Homework Help Needed!!

How the Use of Big Data Analytics Affects Value Creation in Supply Chain Management

DANIEL Q. CHEN, DAVID S. PRESTON, AND MORGAN SWINK

DANIEL Q. CHEN (corresponding author; [email protected]) is an associate professor in the Department of Information Systems and Supply Chain Management in the Neeley School of Business at Texas Christian University. He received his Ph.D. in MIS from the University of Georgia and also holds an M.B.A. degree from Washington University in St. Louis. His research interests lie across the areas of strategic management, supply chain management, and technology management. His work has appeared or is forthcoming in Journal of Management Information Systems, MIS Quarterly, Decision Sciences, Decision Support Systems, Journal of Operations Management, Journal of Strategic Information Systems, MIS Quarterly Executive, and others.

DAVID S. PRESTON is an associate professor in the Department of Information Systems and Supply Chain Management in the Neeley School of Business at Texas Christian University. He received his Ph.D. in MIS from the University of Georgia and also holds an M.B.A. from the University of Georgia. His research interests include information systems (IS) in supply chain management, IS leadership, and IS strate- gic alignment. His work has been published or is forthcoming in a variety of journals, including Information Systems Research, Journal of Management Information Systems, MIS Quarterly, Decision Sciences, Journal of Operations Management, IEEE Transactions on Engineering Management, Journal of Strategic Information Systems, MIS Quarterly Executive, and others.

MORGAN SWINK is the Eunice and James L. West Chaired Professor of Supply Chain Management in the Neeley School of Business at Texas Christian University. He also serves as the executive director of the Supply and Value Chain Center in the Neeley School. He is the former coeditor in chief of the Journal of Operations Management. He has coauthored two textbooks on supply chain operations, one managerial book on supply chain excellence, and more than 50 articles in academic and managerial journals.

ABSTRACT: Despite numerous testimonials of first movers, the underlying mechan- isms of organizations’ big data analytics (BDA) usage deserves close investigation. Our study addresses two essential research questions: (1) How does organizational BDA usage affect value creation? and (2) What are key antecedents of organiza- tional-level BDA usage? We draw on dynamic capabilities theory to conceptualize BDA use as a unique information processing capability that brings competitive advantage to organizations. Furthermore, we employ the technology–organization– environment (TOE) framework to identify and theorize paths via which factors

Journal of Management Information Systems / 2015, Vol. 32, No. 4, pp. 4–39.

Copyright © Taylor & Francis Group, LLC

ISSN 0742–1222 (print) / ISSN 1557–928X (online)

DOI: 10.1080/07421222.2015.1138364

influence the actual usage of BDA. Survey data collected from 161 U.S.-based companies show that: organizational-level BDA usage affects organizational value creation; the degree to which BDA usage influences such creation is moderated by environmental dynamism; technological factors directly influence organizational BDA usage; and organizational and environmental factors indirectly influence organizational BDA usage through top management support. Collectively, these findings provide a theory-based understanding of the impacts and antecedents of organizational BDA usage, while also providing guidance regarding what managers should expect from usage of this rapidly emerging technology.

KEY WORDS: and phrases: big data, data analytics, dynamics capability theory, structural equation modeling, survey research, TOE framework.

The emergence of networked business has dramatically enhanced the volume, variety, and velocity of unstructured as well as structured information, often described via the term “big data.” Managers are seeking to more effectively exploit such data gathered and created within their organizations and also by business partners and third parties (i.e., including commercial data aggregators and government agencies). Firms that use big data effectively are able to convert data into insights and intelligence delivered when and where they are needed throughout various levels of the organization. Anecdotal evidence has shown that the insights derived from big data have the potential to transform business strategies and business models and thereby improve marketing, pro- duct and service development, human resources (HR), operations, and other core business functions.1 Based on this evidence, there have been advances in an array of powerful analytical techniques and information technologies enabling companies to capitalize on the promise of big data [13]. The current study examines the organizational use of big data analytics

(BDA), defined as the process of using advanced technologies to examine big data in order to uncover useful information (e.g., hidden patterns, unknown correlations, etc.) to help with making better decisions across business pro- cesses among functions or companies [84]. Despite successful testimonials of “big data first movers,” a recent industry survey indicates that a majority of companies have still not begun to engage in the practice of capitalizing on big data [60]. Many organizations appear to still be in the stage of learning about the value of big data, the necessary information technology (IT) and analytical skills, the risks involved, and how to make a compelling business case for necessary substantial investments [5]. Therefore, the underlying mechanisms that lead to organizations’ BDA usage, as well as the performance outcomes of such usage, deserve close investigation. Although an extensive body of IT adoption literature exists at the individual

level [20, 30, 82], studies on technology adoption behavior at the organiza- tional level remain relatively scarce, despite the importance of understanding

BIG DATA ANALYTICS IN SUPPLY CHAIN MANAGEMENT 5

this phenomenon [70, 88]. Furthermore, the limited number of organizational IT adoption studies that do exist generally focus on adoption intentions [17, 38, 76], rather than postadoption behaviors (i.e., usage and value creation) [89]. In addition, to comprehend the drivers of BDA usage, managers and researchers also need to better understand the potential business performance impacts of BDA competency. As we have stated, a growing popular press recounts stories of particular applications of BDA in specific functional areas, mostly in market- ing and customer relationship management [60]. However, a more holistic view of BDA and associated capabilities has not been advanced. In summary, the above outlined gaps in the literature limit the current under-

standing of the organizational-level usage of BDA, as well as the related value creation processes. Our study addresses the following research questions in order to bridge these gaps: (1) How does organizational BDA usage influence business productivity and growth? and (2) What factors are key drivers of organizational-level BDA usage? To answer these research questions, we exam- ine BDA usage in the domain of supply chain management (SCM). This is a particularly important domain because firms are increasingly dependent on organizational learning and knowledge creation to achieve SCM success [36, 37, 84]. Furthermore, IT and particularly BDA play instrumental roles in satisfying needed information flows and learning demands in today’s networked economy [44, 84]. The overarching theoretical lens through which we develop the conceptual

model is the theory of dynamic capabilities [25]. We contend that an organiza- tion’s applications of BDA across a breadth of supply chain domains enable greater dynamic information processing capability, which in turn endows the organizational decision makers with knowledge enabling them to produce better resource configurations and reconfigurations that create competitive advantages. Such capability is likely to be especially valuable in highly dynamic environments characterized by high levels of uncertainty. We also incorporate the technology– organization–environment (TOE) model [80] to explain forces that potentially shape organizational initiatives to BDA use. In particular, we extend the TOE model by examining the application of this model with BDA in mind and consider interrelationships with specific TOE elements and managerial factors, thus devel- oping a more complete model of organizational-level IT usage. A test of these propositions using survey data provided by supply chain

managers from 161 U.S.-based companies provides a needed empirical demon- stration of BDA’s benefits, along with managerial implications regarding how BDA usage should be measured, and what operational performance improve- ments should be targeted. As a whole, the integrative research model for BDA usage and its influence developed in our research study offers an original theoretical grounding and explanation in an emerging area of practice that has become widely desirable, yet continues to want for theoretical depth.

