Firm optimization with Big Data Analytics:
The impact of Big Data Analytics on labor productivity.
Eline Kajim S3153843 University of Groningen
MSc BA-CM Supervisor: Mr. J. Dong Co-Assessor: Prof. Dr. J. Surroca
Abstract
In light of the fact that firms progressively adopt big data analytics (BDA) as a driver for enhanced firm performances, this paper is among the first to explore how BDA influences labor productivity. Drawing on the resource-based view (RBV) of information technology (IT), the basic premise we propose is that technical and human IT resources stimulate superior BDA practices. We investigate how the actual practice of BDA at firm level may influence labor productivity and assess whether there is a difference in between small to medium enterprises (SMEs) and large firms. By using a large-scale survey data set from firms in German, we found empirical evidence that technical and human IT resources positively affect BDA and in turn BDA has a positive effect on labor productivity. The latter is more salient for large firms compared to SMEs.
1. Introduction
Big data publications have seen an explosive growth since 2007 (Chen et al., 2012) and more recently businesses are being redefined by a data-driven revolution (Tushman et al., 2017). The concept of big data is primarily used to describe large data sets that include overwhelming amounts of unstructured data that requires more real-time analysis (Chen et al., 2014). Sequentially, organizations are increasingly relying on digital technologies to manage large data sets, but to establish and maintain superior firm performances with these digital technologies change management has become rather important (Whyte et al., 2016). This entails managing the people side of change with processes, tools and techniques to realize business outcome. We are currently living in a “big data society”, which means that even without actively seeking for information we are bombarded with large amounts of data whether we like it or not (Chen et al., 2012). Currently change managers are having a hard time finding scientific proof of work for demonstrating the link between cause and effect (Tushman et al., 2017). This is primarily since change management issues are predominantly related to human behaviors and intangible factors like culture, leadership, and motivation that do not easily yield empirical analysis (Tushman et al., 2017). Additionally, the intensity of tremendous information flows develops a perception of information overload, whereby the supply of information exceeds individual processing capacities (Zelenkauskaite and Simöes, 2014). Within organizations big data is seen as the key to success. Consequently, employees are expected to deal with an overwhelming amount of data from various sources. Edmunds and Morris (2000) suggested that survival in modern society depends on people’s ability to process vast amounts of information. Although information could benefit organizations success, information overload can make people feel stressed, less satisfied or even physical ill, and thereby unproductive (Edmunds and Morris, 2000).
of imperfect, complex and unstructured data into valuable information, to decrease information overload and to simplify decision making processes (Hilbert, 2016).
Although, existing literature highlights the effects of information overload, whereby rapid increasing information streams could create bottlenecks for each individual, former research has paid limited attention to the benefits of BDA on labor productivity. Labor productivity is relevant for IT studies as several scholars indicated that new IT systems help firms to manage scarce and expensive labor more efficiently, which ultimately resulted in enhanced firm performance and competitive advantage (e.g., Hitt and Brynjolfsson, 1996; Mukhopadhyay et al., 1997; Rai et al., 1997).
The RBV of IT is a helpful managerial framework that is commonly used in IT literature to determine why some firms consistently outperform others (e.g., Bharadwaj, 2000; Kearns and Lederer, 2003). Drawing on the RBV of IT is important to our understanding as it explains how BDA allows firms to systematically prioritize, categorize and manage data that provide firms with the opportunity to generate valuable, rare and imitable practices of BDA that positively influence competitive advantage and firm performance (Barney, 1991; Eisenhardt and Martin, 2000; Teece et al., 1997). Moreover, it emphasizes the importance of internal IT resources that are the main predictor of superior financial benefits (Kearns and Lederer, 2003; Wernerfelt, 1984).
To fill the literature gap of BDA and labor productivity, we theorize and empirically examine both technical and human IT resources. That are identified as basic components of analyses. By assembling these resources firms are able to construct firm specific BDA practices (Barney, 2104). We add to the theoretical understanding of RBV of IT by investigating the BDA practices at firm level, leading to enhanced labor productivity. Next, we study the differences in labor productivity derived from BDA among SMEs and large firms. We utilized the 2016 cross-sectional survey data set from the Manheim Innovation Panel (MIP) database. This is a large-scale cross industry data set that has been anonymously conducted across firms located within Germany and generated a total sample size of 4685 firms. We found that firms capitalizing their technical and human IT resources benefit their BDA practices and firms that are able to get ahead of these resources developments continue to experience faster labor productivity growth. Furthermore, we found that BDA is relatively more valuable for large firms compared to SMEs.
2. Theoretical foundation
Since recent years, the concept of BDA has become increasingly popular among both academics and practitioners within the world of IT (e.g., Baesens et al., 2016; Chen et al., 2012; Zakir et al., 2015). Since BDA is relatively new and still in an evolving stage, there seem to be no uniform definition yet. Chen et al. (2012) argued that BDA has been used to define data sets and views BDA as an analytical technique that is (1) large enough to handle exabytes, rather than terabytes that is processed in traditional business analytics techniques and (2) complex as it deals with large amounts of data. Zakir et al. (2015) defined BDA as a way to extract value from large volumes of data, to seize novel market opportunities and to exploit customer retention. This study defines BDA as a firm specific practice that allows firms to manage, process and analyze the 5V data-related dimensions; volume, variety, velocity, veracity and value (Baesens et al., 2016). BDA requires advanced and unique data storage, management, analysis and visualization technologies (Chen et al., 2012), to stimulate improved data-driven decision making and to make firms more forward thinking. Thus, to generate superior BDA practices that are difficult to imitate for others and establish competitive advantages (Wamba et al., 2016). BDA will change our way of doing business and continues to transform not only the technology industry, but many more, like service industries (Demirkan and Delen, 2013), healthcare organizations (Raghupathi and Raghupathi, 2014), and even in the education sector (Siemens and Long, 2011). On the one hand, BDA is prized for its contribution to produce business efficiency (Zheng et al., 2003) and productivity (Ren et al., 2017; Wamba et al., 2016). It is without a doubt that businesses converge into explorations of BDA. On the other hand, BDA brings along various difficulties, for example, its complexity issues to obtain real time value (Baesens et al., 2016) or the increasing struggle of talent shortage in the IT sector (Brown et al., 2011).
