Talent Goes Social: Online Corporate Networking and Business Performance

Abstract

This study examines the effect of online social talent on business performance. The paper uses data from a selected sample of 296 companies from the S&P 500 list with active corporate profiles on LinkedIn. The empirical design consists of non-linear techniques to test the hypothesis that financial performance (i.e., revenue) and online social talent (i.e., employee online profile and skills) have a positive and non-linear relationship. The findings show that internal online social talent measured by employees’ online profiles, and their skills are positively associated with companies’ financial performance. The study provides insights into talent management in the digital age and elucidates the role of online corporate social networking in business performance.

1. Introduction

In 1997, a McKinsey study described the “war for talent” as a strategic challenge and a key driver in achieving competitive advantage [1]. Currently, it seems that the war for talent has gone both online, with more than 4 billion Internet users worldwide [2] and social as online social technologies are playing an increasingly important role at work. Additionally, Covid-19 forced many organizations and individuals to operate in an online setting [3].

Earlier studies in talent management (TM) focused on internal labor market theory, which studied employee promotions and transfers inside the organization [4,5]. More recent concepts refer to hiring from outside the organization [6]. Online corporate social networking (OCSN), in platforms such as LinkedIn (www.linkedin.com), is an online activity that builds professional relationships among corporate members [7]. OCSN mainly supports transfers from outside the organization [8]; however, LinkedIn is not used only for a talent search. Some individuals simply want to network and share their experiences and skills [9]. Thus, LinkedIn may also serve as an internal platform for social talent.

Many studies show interdependence between TM and business performance [10,11] and indicate that obtaining the best talent pool is crucial for organizations [12,13,14]. A few issues are influencing the importance of theories on TM in current organizations. First, many companies struggle with business cycles. During the Great Recession, investments in R&D and human capital were affected by demand reductions and deterioration in credit availability [15], suggesting that companies require new TM methods to gain competitive advantage. Second, a high unemployment rate in many countries worldwide has not led to a talent surplus: companies are still experiencing shortages of experienced professionals [16]. Third, experts pointed out a failure of today’s leadership [17].

Moreover, studies have highlighted the impact of general social media on performance indicators such as movie revenues [18], automobile sales [19], unemployment rates [20], share prices [21], and contract value [22]. Most of the studies focus on how external followers (not affiliated with the firm) affect business outcomes. A smaller portion of the literature focuses on social media activity within firm boundaries [23,24]. However, in our review of the literature, we found no theoretical proposals or empirical descriptions regarding the effect of employee’s online social networking on business performance. However, empirical results of the present study indicate that the relevant academic literature lags behind the practice of the large firms under study (S&P 500). Our results show that employee’s networking is positively associated with corporate financial performance. This finding suggests that a greater theoretical maturity in TM and its application to OCSNs would result in a greater refinement of practice and improved business performance.

This study examines the influence of online social talent (employees and skills) on organizations. First, we identify two types of online social talent defined according to the talent’s origin. External social talent is user-generated by online users or followers. We focus on internal social talent, which is corporate-generated by employees and their skills. Typically, the definition of TM captures the employees of the firm broadly. We extend this definition by considering employees’ corporate profiles on the OCSN platform and their skills as an inseparable part of their talent. Second, we empirically analyze how these types of networks may online social influence business performance. Finally, we conclude with a discussion of results and implications.

In sum, the paper’s main contribution is to continue piling up evidence on social media’s role in business by showing that internal online social talent is positively associated with companies’ financial performance.

The rest of the manuscript is organized and follows. The next section develops the hypotheses levering on the current literature stance. Section 3 describes the data and the empirical methodology. Section 4 reports the regression results and the last section concludes the paper.

2. Theoretical Background and Hypothesis Development

Within the field of human resources, Talent Management deals with managing different types of knowledge [25,26,27,28,29]. Therefore, TM corresponds to gaining awareness of the characteristics of different employee groups and management levels to ensure necessary knowledge levels. Consequently, TM also corresponds to managing people linked to the firm through a network; in other words, managing online corporate talent.