6 CHEN, PRESTON, AND SWINK

Theoretical Background

As described previously, the dynamic capabilities theory in strategic management literature coupled with the TOE framework in IT management literature is the basis for the development of the conceptual model (see Figure 1) for this research. In this section, we briefly discuss the theoretical elements of the conceptual model. The next section explains the research model and hypotheses.

The Emergence of Big Data and Big Data Analytics

Big data has been defined as high-volume (large scale), high-velocity (moving/ streaming), and high-variety (e.g., numerical, text, video, etc.) information assets that demand cost-effective, innovative forms of information processing for enhanced insight and decision making [29]. Big data analytics has recently risen to potential prominence due to greater ability to both capture vast amounts of data and employ more powerful analytical techniques to vast data sets. This recent ability of firms to be able to both collect big (and varied) data and also apply powerful analytical techniques to such data enables the organization to automate highly complex decisions that have traditionally been dependent (primarily or solely) on human judgment and intuition [6, 29]. Despite the purported bandwagon for big data, recent industry reports show that

many chief information officers (CIOs) and business executives have hesitated to make major investments in big data analytics, particularly after either directly experiencing or indirectly observing (i.e., other firms) prior past initiatives aimed at gathering business intelligence (often from terabytes of information), with dis- appointing results [56]. Some CIOs and business executives likely question whether BDA is simply a repackaging of traditional business intelligence and data mining

Big Data Analytics Use

Organizational Value Creation

Environmental Factors

TMT Support

Organizational Factors

Technological Factors

.

T-O-E Framework

Dynamic Capabilities Theory

Figure 1. Conceptual Model

BIG DATA ANALYTICS IN SUPPLY CHAIN MANAGEMENT 7

processes, or whether it truly embodies new capabilities that justify significant investments and offer competitive advantages. To help practitioners as well as academics answer this question, it is essential to

understand some of the differences between big data analytics and traditional business intelligence tools. According to a June 2011 IDC (International Data Corporation) report, since 2005 the quantity of universally available data has been growing at a rate of more than 50 percent a year, and it is expected to exceed 8,000 exabytes by 2015.2 A significant part of this growth is driven by unstructured data, including web content, news feeds, social media postings, video clips, and other data that cannot easily be grouped into recurring fields. Big data is therefore an all- encompassing term for collections of data sets with appreciable levels of size and complexity that they become difficult to capture, process, and manage in a timely fashion using on-hand data management tools and traditional data processing appli- cations [68]. As such, big data analytics has been recognized as a radical departure from traditional business intelligence tools [29].3

It is important to note that big data and analytics connote different precise ideas to different people [86]. The opportunities associated with data and analysis in different organizations have motivated significant advancements of BDA technologies that analyze critical business data to help an enterprise better understand its business and market, and to make timely business decisions. In addition to innovative data processing and analytical technologies, BDA also embraces business-centric prac- tices and methodologies that can be applied to various high-impact applications such as e-commerce, market intelligence, e-government, health care, and security [13]. Although the evolution of BDA is still in its early stage, the literature suggests that, for many organizations, what is considered BDA varies depending on the capabil- ities of the organization managing the data set, and on the capabilities of the applications that are used to process and analyze the data set in various business domains. As such, we conceptualize BDA usage as an important enterprise cap- ability that organizations could leverage to create cutting-edge knowledge in a dynamic environment [31]. From this perspective, the dynamic capabilities theory is a particularly useful lens to understand the impacts of BDA.

Big Data Analytics Usage as an Enterprise Dynamic Capability

Researchers have developed the notion of dynamic capabilities [25, 75] to describe how firms integrate, build, and reconfigure internal and external competences to address rapidly and unpredictably changing environments. Dynamic capabilities can also be defined as unique organizational processes—specifically the processes to integrate, reconfigure, gain, and release resources—used to match and even create market change [25]. Dynamic capabilities are thus the organizational and strategic routines by which firms achieve new resource configurations as markets emerge, collide, split, evolve, and die.

8 CHEN, PRESTON, AND SWINK

We conceptualize organizational use of BDA as creating dynamic capabilities for two reasons. First, using BDA helps organizations establish knowledge creation routines particularly when market dynamism is high. Such knowledge creation routines are essential dynamic capabilities identified in the literature [25]. More specifically, the use of BDA can be viewed as an organizational information processing capability [28] that reduces uncertainty by stimulating insights and knowledge creation, and increases organizational capability for strategic decision making. Second, this view of BDA usage also reflects the key characteristics of dynamic

capabilities suggested by the literature: commonalities in key features, coupled with idiosyncrasy in details [25]. In terms of commonalities in features, we note that, even in an early stage of development, there already exist numerous industry reports describing the features of various BDA tools,4 as well as consulting practices advising companies on how to explore the opportunities presented by BDA. In other words, there are many identifiable “best practices” for firms to select and implement common BDA applications for various purposes [13]. However, the existence of commonality among BDA and its use does not imply that any particular BDA tools will be employed across firms in exactly the same ways. Similar to the deployment of any organizational-level technologies, the use of BDA must be embedded into and assimilated with SCM processes. Given the complexity of varying SCM processes (e.g., sourcing, purchasing, network design, inventory optimization, customer services, etc.), organizational use of common BDA technol- ogies in supply chain management is idiosyncratic across firms. Therefore, viewing BDA use as a dynamic capability makes theoretical sense and such conceptualiza- tion helps with understanding the implications of BDA usage on organizational value creation.