BDA are much faster in anticipating upon customer choices. Also, knowledge management represents BDA intangible IT enabled resources. Since it is an inherent social process of creating, sharing and using knowledge and information, it is embedded in employee skills and experiences. Particular, generating superior BDA practice may soon become a crucial asset to boost information quality, which in turn forms a good foundation for decision-making processes and positive labor productivity (Wamba et al., 2015).
Technical IT resources are defined as the overall IT infrastructure that contains all computer and communication technologies and includes all shared information platforms (Bharadwaj, 2000; Ross et al., 1996). Barton and Court (2012) emphasized the importance of robust and physical IT infrastructures, as these are imperative to BDA. Tremendous streams of data arising from big data made people become more dependent upon particular hardware (e.g., computers) and software (e.g., Apache Hadoop) to assist them in their analyses. Moreover, McKenney (1995) argued that a firm’s IT infrastructure is one of their most valuable sources to sustain competitive advantages. Prevailing infrastructures were developed to deliver data in batches, but managing unstructured data occurring from big data, goes beyond the scope of most existing infrastructures. Therefore, adjusted IT infrastructures that secure a continuous flow of information are required to support BDA, to make real-time decisions leading to competitive advantage (Barton and Court, 2012; Wamba et al., 2015). However, advocates of RBV argued that physical assets like IT infrastructures can be purchased or easily imitated by competitors and therefore are unlikely to offer any advantages over competitors (Mata et al., 1995). An accurate and unique integration of IT infrastructures is therefore extremely important and should help firms to (1) identify and develop key applications rapidly, (2) share information across products, services, and locations and (3) exploit opportunities across business units (Reed and DeFillip, 1990). Adequately, the focus should be on how firms transform their IT infrastructure into inimitable BDA practices that enriches labor productivity (Barua et al., 2004; Brynjolfsson and Yang, 1996; Ren et al., 2017). Nonetheless, implementing such an integrated IT infrastructure, that links entire organizations including all relevant stakeholders such as suppliers and customers, demands time and great expertise. Consequently, human IT resources as commonly believed to be supplementary and equally important to technical IT resources (Chen and Zhang, 2014).
insights and to share them with key firm stakeholders for competitive advantage, co-creation and realization (Wamba et al., 2015). To attain full benefits from BDA, managers must create a skilled internal IT workforce by aligning existing organizational cultures and practices across the entire organization (Wamba et al., 2015). Barton and Court (2012) referred to this as making data trustworthy and understandable for all levels of the organization. Furthermore, Shah et al. (2012) even suggested that investments on BDA do not provide any returns unless all employees are able to embrace BDA in their decision-making process. It could be stated that human IT resources are rather challenging to develop, but in turn makes them more difficult to reproduce by others. Hence, the RBV of IT supports the concept that human IT resources serve as a source for BDA.
BDA could be considered a new IT innovation, as Ren et al. (2017) and Wamba et al. (2016) emphasized it requires novel ways of using computers and technological systems to store, retrieve, transmit and manipulate data that leads to more efficient internal processes and improved alignment between technology initiative and organizational objectives. Since, Mukhopadhyay et al. (1997) verified that new IT systems enable firms to advance their labor productivity and similarly Rai et al. (1997) indicated that IT developments succeeded in reducing production costs and improve the productivity of all personnel, BDA is also expected to produce improved labor productivity. The competency of BDA is mainly determined by efficient information systems to achieve improved firm performance (Grant, 1991) and support productivity in regards of logistics, inventory management and pricing strategies (Davenport and Harris, 2007). Prior literature has proven that BDA enables firms to improve firm performance, in various forms, such as customer retention, return on investment, profitability and general sales growth (Ren et al., 2017; Wamba et al., 2016), but it is rather unknown whether BDA could also drive labor productivity, which is measured by number of sales per employee. It is therefore a good moment to demonstrate potential labor productivity performance of BDA, to further encourage firms to adopt BDA and strengthen their IT resources.
3. Hypotheses
3.1. The Antecedents of Big Data Analytics
Based on RBV of IT, we argue that technical IT resources have a direct effect on exercising BDA. Since increased amounts of data marked the issue of information overload, technical IT resources became imperative for organizations to continuously assemble real-time data (Barton and Court, 2012). Technical IT resources are seen as tangible resources, whereby IT infrastructures are most predominant. One of the most time consuming and labor-intensive task of analytics is the preparation of data prior to performing the actual analyses (Assunção et al., 2015). This is an issue that is even exacerbated by the come of BDA, as it stretches the limits of current infrastructures and traditional business strategies. Significant more firms are adapting their infrastructures towards the specifications of the new digital era (e.g., Banker et al. 2006; Rai et al. 2012; Sambamurthy et al. 2003). These specifications require more efficient data storage and filters, transforms and retrieves data more effectively to service more range and reach (Broadbent et al., 1996).Although, manufactures are fabricating various hardware and software systems, the fact remains that any firm can purchase equivalent IT systems as its competitors. Being competitive, firms should adjust their IT infrastructure aiming to fully intergrade it into their organization in order to create inimitable BDA practices.
McKinsey (2011) reported that there is a projected growth of 40% of global data generation per year versus 5% growth in global IT spending. Within the coming years, the patch of the digital universe will not only become bigger but also more complex (Gantz and Reinsel, 2012). As a result, contradicting growth skills, experiences and resources to deal with big data become scarcer and more specialized. Consequently, it requires more novel, flexible and scalable IT infrastructures that extends beyond current infrastructures. For example, in 2011 Facebook announced their Open Compute Project (OCP), interesting for both datacenter design as well as for server design. With OCP, Facebook openly shared its modular server design to reduce costs and avoid additional complexity of hard and software systems supporting BDA (Villars et al., 2011). As such bundling and altering IT infrastructures demonstrate their influence on BDA.