The most traditional TM approach seeks to increase all corporate employees’ knowledge or talent (the inclusive approach to TM) under the premise that the corporation will benefit from having a better-prepared workforce in all operational areas and management levels [25]. The inclusive approach recommends attracting and selecting the brightest managers and workers and encouraging their training and knowledge. Nevertheless, scholars criticized that approach for not differentiating correctly between different groups of employees or managers (exclusive approach), examining the needs of diverse operational, and applying corporate resources to the identified training or knowledge needs. Accordingly, Collings and Mellahi [30] considered that TM’s real function is to distinguish between different parts of the firm’s structure and between other strategic priorities; and concentrate resources where the corporation has more significant training needs or where more excellent knowledge can yield better performance.

The present research extends the exclusive approach of talent management to social networks. Today, part of the corporate talent resides in online social networks (OSN). Therefore, for this study, we defined online social talent as the aspect of TM rooted in the OCSN.

OCSN involves a specific set of social networks, e.g., LinkedIn and ResearchGate, that target professional and academic networking opportunities. It is relevant to the creation of customer- and human resource-related capabilities. OCSN enables informal ties among corporate staff through a social network platform. Online social platforms provide a low-cost, highly accessible communication method that allows individuals to build relationships with people within and outside the corporation. Online social platforms also support task performance by enabling users to engage in online discussions, share knowledge, and find clients [31].

Online platforms let users contribute, share, and receive information also allows users to form virtual communities that interact, cooperate, and create new content. When these forms of online interaction are linked to a corporation, they form part of its resources pool. Thus, researchers have indicated that this online social interaction is crucial for corporate information [32]. A particular corporation’s link reflects admiration or willingness to obtain information about this organization that transmits a positive message about its reputation [33].

Scholarly studies reveal that online content in social networks influences real-world outcomes. Using user-generated content on Trip Advisor, (www.tripadvisor.com) O’Connor [34] established that customer opinions on the web can manage a hotel’s image. Fieseler, Fleck [35] demonstrated that the blogosphere adds corporate value and stakeholder engagement. Previous studies have examined corporate networking tools (e.g., CRM or e-business) as an asset used to increase operational performance [36,37]. However, few researchers have specifically studied how social media enhance business performance. Trainor, Andzulis [37] have shown how social media usage relates positively to customer relationship performance by creating firm-level capabilities. Paniagua and Sapena [21] lean on resource based-view (RBV) to explain how social media affects business performance. Their approach is to consider user-generated content as a resource that the firm transforms into performance-enhancing capabilities. An empirical analysis relating stock prices with Facebook (www.facebook.com) and Twitter (www.twitter.com) followers backed their hypotheses.

These studies focus on corporate followers who often undertake online activities such, as likes and comments, on OSCN platforms [38]. We define these followers as external online social talent is user-generated content and comprises online followers. Many scholars have indicated the importance of the number of followers as a proxy for popularity, status, and influence [39,40,41]. Paniagua, Korzynski [42] show how corporate followers influence internalization by examining the effect of Facebook and Tweeter activity on foreign direct investment.

To delve into the detailed mechanisms by which online social talent affects business outcomes, we should refine our understanding of online social talent in OCSN sites. We identify two ways through which corporate-generated online social talent can affect business performance: external and internal. External social talent is composed of individuals outside an organization who affect business performance in two ways: revealed preferences and crowdsourcing [21]. Internal social talent refers to social media assets within the firm.

External online social talent provides room to learn and facilitates customers’ and candidates’ preferences and suggestions that emerge from the OCSN. Through their online social portfolio, followers reveal which firms better reflect their job and product preferences. Companies also use external online social talent in their business processes through social media crowdsourcing [43]. Crowdsourcing—using information technologies to outsource business tasks to virtual crowds—is a significant resource to build competitive advantage [44].

Characteristics of customers’ revealed preferences and crowdsourcing as resources that can contribute to better corporate performance are partial—albeit with broad potential for competitive and performance improvement—resources that can create value for the firm (can be firm-specific) and are asset-stocks whose creation requires accumulation over time [45].

Internal Online Social Talent

Internal online social talent comprises the employees registered on the OCSN platforms. Internal online social talent the corporate-generated generated content and includes employee’s online skills. These skills help employees in building an online personal brand [46], professional identity [47], and social interactions [48]. Social corporate ties are established within or outside the company [21]. OCSN within the company increases employees’ social relations and supports communication effectiveness, thereby providing additional interactions parallel to face-to-face meetings and email [8]. With OCSN, the firm establishes links that go beyond its corporate boundaries. Organizations engage actively with customers, stakeholders, competitors, and potential candidates. In particular, organizations can perform searches and engage in display activities through OCSN platforms [49]. Search activities relate to finding candidates whose experience and skills are suitable for an open position. In contrast, display activities include posting information about job offers, products, or services and posting comments and creating corporate groups.