The TOE Framework and BDA Use

Dynamic capabilities are often characterized as unique and distinctive processes that emerge from path-dependent histories of individual firms [75]. This concept of path dependency suggests that the forces driving organizational capabilities are context- specific [25]. Recent development in dynamic capabilities literature suggests two categories of contextual forces: technical and external fitness [34]. In line with these arguments from dynamic capabilities literature, we employ the technological, orga- nizational, and environmental (TOE) framework [80] in the IT management litera- ture to identify unique factors explaining organizational movement toward BDA usage in the SCM context. Indicative of its moniker, the TOE framework posits that various contextual factors influencing the organizational process of adoption and implementation of a technological innovation stem from technological, organiza- tional, and environmental facets of a firm’s context. The TOE framework has generally been empirically supported in the literature,

although the specific measures identified and used within the three contexts vary

BIG DATA ANALYTICS IN SUPPLY CHAIN MANAGEMENT 9

across different studies [8]. Despite the acceptance of this framework as it was originally developed, we find two salient limitations of the existing TOE studies. First, these prior TOE studies have predominantly treated each of the three cate- gories of contextual factors as working in similar fashion to directly impacting an organization’s intention to adopt IT. This treatment reflects a mechanistic perspective of organizations [77]. That is, organizational decision making is viewed as a fully rational process of finding an optimal choice given the information available, with- out considering the fact that in decision making, the rationality of individuals is limited by the information they have at hand, by their cognitive limitations, and by the finite amount of time they have to make a decision [67]. The second limitation of the existing studies is that they have treated each of the three factors separately, ignoring potential differential ways in which TOE factors can influence outcomes. While providing a clean baseline investigation of empirical support for the TOE framework, this approach overlooks the complexity of varying paths via which key contextual factors can affect organizational outcomes. For example, it has been noted that some TOE factors may have stronger effects on organizational IT adoption under certain contextual circumstances and contingencies [81]. In order to offer a more comprehensive picture of the mechanisms driving

organizational BDA use decisions, we propose two extensions of the TOE frame- work. First, we incorporate the behavioral perspective of organizations [54], parti- cularly the role of bounded rational human agency (i.e., organizational decision makers) in translating (i.e., mediating) external forces into managerial actions. Second, we examine the varying paths leading to organizational BDA use from TOE factors. Specifically, we modify the originally developed TOE model to the applicable BDA context by positioning top management support as a mediator between two of the TOE factors (i.e., organizational and environmental) and orga- nizational BDA use (see Figure 1). In other words, we suggest that TOE factors can result in BDA usage via two paths: (1) the characteristics of technologies will have a direct impact; and (2) the impacts of environmental and organizational forces will be mediated via top management behavior.

Research Model and Hypotheses Development

Based on the conceptual model, we develop a research model composed of testable hypotheses (Figure 2) to assess the use and value of BDA in the supply chain management domain of organizations. As can be observed in Figure 2, we examine two key aspects of business value generated by BDA use within the supply chain practices, asset productivity and business growth, as well as the moderating effects of environmental dynamism on the relationship between BDA use and both components of business value. For the factors that are posited to influence BDA use, we identify: (a) two salient technological factors (expected benefits and compatibility); (b) an organizational factor (organiza- tional readiness); and (c) an environmental factor (competitive pressure). In

10 CHEN, PRESTON, AND SWINK

addition, we theorize that the technological factors directly impact BDA use and the influence of both the organizational and environmental factors on BDA use are mediated by top management (TMT) support.

Big Data Analytics Use and Value Creation in Supply Chain Management

Traditionally regarded as a research domain of operations management, SCM has more recently received appreciable attention from information systems (IS) scholars because SCM can also be conceptualized as a collection of digitally enabled interfirm processes [62]. The literature clearly demonstrates that IT-based SCM systems enable a focal firm to integrate the flow of information, materials, and finances with its supply chain partners thereby contributing to organizational outcomes [21, 62]. As such, we contend that the SCM domain provides a useful context to examine BDA use. The focus of the current study does not pertain to SCM systems, but to the BDA

tools that are employed to process data/information generated by various SCM systems. The SCM systems examined in prior IS literature have been categorized as either transactional (e.g., purchase ordering systems, online bidding, etc.) or relational (e.g., private B2B exchange, customer relationship management, etc.) systems [40, 74]. However, the performance impacts of BDA that do not produce but process such transactional and relational information (generated from various SCM systems) have not been empirically examined. Since vast amounts of operational, tactical, and

Big Data Analytics Use

Asset Productivity

Competitive Pressure

Environmental Dynamism

Business Growth

H2a

H1b

H3

H5

H1a

H6

H7

H4

H2b

TMT Support

Organizational Readiness

Expected Benefits

Technology Compatibility

Technological

Organizational

Environmental

Figure 2. Research Model

BIG DATA ANALYTICS IN SUPPLY CHAIN MANAGEMENT 11

strategic information are routinely shared across stages of the supply chain, there is greater need for businesses to make sense of the transactional and relational informa- tion generated from the SCM systems for supply chain success [44]. We suggest that big data analytics, with the capability to convert data into insights and intelligence, has the potential to affect supply chain performance [13]. Following prior IT management literature, we assess the value impacts of BDA use

at the business process level [74]. We define BDA use as the extent to which BDA has been used to generate business insights across primary supply chain activities (e.g., sourcing, purchasing, production, distribution, and customer service). According to the dynamic capabilities perspective [25], we assert that BDA use is instrumental in deriving organizational competitive advantage. Specifically, we examine the impacts of BDA use to supply chain performance along two key dimensions within SCM: asset productivity and business growth.

