and miscellaneous) and the number of vacancies advertised on jobs web sites [10], [11], [9], and [12]. With the number of vacancies far exceeding the number of users sampled, and also to normalise the discrepancy between absolute job offer figures across different web sites, figures for jobs supply were scaled down to match the number of users in the sample. Looking at the results, one must bear in mind that in certain fields like fine arts and journalism vacancies are seldom looked for online, so jobs web sites usually have no open vacancies of that kind.

The pattern of supply and demand in the labour market is far from balanced: there is an obvious acute shortage of engineers, IT specialists, medical professionals, and teachers; also, the sizeable demand for history professionals is largely driven by history teacher vacancies. Sport professionals and architects are also in demand.

Figure 6. Specialisation preferences (according to and factual labour market demand (according to,,, juxtaposed

 Figure 6. Specialisation preferences (according to and factual labour market demand (according to,,, juxtaposedFigure 6. Specialisation preferences (according to and factual labour market demand (according to,,, juxtaposed

Philologists, psychologists, and sociologists will most likely find employment opportunities scarce. Graduates in mathematics and cybernetics will be unlikely to find work entirely within their area of expertise, though some types of economics-oriented businesses tend to prefer mathematicians and cybernetics professionals to applicants who specialised in economic matters. There is a surplus demand for physics and biology professionals: physicists are being sought for in engineering, and biologists in medicine.

Let us review preferences for specialisations in the field of economics in greater detail. Figure 7 shows the break-down of various economics subjects.
Management, finances, marketing, theoretical economics, global economics, business administration, hospitality, accountancy, mathematical methods in economics, logistics, IT in economics, municipal government, HR management, entrepreneurship, retail, insurance.

Figure 7. Economics specialisations preferences of Vkontakte users 

Figure 7. Economics specialisations preferences of Vkontakte usersFigure 7. Economics specialisations preferences of Vkontakte users 

Management, finances, marketing, theoretical economics, and global economics appear to be the most popular specialisations. The total number of male and female economics graduates is broadly the same, but some specialisations show signs of gender prevalence. Hospitality, accountancy, and HR management may be classified as female-dominated areas, while logistics and IT in economics can be called male professions.

Correlating the relevant specialisation preferences with labour market demand for economics professionals was carried out leaving out mathematical methods and IT in economics (as the more versatile specialisations), and also entrepreneurship (as people trained in this area aim to be self-employed rather than jobseekers).

Figure 8. Economics specialisation preferences (according to and factual labour market demand for economics professionals (according to,,, juxtaposed 

Figure 8. Economics specialisation preferences (according to and factual labour market demand for economics professionals (according to,,, juxtaposedFigure 8. Economics specialisation preferences (according to and factual labour market demand for economics professionals (according to,,, juxtaposed

 Figure 8 presents the preferences in economics specialisation for students and graduates and the demand for professionals in these areas in a way similar to labour market supply and demand patterns for non-economics professionals shown earlier.

Labour market does not look well balanced: there is an acute shortage of retail, hospitality, HR management, logistics, and insurance professionals. Accountancy, finances, marketing, and business administration specialists are also in demand. The greatest difficulties in finding work would face graduates who specialised in management, theoretical economics, and global economics, the latter two emphasising theory over skills that an employer might regard of more immediate value to their business.

6. Conclusions

The qualitative study conducted revealed a number of factors influencing the projected length of employment, including sex, age, job experience, job level, and whether it corresponds to (potential) employee’s area of expertise. Comparing it with the findings of the quantitative study a conclusion can be drawn that the Vkontakte social network, with the majority of its users’ age making them disinclined to engage in long-term employment, can be viewed as a practical source of young professionals or contractors for fixed-term projects. However, the employer must either plan for the short projected employment or be recruiting with a view to giving the employee an early promotion.

Working within one’s area of expertise was singled out as one of the main factors determining longer employment, so employers should recruit people whose education matches the vacancy. However, the patterns of supply and demand in the labour market have shown that there is an almost total mismatch between specialisation preferences of Vkontakte users and actual needs of employers.

Workforce shortage is greatest in IT, medicine, education, hospitality, retail, insurance, and HR management. Looking for specialists in these areas in Vkontakte would be largely inefficient.

Given the interest in vertical career path shown by Vkontakte users, and the large numbers of management and business administration graduates among them, it would make sense to use this social network to form a talent pool or to supply candidates for graduate development programmes.

Moreover, it seems practical to use Vkontakte for recruiting into financier posts and jobs requiring complex calculations, because the number of mathematics and physics graduates among its users exceeds the demand for such specialists and a degree in mathematics of physics can prove necessary for complex calculations involved in business evaluation, business policy making, or product development tasks.

7. References

  1. Allen D. G. Retaining talent: a guide to analyzing and managing employee turnover – SHRM Foundation – 2008 – 43 p.
  2. Clinton R. Lanier. Recruiting with Social Media: Using Social Networks to Drive College Admissions - W1C, 2012 – 170 p.
  3. Douglas H. Reynolds, John A. Weiner. Online Recruiting and Selection: Innovations in Talent Acquisition - John Wiley & Sons, 2009 – 232 p.
  4. Raj Anand. Recruiting with Social Media: Social Media’s Impact on Recruitment and HR - Pearson Education, 2010 – 134 p.
  5. VKontakte users life attitude according to sex and age. URL:
  6. VKontakte users’ profiles statistics. URL: 


скачать dle 10.2 јвто “юнинг кузова