Beyond this example, there are many more novel developments and initiatives that demonstrate a growing need for more specialized technical IT resources. Respectively, this represents a grounded reason for measuring the physical technical IT resources. Therefore, we propose to study the influence technical IT resources have on BDA, by means of the following hypothesis:
Hypothesis 1: Technical IT resources are positively associated with big data analytics. Similar to technical IT resources, we argue that human IT resources have a direct effect on practicing BDA, as human IT resources empower firms to perform BDA.
Technical and managerial IT skills are considered critical dimensions of human IT resources. Prior research has proven that solid human IT resources; (1) more effectively integrate IT and business planning processes, (2) more rapidly support business needs by conceiving of and developing reliable and cost-effective applications, (3) communicate and operate more efficiently among different business units, and (4) enable firms to foresee future business needs before or better than competitors (Bharadwaj, 2000). Moreover, several scholars (e.g., Barton and Court, 2012; Shah et al., 2012; Wamba et al., 2015) emphasized the need to involve employees across the entire organization to encourage data trustworthiness and exploit full advantages of BDA. However, while the demand for advanced BDA is growing, talent who get all the insights and IT skills to the front lines remain rather hard to come by (Chen et al. 2012; Davenport & Patil 2012).
For example, a quarterly report of McKinsey Global Institute highlighted that by 2018; approximately 140.000 to 190.000 extra specialists in this field are required, along with 1.5 million managers and analysts, who have a precise understanding of how BDA could be used to benefit them and their organization (Brown et al., 2011). Likewise, Debortoli et al. (2014) and McAfee et al. (2012) expressed that BDA requires capable BDA specialists, who hold abundant technical skills to effectively make sense of statistical data analysis.
Firms that are able to acquire and retain human IT resources can unlock the potential of unique and rare exploitation of BDA. Therefore, we propose that human IT resources have a positive influence on BDA. Leading to the following hypothesis:
3.2. The Consequence of Big Data Analytics
The relevance and importance for investigating the relationship between BDA and labor productivity arose from prior literature indicating that BDA is a novel IT innovation that drives various forms of firm performance (Ren et al., 2017; Wamba et al., 2016) and by following Hitt and Brynjolfsson (1996), Mukhopadhyay et al. (1997) and Rai et al. (1997), who claimed that novel IT innovations commonly result in enhanced labor productivity.
The RBV of IT suggests that firms should possess and exploit resources and capabilities that are both valuable and rare in order to obtain competitive advantage. Also, these should be difficult to imitate and substitute to sustain the competitive advantages (Barney, 1991; Eisenhardt and Martin, 2000; Teece et al., 1997). These advantages may ultimately manifest in improved performance in the short-term (Barney, 1991). Newbert (2008) expressed that the terms competitive advantage and performance are frequently used interchangeably while they are recognized to be conceptually distinct. Competitive advantage is considered to be the implementation of a strategy that has not yet been employed by others and facilitates a cost reduction and exploits market opportunities (Barney, 1991). Whereas performance is theorized as the outcome a firm ensues after implementation of its strategies (Newbert, 2008). Therefore, labor productivity could be considered an indicator of competitive advantage, as firms exploit their technical and human IT resources to superior BDA practices.
Thus, BDA allows firms to systematically prioritize, categorize and manage data that provide firms with the opportunity to generate valuable, rare and imitable practices of BDA that positively influence competitive advantage and performance. Therefore, we propose that BDA has a positive effect on labor productivity. This is presented by the following hypothesis:
Hypothesis 3: Big data analytics are positively associated with labor productivity. 3.3. The Mediating Role of Big Data Analytics
Hypothesis 4: Technological resources are mediated by big data analytics to positively influence labor productivity.
Hypothesis 5: Human resources are mediated by big data analytics to positively influence labor productivity.
3.4. The Moderating Role of Firm Size
In literature it has been widely recognized that firm size is the main source of driving heterogeneity in firm performance and IT innovations (e.g., Geroski, 1998; Griliches and Regev, 1995; Liang et al., 2010; Zheng et al., 2003). When performing regression analyses, is it quite common to control for firm size (Geroski, 1998). Halkos and Tzeremes (2007) explained that firm size may shape the effects of other variables on productivity through other variables, since they are highly likely to develop dissimilar behavior patterns among SMEs and large firms. According to firm size various strategies can be applied to affect productivity, among transferring knowledge, skills and technology, that all reflect attributes of practicing BDA (Rumelt, 1982). We thus investigate the moderating role of firm size in the relationship between BDA and labor productivity. Although, we still argue that BDA has a positive effect on labor productivity, we reason that there is a difference between SMEs and large firms, such that larger firms may experience greater benefits of BDA in productivity improvement. Since, larger firms are systematically found to operate and control a greater scale and greater scope to acquire its data, their data is likely to be more overwhelming. In contrast to SMEs that penetrate a significant smaller scale and scope, they are expected to experience more information overload with processing data. Thus, BDA is more valuable for facilitating decision making in large firms compared to SMEs. Besides, achieving sustainable competitive advantage with BDA heavily relies on decision makers (Court, 2015). McAfee et al. (2012) added that decision making capabilities with BDA are different across functions and business units. This highlights the additional value of BDA for large organizations, as large organizations are comprised of significantly more business units that cause additional complexity with business strategies, structures, management processes and sub-cultures (Boynton et al., 1994). This might be particularly true for firms at which each business unit adopts IT in very distinct ways (Boynton et al., 1994). Therefore, we propose that firm size positively moderates the effect of BDA on firms’ productivity, as it conditions the impact of internal factors of BDA on labor productivity is stronger in large firms than that of SMEs. Hence, the following hypothesis is formulized;
4. Methods
4.1. Data Collection
We used a large-scale, cross-sectional survey data set to test our hypotheses. This survey is authorized by the German Federal Ministry of Education and Research (BMBF) and has been conducted conjunction with the Centre for European Economic Research (ZEW), the Fraunhofer-Institute System & Innovation Research and with the Institute for Applied Social Sciences (Infas). On an annual basis since 1993, ZEW has been conducting this representative firm survey regarding innovation activities of German firms across industries (e.g., mining, food, chemicals, retail, water, wholesale, transport, media, telecommunications, banking and consulting). The survey data used in this study came from MIP database, which contributes to the German part of European Commission’s Community Innovation Surveys (CIS). ZEW is heavily involved in international working groups that established the survey’s methodology and defined innovation activities based on “Oslo Manual”. This resemblescovenant with well-ground guidelines contributing to the reliability of the data and assures a high response rate in data collection. Also, the MIP database has been used in prior IT research (e.g., Dong and Netten, 2017). The sample for this research was drawn from the survey in 2016 from MIP database, which has been anonymously completed by various enterprises in different industries within German. To clarify, we have not personally conducted or have been part of the initial data collection process but chose to use this data set based upon of the following advantages. First, it is part of European Community statistics and together with the relative large sample size it supports generalizability of the research findings within German. Second, this large-scale data set allows us to structurally obtain large amounts of data in a fairly short time (de Leeuw, 2005). Third, the data used for this study is quite recent, which facilitates timely assessment of business values of BDA. Finally, the Global Innovation Index of 2016 (the year during which the data of this study has been collected) indicated that Germany entered the top 10 due to constituent performance in areas such as research & development (R&D), knowledge creation and respectable logistic performances. Comparable innovation driven countries at that time were (1) Switzerland, (2) Sweden and (3) United Kingdom. Hence, findings of this study are likely to be generalizable among other countries in the European Union that possess considerable innovation capacity.