Besides, OCSN platforms contribute to raising the visibility of the corporation’s products or values through social marketing [50]. Firms with OCSN skills can engage in viral advertising and disseminate proposals quicker and more efficiently [51].

Paniagua, Rivelles, and Sapena [24] show how board members’ social media activity is positively associated with financial performance indicators. Their study shows how a very particular group of internal social talent, executives, affects firm performance. We consider that a larger group of employees’ online social talent may also have a positive effect on business performance.

Collective learning and knowledge through interaction with the corporation form a valuable resource for the corporation, is rooted in the specific relationship within the corporation, and requires time to be created, which means that it is rooted in not being improvised [45]. Nevertheless, in addition to fulfilling these conditions for customers’ revealed preferences and crowdsourcing (characterized as specific resources or specific assets in the Resource-Based View), these assets’ complementary and non-central character must be added. In general, these assets derived from OCSN are distinct from the corporation’s core competences in a sense used by Prahalad and Hamel [52]. However, many of the elements that configure them, such as organizational reputation, can generate a high level of prestige and reputation [53,54]. Corporate reputation (to attract users to the network) is enhanced by contributions or suggestions from users [55]. Further, online relations between employees help corporations find the best candidates for specific jobs [56].

Korzynski, Paniagua, and Rodríguez-Montemayor [23] show how online tools, social media, and knowledge management enhance employee’s skills. Precisely, creative skills that allow employees to solve problems and access knowledge more efficiently. The online social skills developed by the corporation and its social media through social marketing and employees’ social networking enable the online relationship conversation, sharing identity aspects on the consolidation of trust and reputation. Online connection with the corporation is a reservoir of experience, knowledge, and talent, and is, therefore, an external source of strategic and operational orientation capabilities.

We, therefore, state our first hypotheses:

Hypothesis 1 (H1a).

Financial performance (i.e., revenue) is positively associated with online social talent measured by the number of online employee profiles.

Hypothesis 1 (H1b).

Financial performance (i.e., revenue) is positively associated with online social talent measured by online employee skills.

Network externalities that capture the effect of advertising thresholds and wear-outs [57] are not uncommon in the context of social media. Several studies show that building social media capabilities is costly for firms. They have to invest time, effort, and resources to manage online profiles, posting, and responding to interactions. Marwell, Oliver [58], in their theory of a critical mass in social networks, showed how the density of social ties in a group improves its prospects for collective action. Lin and Lu [59] empirically determined that the number of peers and perceived complementarity greatly influence social networking site usage. Paniagua and Sapena [21] showed how Facebook and Twitter positively affect the stock prices of traded firms, but only after a critical mass of followers is attained. Therefore, Paniagua and Sapena [21] showed that the effect of social media technology adoption on business performance is non-linear. In particular, it is U-shaped due to economies of scale, meaning that the marginal costs decrease as the number of social users increases for a given firm. Therefore,

Hypothesis 2 (H2).

Financial performance (i.e., revenue) and online social talent have a non-linear U-shaped association.

3. Materials and Methods

3.1. Data

LinkedIn is the most popular and widespread OCSN platform, which reported 706 million users and three million business pages in more than 200 countries in 2020. LinkedIn is the default corporate networking and recruiting platform. This OSCN site counts executives from all Fortune 500 companies as members. The majority of the Fortune 100 companies use LinkedIn’s corporate talent solutions. The results of the 6th annual Social Recruiting Survey from social recruiting platform Jobvite (2013) revealed that 94% of online recruiters turn to LinkedIn, 55% to Twitter, and 65% to Facebook.

LinkedIn is especially well suited to contrast our hypotheses, as it provides information on both corporate and user-generated online social talent. Registered users on LinkedIn share their skills on their online profiles. Since corporate users are linked to a specific company, we can then match corporate skills and companies to obtain the aggregate social networking skillset per corporation. Companies with corporate profiles on LinkedIn normally display the number of followers. These are registered users (mostly non-employees) who receive online updates on the company’s online activity, i.e., posts or job offers.