Asset Productivity and Business Growth

As noted earlier, the functionality of dynamic capabilities is likely to be common (e.g., similar BDA technologies that can be obtained in the open market); therefore, the value of competitive advantage does not lie in the capabilities themselves, but in the resource configurations that such capabilities create [25, 75]. In particular, based on the logic of opportunity [46], dynamic capabilities are frequently used to build new resource configurations in the pursuit of a continual series of temporary advantages. The potential for long-term competitive advantage lies in “using dynamic capabilities sooner, more astutely, or more fortuitously than the competition to create resource configurations that have that advantage” [25, p. 1117]. One such temporary advantage in the supply chain context is asset productivity [73]. Asset productivity is a primary measure used to assess supply chain performance, describing the extent to which a business productively uses both current assets (e.g., cash, inventory) and fixed assets (e.g., plant, property, and equipment). Important and established indicators of asset productivity are the asset turnover rate (i.e., sales/assets) and profitability (i.e., return on assets). In accordance with the dynamic capabilities perspective, we view BDA use as a means

by which organizations can develop information processing capabilities that enable them to interpret and combine information collected from various sources and to direct this synthesized information to appropriate decision makers within functional supply chain departments [65]. In doing so, insights generated from BDA can potentially reduce uncertainties regarding demands, capacities, and supply availability. The absence of such capabilities dictates the need for costly asset-intensive buffers, such as cash, inventory, and excess capacity [27]. These assets exist as hedges against uncertainty. We expect that BDA use enables organizations to make more accurate and robust resource configurations by virtue of being able to make better predictions about future resource requirements. An often-quoted mantra in supply chain management is “informa- tion replaces inventory” [85]. This idea can be extended to other safety stocks (e.g., cash

12 CHEN, PRESTON, AND SWINK

held as a buffer, planned underutilization of equipment) that are created in order to deal with uncertain variability in the demand on and supply of resources. Similarly, we suggest that insights developed through BDA usage create opportu-

nities for organizations to reconfigure their resources in ways that are in greater alignment with trends and shifts in both demand and supply markets. For example, BDA enables faster and more complete information processing, leading to more accurate predictions that give asset managers greater advanced notice of the need to scale asset resources up or down [85]. Such insights can potentially lead to better asset utilization over time. Furthermore, extensive use of BDA across areas of SCM can stimulate interorgani-

zational learning between a firm and its supply chain partners [44, 75]. For example, analysis of data describing procurement and delivery patterns can potentially lead to greater optimization of transportation resources in ways that increase the use of transportation assets. In similar ways, production schedules can be enhanced to more fully use production assets. Collectively, such improvements can potentially be reflected in higher asset turnover rates and other indicators of asset productivity. Thus, we hypothesize:

Hypothesis 1a: Big data analytics use is positively associated with asset productivity.

We next evaluate the importance of BDA for organizational growth opportunities. Business growth in a dynamic market is a function of the capability of creating a series of temporary advantages. Long-term competitive advantage is difficult to obtain simply by enhancing existing resource configurations in dynamic markets [25]. By offering new insights in various areas (e.g., customer insights, marketing, operations, inventory man- agement, etc.), the use of BDA enhances the innovation opportunity for organizations to continually develop a series of temporary advantages in supply chain practices. For example, analysis of point of sale (POS) data can potentially lead to more attractive pricing/service offerings for specific customer segments. Similarly, analysis of inventory and shipping data can potentially create opportunities for lead time reduction and better product availability, thus increasing sales. The extant literature indicates that firms often seek to derive insights from opera-

tional data in order to create supply chain process improvements. For example, Lummus et al. [52] demonstrated that managers strongly associate analysis of demand and inventory data with greater responsiveness in the supply chain. Managers indi- cated that such information processing capability can enable their organizations to better anticipate and exploit rapidly emerging business opportunities. In addition, several research studies maintain that greater information visibility and processing capabilities generate greater supply chain responsiveness [16, 48, 87]. Thus, these capabilities can potentially make managers more alert to opportunities for driving business growth [22, 35]. Accordingly, we hypothesize:

Hypothesis 1b: Big data analytics use is positively associated with business growth.

BIG DATA ANALYTICS IN SUPPLY CHAIN MANAGEMENT 13

Environmental Dynamism

Environmental dynamism is a key situational parameter in dynamic capabilities theory, which suggests that the variance of competitive advantage created by organizational capability is contingent on environmental dynamism [25]. Some prior research has espoused the idea that the outcomes of dynamic capabilities in high-velocity markets are unpredictable. Eisenhardt and Martin [25] asserted that in moderately dynamic markets, firms generally follow predictable and linear paths (with these markets characterized by stable industry structures with defined market boundaries). As such, effective dynamic capabilities in moderately dynamic envir- onments rely on exploiting existing knowledge. In contrast, changes in high-velocity markets are generally nonlinear and less predictable and these markets are character- ized by volatile industry structures (i.e., market boundaries and players are ambig- uous and shifting) [55]. Despite the assessment of some researchers with regard to the unpredictable nature

of environmental dynamism on organizational outcomes, we posit that such an environment will provide greater opportunity for firms to capitalize on BDA. Prior research has shown that a turbulent external environment can either enhance or destroy a firm’s most critical competencies [1]. We contend that it is highly likely that the influence of BDA use on supply chain performance will be amplified in high-velocity markets. There is a basis for this argument within extant research that espouses the idea that knowledge dissemination can lead to greater variance in performance outcomes when faced with dynamic environments. Environmental dynamism puts great pressure on firms to leverage organizational knowledge to guide their courses of action [23, 42, 71]. Specifically, when facing a dynamic environment, organizational key decision makers are required to evaluate situations rapidly and execute effectively [7]. However, a high-velocity marketplace can place greater stress and cognitive demands on them, even impeding their abilities to make sense of situations and to implement imperative ideas [61]. A turbulent environment can thus diminish the confidence of key organizational decision makers when they are making strategic and operational decisions [9]. Hence the need for BDA becomes of critical importance to organizational decision makers when they face such an environment. Furthermore, in volatile markets, dynamic capabilities rely less on existing knowledge, and considerably more on rapidly creating new knowledge that is situation-specific [25]; this condition provides additional potential impact of BDA. The above discussion suggests that the value of BDA use is most salient in highly

dynamic environments because BDA is expected to increase an organization’s capacity to discover new knowledge and insight [13]. This is particularly relevant to SCM practices in which large amounts of data are frequently collected from multiple sources at multiple points within a firm (e.g., demand data, operating process data, inventory records, shipping data, supplier transactional data, etc.). In order to make effective decisions in a highly dynamic market, such data must be processed, integrated, ana- lyzed, and interpreted in holistic ways. The use of BDA facilitates the generation of

14 CHEN, PRESTON, AND SWINK

insights quickly and effectively across essential SCM activities (e.g., logistics improve- ment, forecasting and planning, product run optimization, etc.) for cost saving or growth opportunities. Prior literature also indicates that in a high-velocity market, new knowl- edge and insight provide a sense of confidence that helps managers to overcome emotional difficulties when coping with uncertainty and will therefore promote manage- rial confidence in such situations [24]. Furthermore, new situational knowledge can allow managers to develop intuition about the marketplace so that these managers can quickly understand the changing situation and appropriately adapt to it. In order to examine the theoretical relationship between BDA use and supply chain performance under different levels of environmental dynamism, we hypothesize the following:

Hypothesis 2a: Environment dynamism positively moderates the impact of big data analytics use on asset productivity.