into what factors enable firms to invest in their future or hinder them from doing so. But most importantly for this paper, only the 2016 survey provided thoroughly investigated information regarding organizational usage of digitalization. The latter formed the basis of the selection criteria, namely delivering relevant insights regarding firm’s usages of BDA, the appropriate IT resources and related business performances in form of labor productivity. These selection criteria limited the sample to a specific year. After neglecting missing observations and pre-selecting participants that generated a valid response for this study, a response rate of approximately 84% was established and let to a final sample of 3923 firms, out of a total sample of 4685 firms.
4.2. Measures
4.2.1. Technical IT Resources
Technical IT resources are measured by a survey question asking; ‘How strongly does your enterprise currently experience difficulties with IT infrastructure when trying to use the opportunities of digitalization?’ The variable is based on a 4-point scale, whereby 0=no, 1=low, 2=medium and 3=high. This measure emphasizes a firm’s overall understanding of their IT infrastructures. A low score would suggest that a firm is highly capable to integrate and operate their BDA related technologies. New digital technologies, shape novel infrastructures and influence organizational logic and patterns of coordination within and across firms (Bharadwaj et al., 2013). Specific IT infrastructures fuel the development of data analytics and enable innovative breakthroughs (Chen and Zhang, 2014). These scholars also argue that scalable, accessible and sustainable data infrastructures stimulate peoples understanding about human interactions and social processes. As consequence of this development, infrastructure vendors are offering different hardware and software systems that support the usage of BDA. Respectively, this growing interest represents a grounded reason further investigating technical IT resources.
4.2.2. Human IT Resources
Bharadwaj (2000) and Chen (2012), this paper measures specific human IT resources, by assessing technical and managerial IT skills of employees.
4.2.3. Big Data Analytics
For the BDA variable, we rely upon a survey question asking; ‘To what extent does your enterprise currently use BDA in different business function areas?’ Similar to the two variables above (technical and human IT resources), possible answers to this question are built upon a 4-point scale (0=no, 1=low, 2=medium, 3=high). This measure indicates the extensiveness of BDA practices within an organization. Whereby, a high score suggests that a firm is highly capable to integrate BDA practices throughout their organization intensively. Along with Bharadwaj (2000), the analytical mandate of Kiron et al. (2014) is followed to explore BDA practice. This mandate emphasizes that culture, skills and data management are the key dimensions of an analytical capability. Whereby, BDA practices are broadly defined as the competence to provide business insights by means of profound data management, technical infrastructure and talent to transform into a competitive force. Several studies have explored the actual execution of BDA. For example, Wamba et al. (2015) investigated how BDA practices are measured and how these are linked to firm performance. They explored the influence of infrastructure flexibility, management capabilities and personal expertise as elements for BDA. Gupta and George (2016) contributed to literature by developing theoretically grounded constructs of BDA practices and created a comprehensive survey instrument to measure firms BDA practices. Before, Chen et al. (2015) draw upon the dynamic capabilities theory to conceptualized BDA use. More recently, Dong and Yang (2018) studied the use of BDA to derive customer insights. Thus, prior literature demonstrates that BDA could be measured in various ways, however our single indicator of superior BDA practices applied in this study allows for generalizability of the findings among various firms cross and beyond Germany. 4.2.4. Labor Productivity
4.2.5. Firm Size
Within innovation literature firm size is one of the most commonly cited factors that could have a great impact on firm performance (Wu et al., 2006; Zhu and Kraemer, 2005). Large firms are commonly assumed to derive greater synergy effects from human and financial resources that lead to better firm performance. Therefore, firm size is used as moderator factor in this study and is measured by the number of full-time employees operating within an organization. As firm size is represented by continues variables, a natural logarithm transformation is performed to normalize the measures (Bartlett, 1947). Subsequently, an interaction term of BDA and firm size has been generated to identify the moderating effect on the relationship between BDA and labor productivity.