Our data came directly from the linkedin.com webpage. We selected a sample of companies with active corporate profiles on LinkedIn. Our sampling strategy randomly selected 300 companies from the S&P 500 to ensure a representative sample. After a detailed analysis, 296 profiles with full available data are included in our study. The entire company list is available upon request.

We collected the number of followers and skills in their LinkedIn profile for each company in

Table 1. Summary statistics.

Variable	Observations	Mean	Std. Dev.	Min	Max	Revenue	Employees
Revenue	296	38.795	81.298	0.01	829.73
(Billion USD)
Employees	296	19,317.25	34,281.01	9	357,455	0.556 ***
(Number)
Skills	296	559,799.4	829,675.7	8138	7,126,588	0.408 ***	0.408 ***
(Number)

Notes: * p < 0.05, ** p < 0.01, *** p < 0.001.

. Each user may define their skills, e.g., sales and IT, on the networking site. For practical purposes, we selected the five most popular skills per company. Our measure for corporate social networking talent is the aggregated number of skills per company. User-generated social networking talent is simply the number of overall followers. We collected those data at the end of the year 2013.

Revenue is our indicator of firm performance. Financial performance is the center of the three business performance domains [60]. Revenue is income a company receives from its regular business activities and is frequently used in studies in this area [24]. Sales of goods and services to customers are the primary source of any company’s revenue. Therefore, it is an adequate measure of business performance since revenue is affected by the four channels (customer’s preferences, crowdsourcing, marketing, and employee networking). We collected data on the firm’s annual revenue, as reported by S&P IQ Capital. Table 1 reports the descriptive statistics and correlation matrix of our dataset.

3.2. Empirical Method

We apply the following log-linear fixed-effects regression equation to test our first hypothesis:

$$\ln R_{is}={\beta }_1\ln employe{e}_i+{\mathsf{\beta }}_2\ln skill{s}_i+{\lambda }_s+e_{is}$$

(1)

where R is the revenue of a firm i in sector s; employee is the number of LinkedIn employee profile of a particular firm i; skills are the number of aggregate skills reported by the employees; ${\lambda }_s$ are a set of fixed-effect dummies per sector; and e is a stochastic error term.

Convenient research design can hedge most limitations related to cross-sectional data [61]. The log-linearization of the variables and the clustering of robust standard errors at the firm level helps increase our inferences’ validity. A potential concern is associated with the firm size. McCann and Barlow [62] show that social media adoption by Small and Medium Enterprises (SME) is distinct from large corporations. Additionally, the influence of social media on firm performance has particular traits in SMEs [63]. It is, therefore, relevant to control for firm size to avoid omitted variable bias. The first step in this direction is the use of sector fixed effects. Ideally, we would include firm-fixed effects to control all constant firm characteristics, including firm size. Unfortunately, this is not possible in our cross-sectional data because the firm-fixed effects would absorb all the firm-level variation.

We take two actions to control omitted variable bias and increase the confidence in our inferences. Firstly, we introduce sector fixed-effects (${\lambda }_s$). The fixed effects control for the heterogeneity of firms across sectors. The sectorial dummies also capture any omitted variable that affects companies within the same industry. Therefore, under the plausible assumption that firms are similar within each sector, these fixed-effect absorb the firm heterogeneity in terms of firm size and reduce the omitted variable bias.

Secondly, we adopt a second measure that several scholars used to combat this bias: quantile regressions [64,65,66]. This type of regression performs conditional regressions for each quantile of the dependant variable. Quantile regression has the benefit of correcting a potential estimation bias derives from several companies’ weight in the sample. Standard regression techniques assume symmetric distributions, and if the data are skewed, the mean estimates of the coefficients are potentially biased.

In sum, the combination of a structural estimation along with quantile regression limits the potential harm of omitted variable bias.

The test hypothesis 2, related to the non-linearity of the variables of interest, is performed including the quadratic form of the variables of interest in our regression equation. Particularly:

$$\ln R_{is}={\beta }_1\ln employe{e}_i+{\beta }_2\ln employe{e}_i^2+{\mathsf{\beta }}_3\ln skill{s}_i+{\mathsf{\beta }}_4\ln skill{s}_{si}^2+{\lambda }_s+e_{is}$$

(2)

The interpretation of the $\beta$ coefficients in Equation (2) is different from the marginal effects in Equation (1). Focusing on employees, ${\beta }_1$ is the slope of the non-linear regression line at the origin and ${\beta }_2$ is a measure of the curvature of this relationship. With the first-order conditions of (2), i.e., the derivative of revenue, we obtain the critical mass of employees upon which the sign of the effect changes.