Hypothesis 2b: Environment dynamism positively moderates the impact of big data analytics use on business growth.

TOE Factors as Antecedents to BDA Use

To develop a parsimonious research model, through an extensive search of the literature we identified four TOE factors that have been noted as determinants of organizational IT adoption and are highly relevant explanatory factors for BDA usage: expected benefits, technology compatibility, organizational readiness, and competitive pressure. Among these factors, expected benefits and technology com- patibility are categorized as technological factors because they represent the instru- mental benefits of using the technology and the congruence of technology use with existing practices of potential users [17, 78, 79, 81]. Organizational readiness is categorized as an organizational factor because it describes a firm’s resource avail- ability to pursue the technology [17, 89]. Competitive pressure is categorized as an environmental factor because it captures the current status of the technology use in the competitive environment [17, 78, 90]. Table 1 provides a summary of the definitions of the four factors that we identified from the literature as well as the possible paths through which these factors may influence organizational BDA use.

Technological Drivers of BDA Use

Expected benefits: Expected benefits describes the anticipated benefits, including both the operational and strategic advantage to the organization, gained from using a new technology [81]. Prior IT literature has suggested that these benefits are both direct and indirect in nature. Direct benefits include operational cost savings and other internal efficiencies arising from, for example, reduced paperwork, data reentry, and error rates. Likewise, indirect benefits are opportunities that emerge from the use of a technology, such as improved customer service and the potential for process reengineering [17, 78].

BIG DATA ANALYTICS IN SUPPLY CHAIN MANAGEMENT 15

T ab le

1 . D ef in it io n , S o u rc e,

an d P ro p o se d E ff ec ts o f T O E F ac to rs

C o n te x t

F ac to r

D ef in it io n

S o u rc e

Im p ac t p at h

T e ch

n o lo g ic a l

E xp

e ct e d B e n e fit s

T h e e xt e n t to

w h ic h B D A is

a n tic ip a te d to

b ri n g

o p e ra tio

n a l a n d st ra te g ic

a d va

n ta g e to

th e o rg a n iz a tio

n .

[1 7 , 3 8 , 7 8 , 7 9 , 8 1 ]

D ir e ct

T e ch

n o lo g y C o m p a tib

ili ty

T h e e xt e n t to

w h ic h B D A is

p e rc e iv e d a s b e in g

co n si st e n t w ith

e xi st in g

o rg a n iz a tio

n a l cu

ltu re

a n d p ra ct ic e s.

[1 8 , 7 8 , 7 9 , 8 1 ]

D ir e ct

O rg a n iz a tio

n a l

O rg a n iz a tio

n a l R e a d in e ss

T h e e xt e n t to

w h ic h o rg a n iz a tio

n a l re so

u rc e s a re

a va

ila b le

fo r u si n g B D A .

[1 7 , 3 8 , 7 8 , 8 9 , 9 1 ]

In d ir e ct

vi a T M T S u p p o rt

E n vi ro n m e n ta l

C o m p e tit iv e P re ss

u re

T h e in flu

e n ce

s fr o m

th e co

m p e tit iv e e n vi ro n m e n t

fo r th e o rg a n iz a tio

n to

u se

B D A to

m a in ta in

o r in cr e a se

co m p e tit iv e n e ss

.

[1 7 , 3 8 , 7 8 , 8 9 , 9 0 ]

In d ir e ct

vi a T M T S u p p o rt

16

In addition to these tangible benefits, organizational decision makers may also value intangible benefits such as improved knowledge sharing and coordination [81]. Even though the use of BDA is at an early stage, industry press has reported a wide

range of potential operational and strategic benefits of BDA usage that are expected by organizational leaders. These benefits often include cost reduction, work improvement, new product offering, new revenue stream development, tailored customer/patient care, and so on [8]. Organizational decision makers should assess the expected benefits (versus the costs) of employing BDA in supply chain practices. Activities included in SCM are highly data-intensive and involve complex decision processes. If managers of a focal firm expect that the use of BDA is highly beneficial, they should be more likely to move on and determine how their firm can leverage these benefits by extensively using BDA to serve business needs. Thus, we hypothesize:

Hypothesis 3: Expected benefits is positively associated with big data analytics use.

Technology compatibility: Compatibility has been the most frequently cited factor leading to innovation adoption [79]. Rogers defined compatibility as “the degree to which the innovation is perceived as consistent with the existing values, past experi- ences, and needs of the potential adopter” [63, p. 223]. Applying this definition to the organizational context, it suggests that compatibility of an innovation may refer to its congruence with (1) the value systems (e.g., culture), and (2) the business practices of an organization. Tornatsky and Klein [79] reported that both types of compatibility (which they described as cognitive and operational) are, theoretically, positively related to innovation adoption and implementation and sometimes difficult to differentiate. As such, in the current study, we assess the compatibility of BDA to an organiza-

tion’s existing culture and business practices (including the existing SCM practices). We contend that compatibility is a highly relevant attribute of BDA when examining the impacts of technological factors on BDA use, particularly when BDA use is conceptualized as a dynamic capability as we have proposed. As we described earlier, an important characteristic of organizational dynamic capability is its “path dependency” (i.e., organizational capability is shaped by its history and past experi- ence) [25, 75]. From this perspective, we suggest that if organizational decision makers view the adoption of BDA as compatible with existing organizational values (which also reflect history) and work practices, they will be more likely to use and apply it in various aspects of supply chain processes. Thus, we hypothesize:

Hypothesis 4: Technology compatibility is positively associated with big data analytics use.