4.2.6. Control Variables
Table 1: Industry Dummies Classification
Industry Classification Frequency Percentage
1) Mining 197 4.20 2) Food/Tobacco 223 4.76 3) Textiles 157 3.35 4) Wood/Paper 151 3.22 5) Chemical 126 2.69 6) Plastics 154 3.29 7) Glass/Ceramics 101 2.16 8) Metals 332 7.09 9) Electrical equipment 292 6.23 10) Machinery 194 4.14 11) Retail/Automobile 102 2.18
12) Furniture/Toys/Medical Technology and Maintenance 273 5.83
13) Energy/Water 311 6.64
14) Wholesale 208 4.44
15) Transport equipment/Postal Service 409 8.73
16) Media services 246 5.25
17) IT/Telecommunications 187 3.99
18) Banking/Insurance 186 3.97
19) Technical service/R&D Services 310 6.62
20) Consulting/Advertisement 283 6.04
21) Firm-related services 243 5.19
Table 2: Descriptive Statistics
Mean SD
1) Labor Productivity 0.266 0.168 2) Technical IT Resource 1.446 1.050 3) Human IT Resource 1.275 1.046 4) Big Data Analytics 0.486 0.775 5) R&D intensity 0.009 0.029 6) Location 0.333 0.471 7) IT Flexibility 1.526 1.002 8) Size 3.345 1.478 Table 3: Correlations (1) (2) (3) (4) (5) (6) (7) 1) Labor Productivity 2) Technical IT Resource 0.038* 3) Human IT Resource 0.042** 0.402*** 4) Big Data Analytics 0.109*** 0.143*** 0.206*** 5) R&D intensity -0.038* 0.037* 0.113*** 0.139*** 6) Location -0.176*** 0.011 0.013 -0.032* 0.042** 7) IT Flexibility 0.093*** 0.580*** 0.511*** 0.144*** 0.068*** 0.001 8) Size 0.254*** 0.089*** 0.183*** 0.234*** 0.057*** -0.099*** 0.196***
5. Results 5.1. Hypotheses Testing
Since, all variables in this study contain only one singular measurement item, it is not necessary to estimate a separate measurement model by structural equation modeling (SEM). Therefore, ordinary least squares (OLS) regression has been performed to test our hypotheses. SEM could roughly be seen as equivalent to OLS regression, but one of the biggest concerns with SEM is its complexity (Xiao, 2013). Although, SEM has numerous advantages over OLS regression (e.g., at SEM a variable can be both predicted and explanatory, also it allows an inclusion of unobserved latent variables), when data satisfies the assumptions of OLS regression sufficiently and fitting abilities are similar, OLS regression should be preferred over SEM (Xiao, 2013). Even though SEM will fabricate essentially the same results, using a considerably more complex model will be considered overkill (Xiao, 2013). In order to perform these OLS regressions, the research model has been split into two separate parts. At the first part BDA is viewed as the dependent variable. Results of these OLS regressions for BDA are presented in table 4. At the second part labor productivity is considered as the dependent variable, of which the regression results for labor productivity are presented in table 5.
Table 4: OLS Regression Results (part 1) (1) (2) (3) Technical IT Resource 0.077*** (0.014) 0.066*** (0.014) Human IT Resource 0.079*** (0.013) R&D Intensity 2.130*** (0.413) 2.086*** (0.412) 1.951*** (0.412) Location -0.004 (0.025) -0.001 (0.025) -0.005 (0.025) IT Flexibility 0.082*** (0.012) 0.035* (0.014) 0.003 (0.015) Size 0.117*** (0.008) 0.118*** (0.008) 0.113*** (0.008)
Industry Yes Yes Yes
Constant -0.243 (0.086) -0.152 (0.090) -0.169 (0.091) R2 0.138 0.146 0.155 Adj. R2 0.133 0.141 0.149 F 25.150*** 25.660*** 26.210*** n 3,793 3,770 3,739
Note: * p < 0.05; ** p < 0.01; *** p < 0.001. Standard errors are in parentheses. Dependent variable is big data analytics.
Table 5: OLS Regression Results (part 2) (1) (2) (3) (4) Technical IT Resource -0.002 (0.003) -0.004 (0.003) -0.003 (0.003) Human IT Resources -0.001 (0.003) -0.003 (0.003) -0.003 (0.003)
Big data analytics 0.019***
(0.003)
-0.001 (0.009) Big data analytics
x Size 0.006* (0.002) R&D intensity -0.016 (0.085) -0.169* (0.086) -0.204* (0.087) -0.199* (0.087) Location -0.048*** (0.005) -0.048*** (0.005) -0.049*** (0.005) -0.049*** (0.005) IT Flexibility 0.005* (0.002) 0.007* (0.003) 0.008* (0.003) 0.008* (0.003) Size 0.022*** (0.002) 0.022*** (0.002) 0.019*** (0.002) 0.016*** (0.002)
Industry Yes Yes Yes Yes
Constant 0.205 (0.018) 0.211 (0.018) 0.220 (0.018) 0.228 (0.019) R2 0.241 0.239 0.244 0.245 Adj. R2 0.237 0.234 0.238 0.240 F 51.690*** 46.250*** 44.130*** 42.860*** n 3,923 3,854 3,722 3,722
Note: * p < 0.05; ** p < 0.01; *** p < 0.001. Standard errors are in parentheses. Dependent variable is firm performance expressed in labor productivity.
Table 6: Sobel test 1
Input Test statistic Std. Error p-value
(a) 0.112 Sobel test: 3.667 0.001 0.000
(b) 0.019 Aroian test: 3.635 0.001 0.000
(Sa) 0.020 Goodman test: 3.701 0.001 0.000
(Sb) 0.004
Note: Technical IT resources as independent variable in this test.
Next a second Sobel test has been performed to test whether human IT resources and labor productivity are mediated by BDA. Likewise, the second Sobel test (see table 7), exhibits statistical significance with p-value of 0.000. Thus, also H5 was supported.
Table 7: Sobel test 2
Input Test statistic Std. Error p-value
(a) 0.135 Sobel test: 3.885 0.001 0.000
(b) 0.019 Aroian test: 3.856 0.001 0.000
(Sa) 0.020 Goodman test: 3.914 0.001 0.000
(Sb) 0.004
Note: Human IT resources as independent variable in this test.
Finally, the moderating effect of firm size has been tested. The interaction term of BDA and firm size was included into the full control model with both IT resources and BDA (Table 5). This revealed a very strong moderating effect (b = 0.006, p < 0,011), as the significance of the main effects disappeared. Meaning that for large firms BDA has a stronger effect on labor productivity. Hence, H6 was supported. 5.2. Robustness Check
To address this issue, this paper applied the marker-variable technique, as Malhotra et al., (2006) advocate that this approach is one of the most effective methods to account for CMV and assess CMB. Consistent with Lindell and Whitney (2001) the marker-variable technique has been performed by using the smallest correlation (i.e. 0.001) and, more conventionally, the second smallest correlation (i.e. 0.011) as proxies of CMV. After eliminating CMV from the zero-order correlations between labor productivity and other variables, partial correlations remained statistically significant after decreases of up to 2% and 30% (see table 8). Hence, CMB is not viewed as substantial in the data and as such there is no real threat to the validity of this research (Doty and Glick, 1998).