4. Results

Table 2. Regression results.

	(1)	(2)
Log employees	0.503 ***	−1.132 **
	(0.0731)	(0.437)
Log skills	0.161 *	1.159
	(0.0879)	(1.041)
Log employee2		0.104 ***
		(0.0264)
Log skills2		−0.0494
		(0.0423)
${\lambda }_s$s == Consumer Goods	−0.836 **	−0.636 *
	(0.328)	(0.338)
${\lambda }_s$sector == Financial	−1.562 ***	−1.343 ***
	(0.361)	(0.383)
${\lambda }_s$sector == Healthcare	−1.183 ***	−1.028 **
	(0.425)	(0.434)
${\lambda }_s$sector == Industrial Goods	−1.405 ***	−1.307 ***
	(0.365)	(0.379)
${\lambda }_s$sector == Services	−1.422 ***	−1.240 ***
	(0.333)	(0.342)
${\lambda }_s$sector == Technology	−1.810 ***	−1.757 ***
	(0.333)	(0.339)
${\lambda }_s$sector == Utilities	−0.981 *	−0.829
	(0.553)	(0.546)
${\lambda }_s$Constant (Basic materials)	−2.676 ***	−1.462
	(0.807)	(5.966)
Observations	296	296
R2	0.394	0.427

Robust standard errors in parentheses, clustered by company. * p < 0.10, ** p < 0.05, *** p < 0.01.

reports the results of the regression analysis. The dependent variable is the log of each firm’s revenue. Columns 1 reports a benchmark estimate without adding sector fixed effects to the regression equation, which are introduced in column 2. Column 3 reports the results of the non-linear specification by adding the quadratic terms.

The results in Table 2 confirm both hypotheses. The linear estimates of columns 1 show that both employees and skills are positively and significantly associated with revenue. However, the statistical significance of employees (p < 0.01) is higher than that of skills (p < 0.1). The effect’s magnitude is also higher for employees (0.503 in column 1) than for skills (0.161). Since we have a log-linear specification, the estimates are interpreted as elasticities: all the rest considered, companies with 1% more employees have higher revenues by 0.5%.

We also report the coefficient of dummy variables for each sector or fixed-effects. The constant is the intercept of the base category, basic materials, and each dummy variable is the differential slope concerning this category. All coefficients are statistically significant, revealing that the data had constant heterogeneity at the sector level, which would have been unaccounted for with the fixed effects. We also ran regressions without these fixed effects (available upon request), showing that the estimates of our variables of interest are underestimated without the sector effects.

Hypothesis 2 is tested with the quadratic terms included in column 3. We obtain empirical evidence sustaining the non-linear hypothesis only for employees (the quadratic terms for skills are not significant).

In column 3, the negative coefficient of employee hints that the relationship is negative near the origin. However, the positive sign of employee² indicates that at some point, the relationship turns positive. We can easily calculate this critical mass of employees upon which the relationship tends to increase:

$$\frac{\partial R}{\partial \ln employee}= {\widehat{\beta }}_1+2{\widehat{\beta }}_2\ln employe{e}^{*}=0 \to employe{e}^{*}=\exp -\frac{{\widehat{\beta }}_1}{2{\widehat{\beta }}_2}=231$$

(3)

Figure 1 depicts the scatter plot of revenue vs. employees (in logs) and the quadratic fit regression line. It is readily seen how the positive relationship between employees and revenue catches early on the data (231 employees or 5 in logs), and only for a few firms in the left tail of the employee distribution is this relationship less positive.

Lastly,

Table 3. Quantile regression results.