Organizational and Environmental Drivers of BDA Use

Top management support: This defines the degree to which supply chain managers understand and appreciate the technological capabilities of the new IT systems (e.g.,

BIG DATA ANALYTICS IN SUPPLY CHAIN MANAGEMENT 17

BDA) and that the data they produce varies across organizations [64]. Prior literature has suggested that when top managers have positive beliefs about the potential usefulness of an IT system to the organization, they will most often take actions to support the employment of such a system [49]. Through this support (i.e., cham- pioning and promoting the IT implementation and usage within the organization), TMT acts as the agent facilitating the changes of organizational norms, values, and cultures, which enables other organizational members to use and adapt to the new technology [41, 43, 50]. We note the existence of a broad base of literature providing a theoretical foundation that top management support drives IT usage within orga- nizations [2, 11, 49]. As such, we posit that TMT support is a key driving force of organizational BDA use. Thus, we hypothesize:

Hypothesis 5: Top management support is positively associated with big data analytics use.

Organizational readiness: We suggest that the degree to which TMT provides support for organizational IT initiatives is determined by the availability of organiza- tional resources. In accordance with the extant literature [17, 38], we define orga- nizational readiness as the availability of the necessary organizational resources for using BDA. There has been some dispute within the extant literature regarding whether organizational readiness is indeed a direct driver of technology usage [38, 81]. As such, there is support in the literature that intermediary forces are necessary if organizational readiness is to impact BDA usage. Specifically, prior literature has suggested that these needed resources include financial capital and IT sophistication [47]. In particular, IT sophistication captures not only the technical components of IT infrastructure but also the IT human resources (i.e., technological knowledge and expertise) within the organization [17, 38, 81, 90]. In the context of BDA usage, we suggest that the availability of professionals with the skills or capability to perform business analytics is a critical indicator of organizational readiness. Prior IT manage- ment literature suggests that organizational conditions are key factors that influence TMT support [39]. Accordingly, we hypothesize that organizational readiness is an important condition for top managements to form their attitude toward wide orga- nizational use of BDA. Top management will be more supportive when they believe that the firm has sufficient resources and that capabilities are in place (or can be readily developed) to promote the diffusion of BDA. Thus, we hypothesize:

Hypothesis 6: Organizational readiness is positively associated with top man- agement support.

Competitive pressure: In the current study, competitive pressure refers to influ- ences from the external environment that prompt the organization to use BDA. We are particularly interested in assessing the pressure that the organization receives from its stakeholders (e.g., customers, suppliers, and competitors). Although the effects of competitive pressures on organizational IT adoption have been reported in extant studies [11, 76, 89, 90], with the exception of Liang et al. [49], the literature

18 CHEN, PRESTON, AND SWINK

has almost universally ignored the theoretical “black box” with regard to the mediating role of TMT support. We posit that human agencies (i.e., top manage- ment) play significant roles in translating external environment characteristics into organizational responses [32]. As observed by Liang et al. [49], environmental pressures can positively affect the degree of TMT support. For example, the adoption of BDA by competitors will likely stimulate imitative tendencies of the TMT as the successful behaviors of competitors help with reducing the uncertainty of using a new technology. Furthermore, the increasing use of BDA by competitors and trading partners in the supply chain will also likely apply some coercive pressure [76] to the top corporate officials to more effectively and efficiently capture market intelligence in order to maintain the firm’s competitive position. Therefore, we hypothesize:

Hypothesis 7: Competitive pressure is positively associated with top manage- ment support.

Research Methodology

To test the research hypotheses, we employed a field study approach to collect data from supply chain executives through a questionnaire. In the following subsections we discuss the instrument, including measures, used for this study and the data collection process.

Sample

We collected data using a cross-sectional survey developed as part of an ongoing collaborative effort between university professors and Computer Sciences Corporation (CSC, a $12+ billion business technology consulting firm), which has a large supply chain IT practice. For the purposes of this study, we partnered with CSC and Supply Chain Management Review (SCMR, a trade journal) to design a survey instrument. The data were collected in the tenth iteration of an annual global survey process, administered in summer 2013. The survey instrument was intended for a single respondent, who represented the

supply chain management organization as the unit of analysis. As such, the survey respondents were supply chain executives located in firms around the world. Survey invitations were e-mailed to the clients of CSC, members of the Council of Supply Chain Management Professionals (CSCMP), and a list of university contacts. Three survey invitations were sent over a six-week period. In order to enhance the quality of the key informants, we screened the responses and eliminated informants whose titles were not directly related to a supply chain function [83]. A total of 161 usable responses were collected. Data were obtained from respondents in a wide variety of firms representing multiple industries and geographic regions. Respondents held a range of managerial positions, including executive managers (C-level executive, president, senior vice president, and vice president), upper managers (senior director,

BIG DATA ANALYTICS IN SUPPLY CHAIN MANAGEMENT 19

director, head), and middle managers (senior manager, manager). This distribution of respondents suggests that they had relevant knowledge regarding the study ques- tions. We observe that the firms in our sample have annual revenues that range from less than $250 million to greater than $10 billion. Furthermore, we find that the organizations in this sample range from less than 250 employees to more than 30,000, and a total number of IT professionals in the organization ranging from 1 to 20,000. A limitation of the method used to invite respondents to participate in the survey is

that it did not allow calculation of a conventional response rate. Therefore, we assess nonresponse bias, in two different ways. First, we compared return on assets of the sampled firms with their respective industry median values using paired sample t-tests. Our results reveal that there were no statistically significant differences (p > .05) between the sample firms and industry median values, suggesting that a performance bias is not present in our sample. Second, there were no statistically significant differences in responses to any of the measurement items between early (first 25 percent) and late respondents (last 25 percent) [3]. Thus, we find no cause for concern with regard to nonresponse bias.