Table 8: Assessment of Common Method Bias
Antecedents of labor productivity
Zero-order correlation
First smallest correlation as the proxy of CMV
Second smallest correlation as the proxy of CMV Percentage of change (%) Partial correlation Percentage of change (%) Partial correlation Technical IT resources 0.038* 2.29 0.037* 27.97 0.027 Human IT resources 0.042** 2.08 0.041** 25.39 0.031
Big data analytics 0.109*** 0.74 0.108*** 9.02 0.099***
RD -0.038* -2.49 -0.038* -30.49 -0.049*
Location -0.176*** -0.60 -0.177*** -7.35 -0.189***
IT Flexibility 0.093*** 0.88 0.092*** 10.80 0.083***
Size 0.254*** 0.26 0.253*** 3.24 0.246***
6. Discussion
6.1. Theoretical Implications
In this paper, we theorize and empirically examine antecedents of BDA and the influences of BDA on labor productivity. Our study adds to the theoretical understanding of the RBV of IT and explores the differences in labor productivity among SMEs and large firms. By doing so, we make several contributions to BDA and productivity research.
First, by following the RBV of IT, we provide empirical evidence that technical and human IT resources are significant enablers of superior BDA practices. This theoretical explanation contributes to our understanding of which antecedents are relevant to acquire superior BDA practices. These results are consistent with what has been found in previous studies (e.g., Barney, 1991; Kearns and Lederer, 2003; Wernerfelt, 1984). Hence, our findings support the proposition that technical and human IT resources provision BDA.
Second, we contribute by demonstrating that BDA practice is an enabler of enhanced firm’s labor productivity. Although past research had shown the impact of BDA on productivity, for example on asset productivity (Chen et al., 2015) or on productivity in terms of logistics and inventory management (Davenport and Harris, 2007), apart from one related study of Tambe (2014) no prior work has systematically demonstrated that labor productivity could be derived from superior BDA practices. Tambe (2014) documented that there is no superficial difference in firms’ labor productivity with the specific use of Apache Hadoop, which is the most widely known software platform of BDA. To the best of our knowledge, our study is the first to explore and document the impact of BDA practices, which goes beyond just software, on labor productivity.
Finally, Wade and Hulland (2004) proposed two characteristics of the RBV of IT that may offer rare and valuable benefits for literature related to the novel field of big data.First, RBV assists academics and practitioners to provide a solid foundation for identifying idiomatic firm level technical resources. Second, RBV of IT permits researchers to systematically test a relationship between firms’ internal resources and its performances. By drawing on the RBV of IT and by thoroughly applying these characteristics to BDA, we have fostered the generalizability of RBV of IT to the emerging field of BDA. This is consistent with prior research (e.g., Kozlenkova et al., 2014) revealing that given the solidity of RBV, the attention for RBV in research and practice is growing. In turn RBV is evolving through clarification, modification and transformations, of which RBV of IT presents a great example. 6.2. Practical Implications
6.3. Limitations and Future Research
This study entails some limitations, several of which may occasion opportunities for future research. First, although we have conducted this study upon a large scale, cross-industry sample from a recent year, the fact remains that reliance on a cross-sectional data limits this study. While this type of data succeeds to provide breadth, it fails to provide extensive depth (Nevo and Wade, 2011). Hence, alternative research methods such as ethnographic, which are scientific descriptions of nations, race, customs, habits and differences (Lofland and Lofland, 2001), and case studies, would enable researchers to track firms’ movements and identify the way firms manage their IT resources to optimize BDA. This would undoubtedly help to further elucidate the path from BDA practices to labor productivity. These kinds of studies may provide valuable insights upon how to generate a competitive advantage in a fast-changing IT world. Additionally, a cross-sectional design does not allow us to evaluate the sustainability of the enhanced firm performance associated with BDA nor does it allow us to test causality. Hence, future research could go beyond our initial findings by undertaking a longitudinal research to examine the durability of labor productivity and related competitive gains.
Second, in this study we measure all concepts using one single indicator. Although, the majority of studies related to IT literature employ just one single indictor (Hagedoorn and Cloodt, 2003), we encourage future research to engage multiple indicators to generate one construct. Hagedoorn and Cloodt (2003) highlighted the possible use of combining multiple indicators in a composite measure of performance. The benefit of a multi-indicator approach is that it allows for more complex, informative and composite measures rather than assuming ‘correctness’ of a single indicator (Hagedoorn and Cloodt, 2003).
7. Conclusion
Reference
Anderson, R. C., & Reeb, D. M. (2003). Founding-family ownership and firm performance: evidence from the S&P 500. The Journal of Finance, 58(3), 1301-1328.
Assunção, M. D., Calheiros, R. N., Bianchi, S., Netto, M. A., & Buyya, R. (2015). Big Data computing and clouds: Trends and future directions. Journal of Parallel and Distributed Computing, 79, 3-15. Baesens, B., Bapna, R., Marsden, J. R., Vanthienen, J., & Zhao, J. L. (2016). Transformational issues of big data and analytics in networked business. MIS quarterly, 40(4).
Balk, B. M. (2001). Scale efficiency and productivity change. Journal of Productivity Analysis, 15(3), 159-183.
Barney, J. B. (1991). Firm resources and sustained competitive advantage. Journal of Management, 17(1), 99-120.
Barney, J. B. (2014). How marketing scholars might help address issues in resource-based theory. Journal of the Academy of Marketing Science, 42(1), 24-26.
Bartlett, M. S. (1947). The use of transformations. Biometrics, 3(1), 39-52.
Barton, D., & Court, D. (2012). Making advanced analytics work for you. Harvard Business Review, 90(10), 78-83.