	Q(0.1)	Q(0.25)	Q(0.5)	Q(0.75)	Q(0.90)
Revenue(billions)→	1.53	3.42	13	39	93
Log employees	0.655 ***	0.572 ***	0.509 ***	0.416 ***	0.515 ***
	(0.090)	(0.053)	(0.067)	(0.096)	(0.126)
Log skills	0.0933	0.258 ***	0.216 *	0.0469	−0.153
	(0.121)	(0.071)	(0.090)	(0.129)	(0.168)
N	296	296	296	296	296

Standard errors in parentheses. * p < 0.05, ** p < 0.01, *** p < 0.001.

shows the results of the quantile analysis. We observe that the effect of employees is positive and significant in all revenue quantiles. This result increases our confidence, supporting our hypotheses. Interestingly, the effect is higher in the lower tail of the revenue distribution. In particular, the largest effect of employees is for companies with about 1 billion annual revenues. As the firm size increases (in terms of revenue), employees’ marginal effect is still positive and significant, but with a smaller effect.

The effect of skills is only positive and significant for medium firms between 3 and 13 billion in revenue. Other than those quantiles, the effect of skills is not significant. These results explain why skills had no significant effect on the previous quadratic regression, which could not capture the effect of skills only on a subset of firms.

In sum, online social talent has a larger effect on small and medium-sized firms (based on revenue). The marginal effects of the employees and skills are plotted in Figure 2. As expected from the quantile regression results in Table 3, the magnitude of these effects decrease for employees and is only significant in a small portion of the revenue distribution for skills.

5. Discussion

This research broadens the theoretical reach of talent management by developing the online social talent management concept. We show that online social talent is relevant for corporations as it significantly affects their financial performance. We argue that internal (generated by corporate online users) are resources that firms transform into performance-enhancing capabilities.

We have analyzed the impact of internal online social talent measured by aggregate social networking employees and their skillset of employees registered on LinkedIn. Bhattacharya, Gibson [67] indicate that flexibility in employee skills positively influences corporate results. This statement is aligned with our findings because we have discovered a positive relationship between business performance and employee skills on online corporate social networking (OCSN) platforms, which, according to many scholars, ensure flexibility and adaptability [68].

Second, we have uncovered several mechanisms that help better understand social talent’s effect on a firm’s performance. Remarkably, the study shows that the effect is non-linear, and a small critical mass of online employees is needed to jump-start positive results. Secondly, the impact of social talent is not uniform across all firms. Relatively smaller firms benefit the most from employees socializing online.

Our results build upon the literature initiated by Paniagua and Sapena [21], who were among the first to conceptualize the mechanisms that underlie in our findings. They showed how social media has a non-linear relationship with financial performance. They also suggested that corporate networking was an important channel for this relationship. Other previous studies picked up the gauntlet before studying OCSN in engineers [23] and board members [24]. We advance the knowledge frontier by examining the role of an employee’s professional activity and skills on business performance.

The study advances the conceptualization of social media’s role in business by interviewing them with talent management. Until now, these bodies of literature were relatively isolated from one another. The paper pushed forward in ways to understand how they interrelate. The paper introduces internal and external social talent and associates them with how individuals interact in OCSN.

Our findings have significant implications for organizations and their talent management activities. First, organizations may influence the number of followers by outsourcing business tasks to virtual crowds and facilitating customer and candidates’ revealed preferences. Second, they can affect employees’ corporate skill sets on an OCSN platform by supporting and encouraging their employees to sign-in on the platform and use online networking tools to interact with other individuals. Third, organizations may provide their employees with premium accounts that raise corporate employees’ visibility and skills.

The study is not without limitations. The data’s cross-section nature limits causal claims on our results, showing promising and exciting developments that align with our theoretical expectations. A second limitation is that the sampled firms are relatively large. Although we observed differences according to revenue, our study speaks little to the effect of SMEs. Thus, this study puts forward an exciting research agenda. A third limitation is that new dynamics might alter the mechanisms unveiled by our data snapshot, limiting to a single year. Fourthly, the study uses only one performance indicator (revenue).

Scholarly work aimed to complement and refine our findings is positively encouraged. For example, future research aimed at explaining the employee’s activity on online social networks. Another interesting path would be to study the effect of social talent in SMEs. Additionally, due to Covid-19, many organizations and individuals started or increased online activities, presenting promising new research avenues.

6. Conclusions

Talent management goes social through online OSCN platforms. Management theory is only beginning to grasp how social media affect organizations. The purpose of this study was to examine and quantify the role of online social talent in business performance. The findings support the study’s hypotheses: online social talent is positively associated with business performance in a non-linear way.