Measures and Questionnaire Administration

We designed survey items to ensure face validity, and to maximize relevance and readability for the respondents. The questionnaire contains a number of existing valid instruments that were adapted to the current study. In particular, we measure the following constructs in our research model: BDA usage, technology compat- ibility, organizational readiness, expected benefits, competitive pressure, TMT sup- port, environmental dynamism, asset productivity, and business growth. To account for the differences among organizations, we also include control variables (annual sales, number of employees, number of IT professionals in the organization, and industry) for BDA usage and both dimensions of performance outcomes in the research model. All survey items were suggested, reviewed, and validated by three academic researchers, two CSC partners, and five supply chain practitioners.5

As noted previously, BDA usage is defined in the current study as the extent to which organizations use BDA to process information generated across key supply chain processes. In order to develop valid measurement items for this focal construct of the study, we made an effort to seek both conceptual and practical support to identify supply chain processes to which BDA usage is mostly relevant. For example, we reviewed the literature to theoretically categorize the highly complex and information-intensive supply chain processes into three types as outlined by Teece et al. [75]: (a) coordination/integration (a static concept), (b) learning (a dynamic concept), and (c) reconfiguration (a transformational concept). We also approached two partners of CSC supply chain IT practice and two editors of SCMR to suggest and validate the items of the BDA use construct from the perspective of how BDA is being applied in SCM practices. These efforts ensured that we identify

20 CHEN, PRESTON, AND SWINK

and use only terms and applications of BDA with which our respondents (i.e., supply chain executives) are most likely familiar. Through multiple rounds of iterative discussions and refinements with our colla-

borators in industry, we developed 10 measurement items of BDA usage covering the three dimensions of supply chain process, including (1) coordination/integration processes (warehouse operations improvements, process/equipment monitoring, logistics improvements), (2) learning processes (sourcing analysis, purchasing spend analytics, CRM/customer/patient analysis, forecasting/demand management), and (3) reconfiguration processes (network design/optimization, production run optimization, inventory optimization). To help the respondents more effectively answer the BDA usage questions, at the

beginning of the survey we provide definitions and examples of big data and big data analytics to ensure that the respondents have a common understanding of the research. These definitions and examples are provided in Appendix A. The survey items and scales (including the literature sources) used to measure all other con- structs are also included in Appendix A.

Data Analysis and Results

To establish the nomological validity of the research model, we analyzed the survey data using partial least squares (PLS) with a two-step analytic approach. First, the measurement model was evaluated to assess the validity and reliability of the measures. Second, the structural model was evaluated to assess the strength of the hypothesized links among the variables. The psychometric properties of all scales were assessed within the context of the structural model through an assessment of discriminant validity and reliability.

Measurement Model

Summary statistics of the variables are presented in Table 2. The psychometric properties of the scales are assessed in terms of item loadings, internal consis- tency, and discriminant validity (Tables 3 and 4). Item loadings and internal consistencies greater than 0.70 are generally considered acceptable [26]. As shown by the factor analysis results (Table 3) and composite reliability scores (Table 4), the scales used in the study largely meet these guidelines. To assess the discriminant validity of the measures, we follow the guidelines suggested by the literature [15]: (1) indicators should load more strongly on their corresponding constructs than on other constructs in the model, and (2) the square root of the average variance extracted (AVE) should be larger than the interconstruct correla- tions. As shown by the factor analysis results and the comparison of the inter- construct correlations and AVE (shaded leading diagonal), the constructs meet these guidelines. Thus, the results support the discriminant validity of the con- structs in the model.

BIG DATA ANALYTICS IN SUPPLY CHAIN MANAGEMENT 21

In addition to establishing measurement reliability and validity, it is important to consider the potential effects of common method variance (CMV) on the measurement model. For several reasons, CMV is not likely to be a source of bias in this study. First, the potential for CMV is lessened somewhat by the fact that the supply chain executive was the appropriate respondent for the data, because he or she was the subject matter expert most knowledgeable about internal and external supply chain related matters [19]. Nevertheless, we assessed potential CMV using several observational and statistical techniques. In accor- dance with Craighead et al. [19], we first observe that the latent variable correla- tion coefficients between a subset of the subjectively measured constructs (i.e., not part of a causal relationship) in our model were of reasonable magnitude (e.g., Technology Compatibility with Organizational Readiness [0.289]; Competitive Pressure [0.403]; Asset Productivity [0.255]; and Business Growth [0.321]). These observed correlations provide evidence that common method variance is not prevalent [19, 72]. Furthermore, researchers have shown that CMV is usually not a concern in the assessment of moderation effects [66]. We observe that there are three theoretically justified, statistically significant interaction effects as noted in our research models and findings. The presence of significant interaction effects between these latent variables provides additional evidence that CMV is not a concern [19, 66]. We also conducted two statistical tests to assess the potential of CMV, including

Harmon’s single-factor test [33, 57, 58] and the correlational marker variable technique [51, 53]. In accordance with established guidelines pertaining to Harman’s single-factor test, CMV does not appear to be problematic because: (1) several factors were identified; (2) the first factor did not account for the majority of the variance; and (3) there is no general factor in the unrotated factor structure. While the single-factor test is not generally regarded as the most robust of tests for CMV, the use of a marker variable has been suggested as the most essential statistical assessment of CMV, specifically within the fields of

Table 2. Summary Statistics of Variables

Variable (# items) N Mean Std. dev. Min Max

Technology Compatibility (3) 161 3.16 0.82 1.00 5.00 Organizational Readiness (4) 161 2.85 0.88 1.00 5.00 Expected Benefits (6) 161 3.77 0.68 1.00 5.00 Competitive Pressure (3) 143 2.59 0.84 1.00 5.00 TMT Support (3) 161 2.99 1.05 1.00 5.00 Big Data Usage (10) 161 2.50 1.04 1.00 5.00 Environmental Dynamism (4) 161 2.95 0.82 1.00 4.50 Asset Productivity (4) 138 3.81 0.97 1.00 5.00 Business Growth (3) 141 3.77 1.08 1.00 5.00

Note: All measures represent five-point scales ranging from (1) to (5).