Barua, A., Konana, P., Whinston, A. B., & Yin, F. (2004). An empirical investigation of net-enabled business value. MIS Quarterly, 28(4), 585-620.
Bharadwaj, A. S. (2000). A resource-based perspective on information technology capability and firm performance: an empirical investigation. MIS Quarterly, 169-196.
Bharadwaj, A., El Sawy, O., Pavlou, P., & Venkatraman, N. (2013). Digital business strategy: toward a next generation of insights. MIS Quarterly, 471-482.
Brown, B., Chui, M., & Manyika, J. (2011). Are you ready for the era of ‘big data’? McKinsey Quarterly, 4(1), 24-35.
Brynjolfsson, E., & Yang, S. (1996). Information technology and productivity: a review of the literature. Advances in Computers (Vol. 43, pp. 179-214). Elsevier.
Capon, N., & Glazer, R. (1987). Marketing and technology: a strategic co-alignment. The Journal of Marketing, 1-14.
Chen, J. L. (2012). The synergistic effects of IT-enabled resources on organizational capabilities and firm performance. Information & Management, 49(3-4), 142-150.
Chen, C. P., & Zhang, C. Y. (2014). Data-intensive applications, challenges, techniques and technologies: A survey on Big Data. Information Sciences, 275, 314-347.
Chen, H., Chiang, R. H., & Storey, V. C. (2012). Business intelligence and analytics: From big data to big impact. MIS Quarterly, 36(4).
Chen, Q., Hsu, M., & Zeller, H. (2011, March). Experience in Continuous analytics as a Service (CaaaS). Proceedings of the 14th International Conference on Extending Database Technology (pp. 509-514). ACM.
Chen, M., Mao, S., & Liu, Y. (2014). Big data: A survey. Mobile Networks and Applications, 19(2), 171-209.
Chen, D. Q., Preston, D. S., & Swink, M. (2015). How the use of big data analytics affects value creation in supply chain management. Journal of Management Information Systems, 32(4), 4-39.
Copeland, D. G., & McKenney, J. L. (1988). Airline reservations systems: lessons from history. MIS Quarterly, 353-370.
Court, D. (2015). Getting big impact from big data. McKinsey Quarterly, 1, 52-60.
Davenport, T. H., & Harris, J. G. (2007). Competing on analytics: The new science of winning. Harvard Business Press.
De Leeuw, D. (2005). To mix or not to mix data collection modes in surveys. Journal of Official Statistics, 21(2), 233.
Debortoli, S., Müller, O., & vom Brocke, J. (2014). Comparing business intelligence and big data skills. Business & Information Systems Engineering, 6(5), 289-300.
Demirkan, H., & Delen, D. (2013). Leveraging the capabilities of service-oriented decision support systems: Putting analytics and big data in cloud. Decision Support Systems, 55(1), 412-421.
Dong, J. Q., & Netten, J. (2017). Information technology and external search in the open innovation age: New findings from Germany. Technological Forecasting and Social Change, 120, 223-231. Dong, J. Q., & Yang, C. (2018). Business Value of Big Data Analytics: The Case of Social Media Analytics. Information and Management, forthcoming.
Doty, D. H., & Glick, W. H. (1998). Common methods bias: does common methods variance really bias results? Organizational Research Methods, 1(4), 374-406.
Edmunds, A., & Morris, A. (2000). The problem of information overload in business organisations: a review of the literature. International Journal of Information Management, 20(1), 17-28.
Eisenhardt, K. M., & Martin, J. A. (2000). Dynamic capabilities: what are they? Strategic Management Journal, 1105-1121.
Geroski, P. A. (1998). An applied econometrician's view of large company performance. Review of Industrial Organization, 13(3), 271-294.
Grant, R. M. (1991). “The Resource-based Theory of Competitive Advantage: Implications for Strategy Formulation.” California Management Review 33 (3): 114–135.
Griliches, Z., & Regev, H. (1995). Firm productivity in Israeli industry 1979–1988. Journal of Econometrics, 65(1), 175-203.
Gupta, M., & George, J. F. (2016). Toward the development of a big data analytics capability. Information & Management, 53(8), 1049-1064.
Halkos, G. E., & Tzeremes, N. G. (2007). Productivity efficiency and firm size: An empirical analysis of foreign owned companies. International Business Review, 16(6), 713-731.
Hilbert, M. (2016). Big data for development: A review of promises and challenges. Development Policy Review, 34(1), 135-174.
Hitt, L. M., & Brynjolfsson, E. (1996). Productivity, business profitability, and consumer surplus: three different measures of information technology value. MIS Quarterly, 121-142.
Jones, C. I., & Williams, J. C. (1998). Measuring the social return to R&D. The Quarterly Journal of Economics, 113(4), 1119-1135.
Kearns, G. S., & Lederer, A. L. (2003). A resource-based view of strategic IT alignment: how knowledge sharing creates competitive advantage. Decision Sciences, 34(1), 1-29.
Kemery, E. R., & Dunlap, W. P. (1986). Partialling factor scores does not control method variance: A reply to Podsakoff and Todor. Journal of Management, 12(4), 525-530.
Kim, G., Shin, B., Kim, K. K., & Lee, H. G. (2011). IT capabilities, process-oriented dynamic capabilities, and firm financial performance. Journal of the Association for Information Systems, 12(7), 487.
Kiron, D., Prentice, P. K., & Ferguson, R. B. (2014). The analytics mandate. MIT Sloan Management Review, 55(4), 1.
Koch, M. J., & McGrath, R. G. (1996). Improving labor productivity: Human resource management policies do matter. Strategic Management Journal, 335-354.
Kozlenkova, I. V., Samaha, S. A., & Palmatier, R. W. (2014). Resource-based theory in marketing. Journal of the Academy of Marketing Science, 42(1), 1-21.
Lofland, J., & Lofland, L. (2001). Handbook of ethnography. P. Atkinson, A. Coffey, & S. Delamont (Eds.). London: Sage.
Lucas Jr, R. E. (1988). On the mechanics of economic development. Journal of Monetary Economics, 22(1), 3-42.
Malhotra, N. K., Kim, S. S., & Patil, A. (2006). Common method variance in IS research: A comparison of alternative approaches and a reanalysis of past research. Management Science, 52(12), 1865-1883. Mata, F. J., Fuerst, W. L., & Barney, J. B. (1995). Information technology and sustained competitive advantage: A resource-based analysis. MIS Quarterly, 487-505.
McAfee, A., Brynjolfsson, E., Davenport, T. H., Patil, D. J., & Barton, D. (2012). Big data: the management revolution. Harvard Business Review, 90(10), 60-68.
McKenney, J. L., Mason, R. O., & Copeland, D. C. (1995). Waves of change: Business evolution through information technology. Harvard Business School Press.
Mukhopadhyay, T., Lerch, F. J., & Mangal, V. (1997). Assessing the impact of information technology on labor productivity A field study. Decision Support Systems, 19(2), 109-122.
Ness, L. R. (2005). Assessing the relationships among IT flexibility, strategic alignment, and IT effectiveness: study overview and findings. Journal of Information Technology Management, 16(2), 1-17.
Nevo, S., & Wade, M. (2011). Firm-level benefits of IT-enabled resources: A conceptual extension and an empirical assessment. The Journal of Strategic Information Systems, 20(4), 403-418.
Newbert, S. L. (2008). Value, rareness, competitive advantage, and performance: a conceptual-level empirical investigation of the resource-based view of the firm. Strategic Management Journal, 29(7), 745-768.
Podsakoff, P. M., MacKenzie, S. B., Lee, J. Y., & Podsakoff, N. P. (2003). Common method biases in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88(5), 879.
Raghupathi, W., & Raghupathi, V. (2014). Big data analytics in healthcare: promise and potential. Health Information Science and Systems, 2(1), 3.
Rai, A., Patnayakuni, R., & Patnayakuni, N. (1997). Technology investment and business performance. Communications of the ACM, 40(7), 89-97.
Ramírez, Y. W., & Nembhard, D. A. (2004). Measuring knowledge worker productivity: A taxonomy. Journal of Intellectual Capital, 5(4), 602-628.
Reed, R., & DeFillippi, R. J. (1990). Causal ambiguity, barriers to imitation, and sustainable competitive advantage. Academy of Management Review, 15(1), 88-102.
Ren, S.J., Wamba, S.F., Akter, S., Dubey, R., & Childe, S. J. (2017). Modelling quality dynamics, business value and firm performance in a big data analytics environment. International Journal of Production Research, 55(17), 5011-5026.
Ross, J. W., Beath, C. M., & Goodhue, D. L. (1996). Develop long-term competitiveness through IT assets. Sloan Management Review, 38(1), 31.
Rumelt, R. P. (1982). Diversification strategy and profitability. Strategic Management Journal, 3(4), 359-369.
Schwartz, B. (2004). The paradox of choice: Why more is less (Vol. 6). New York: HarperCollins. Shah, S., Horne, A., & Capellá, J. (2012). Good data won't guarantee good decisions. Harvard Business Review, 90(4).
Siemens, G., & Long, P. (2011). Penetrating the fog: Analytics in learning and education. Educause Review, 46(5), 30.
Sobel, M. E. (1982). Asymptotic confidence intervals for indirect effects in structural equation models. Sociological Methodology, 13, 290-312.
Stocker, A., Gerold, S., Hinterberger, F. (2015). The interaction of resource and labour productivity. Bio by Deloitte. Retrieved from http://ec.europa.eu/environment/enveco/jobs/pdf/Study%20Resource %20labour%25
Tambe, P. (2014). Big data investment, skills, and firm value. Management Science, 60(6), 1452-1469. Teece, D. J., Pisano, G., & Shuen, A. (1997). Dynamic capabilities and strategic management. Strategic Management Journal, 18(7), 509-533.
Tourangeau, R., Rips, L. J., & Rasinski, K. (2000). The psychology of survey response. Cambridge University Press.
Tushman, M. L., Kahn, A., Porray, M. E., & Binns, A. (2017). Data can do for change management what it did for marketing. Harvard Business Review. Retrieved from https://hbr.org/2017/07/data-can-do-for-change-management-what-it-did-for-marketing
Villars, R. L., Olofson, C. W., & Eastwood, M. (2011). Big data: What it is and why you should care. White Paper, IDC, 14.
Wade, M., & Hulland, J. (2004). The resource-based view and information systems research: Review, extension, and suggestions for future research. MIS Quarterly, 28(1), 107-142.
Wamba, S. F., Akter, S., Edwards, A., Chopin, G., & Gnanzou, D. (2015). How ‘big data’ can make big impact: Findings from a systematic review and a longitudinal case study. International Journal of Production Economics, 165, 234-246.
Wamba, S. F., Gunasekaran, A., Akter, S., Ren, S. J. F., Dubey, R., & Childe, S. J. (2016). Big data analytics and firm performance: Effects of dynamic capabilities. Journal of Business Research, 70, 356-365.
Whyte, J., Stasis, A., & Lindkvist, C. (2016). Managing change in the delivery of complex projects: Configuration management, asset information and ‘big data’. International Journal of Project Management, 34(2), 339-351.
Wu, F., Yeniyurt, S., Kim, D., & Cavusgil, S. T. (2006). The impact of information technology on supply chain capabilities and firm performance: A resource-based view. Industrial Marketing Management, 35(4), 493-504.
Xiao, X. (2013). Structural equation modeling compared with ordinary least squares in simulations and life insurers’ data. University of Texas at Austin.
Zakir, J., Seymour, T., & Berg, K. (2015). Big Data Analytics. Issues in Information Systems, 16(2). Zelenkauskaite, A., & Simões, B. (2014). Big data through cross-platform interest-based interactivity. In Big Data and Smart Computing (BIGCOMP), 2014 International Conference on (pp. 191-196). IEEE.
Zheng, J., Liu, X., & Bigsten, A. (2003). Efficiency, technical progress, and best practice in Chinese state enterprises (1980–1994). Journal of Comparative Economics, 31(1), 134-152.