22 CHEN, PRESTON, AND SWINK

operations management, supply chain management, and information systems research [19]. Using this method, we compared the proposed research model

Table 3. Results of Factor Analysis

Indicators Tech Comp

Org Read

Exp Ben

Comp Press

TMT Supp

BD Usage

Env Dyn

Asset Prod

Bus Grow

TechComp1 0.852 0.147 0.243 0.322 0.461 0.493 0.102 0.165 0.249 TechComp2 0.876 0.343 0.114 0.382 0.532 0.422 0.181 0.289 0.301 TechComp3 0.717 0.238 0.117 0.284 0.418 0.368 –0.001 0.178 0.240 OrgRead1 0.103 0.618 –0.104 0.037 0.153 0.206 –0.097 0.263 0.182 OrgRead2 0.279 0.743 –0.170 0.066 0.204 0.234 –0.124 0.263 0.171 OrgRead3 0.268 0.863 –0.047 0.110 0.269 0.194 0.043 0.314 0.216 OrgRead4 0.198 0.813 –0.107 0.016 0.174 0.118 0.007 0.313 0.131 ExpBen1 0.226 –0.099 0.844 0.317 0.260 0.322 0.060 0.068 0.172 ExpBen2 0.129 –0.090 0.888 0.216 0.195 0.295 0.043 0.134 0.139 ExpBen3 0.152 –0.105 0.841 0.208 0.189 0.281 –0.025 0.129 0.098 ExpBen4 0.188 –0.113 0.809 0.207 0.243 0.260 0.061 0.133 0.104 ExpBen5 0.157 –0.103 0.698 0.105 0.174 0.133 0.085 0.053 0.182 ExpBen6 0.057 –0.194 0.692 0.145 0.084 0.107 0.040 –0.023 0.088 CompPress1 0.402 0.080 0.233 0.872 0.415 0.389 0.035 0.016 0.028 CompPress2 0.262 0.093 0.177 0.827 0.356 0.440 0.057 0.107 0.116 CompPress3 0.322 0.027 0.257 0.764 0.306 0.380 0.030 0.009 0.051 TMT-Supp1 0.554 0.181 0.269 0.395 0.929 0.517 0.014 0.236 0.222 TMT-Supp2 0.559 0.253 0.248 0.438 0.966 0.545 0.082 0.269 0.292 TMT-Supp3 0.522 0.330 0.203 0.420 0.949 0.512 0.080 0.311 0.234 BD-Usage1 0.506 0.226 0.294 0.444 0.474 0.841 –0.013 0.195 0.247 BD-Usage2 0.471 0.175 0.365 0.423 0.445 0.783 0.012 0.199 0.224 BD-Usage3 0.394 0.199 0.218 0.334 0.412 0.736 0.099 0.256 0.238 BD-Usage4 0.454 0.214 0.261 0.379 0.503 0.804 0.160 0.224 0.274 BD-Usage5 0.413 0.184 0.285 0.360 0.424 0.806 0.037 0.241 0.218 BD-Usage6 0.317 0.121 0.263 0.381 0.359 0.815 0.103 0.174 0.235 BD-Usage7 0.393 0.100 0.212 0.347 0.369 0.797 0.039 0.175 0.224 BD-Usage8 0.366 0.254 0.330 0.399 0.485 0.808 0.076 0.355 0.336 BD-Usage9 0.400 0.198 0.140 0.404 0.462 0.761 –0.003 0.238 0.163 BD-Usage10 0.468 0.244 0.130 0.401 0.432 0.801 0.029 0.229 0.208 EnvDyn1 0.049 –0.052 0.147 –0.013 –0.018 –0.040 0.700 –0.008 0.046 EnvDyn2 0.115 –0.019 –0.098 0.016 0.012 –0.002 0.631 0.060 0.039 EnvDyn3 0.161 –0.021 0.097 0.116 0.092 0.156 0.812 0.077 0.096 EnvDyn4 0.055 –0.049 0.032 0.007 0.050 0.025 0.887 0.094 0.152 AssetProd1 0.187 0.350 0.126 0.090 0.248 0.242 0.032 0.859 0.405 AssetProd2 0.207 0.310 0.042 0.048 0.240 0.272 0.070 0.885 0.436 AssetProd3 0.226 0.367 0.084 0.051 0.255 0.255 0.081 0.931 0.527 AssetProd4 0.285 0.321 0.173 0.011 0.284 0.270 0.139 0.900 0.518 BusGrow1 0.288 0.198 0.167 0.061 0.211 0.267 0.066 0.482 0.853 BusGrow2 0.288 0.222 0.108 0.008 0.189 0.241 0.186 0.491 0.933 BusGrow3 0.288 0.203 0.158 0.131 0.302 0.301 0.103 0.455 0.904

BIG DATA ANALYTICS IN SUPPLY CHAIN MANAGEMENT 23

T ab le

4 . In te rc o n st ru ct

C o rr el at io n s

R el ia b il it y a (#

it em

s) T ec h C o m p

O rg

R ea d

E x p B en

C o m p P re ss

T M T S u p p

B D

U sa g e

E n v D y n

A ss et

P ro d

B u s G ro w

T e ch

C o m p

0 .8 5 7 (3 )

0 .8 1 8

O rg R e a d

0 .8 4 7 (4 )

0 .2 8 9

0 .7 6 4

E xp

B e n

0 .9 1 3 (6 )

0 .2 0 0

– 0 .1 3 0

0 .8 0 0

C o m p P re ss

0 .8 6 2 (3 )

0 .4 0 3

0 .0 8 3

0 .2 8 3

0 .8 2 2

T M T -S u p p

0 .9 6 4 (3 )

0 .5 7 5

0 .2 7 1

0 .2 5 7

0 .4 4 1

0 .9 4 8

B D -U

sa g e

0 .9 4 5 (1 0 )

0 .5 2 9

0 .2 4 5

0 .3 2 5

0 .4 8 8

0 .5 5 3

0 .7 9 6

E n vD

yn 0 .8 4 6 (4 )

0 .1 2 0

– 0 .0 4 4

0 .0 6 3

0 .0 4 9

0 .0 6 3

0 .0 6 8

0 .7 6 4

A ss

e tP ro d

0 .9 4 1 (4 )

0 .2 5 5

0 .3 7 6

0 .1 1 2

0 .0 5 3

0 .2 8 8

0 .2 9 1

0 .0 9 3

0 .8 9 4

B u sG

ro w

0 .9 2 5 (3 )

0 .3 2 1

0 .2 3 2

0 .1 6 0

0 .0 7 7

0 .2 6 4

0 .3 0 1

0 .1 3 3

0 .5 3 0

0 .8 9 7

24

against a revised model that introduced a marker variable having no theoretical relationship to the other constructs in our research model. Our marker variable was a multi-item scale that assessed the general level of cloud-based computing used by the organization. The revised model included paths between this marker variable and each of the dependent variables in the research model. Correlations and path coefficients were not substantially different between the original and revised models, and the paths from the marker variable to each of the constructs in the model were nonsignificant. These results provide further evidence to alleviate potential concerns associated with CMV [58]. Based on the study design and these observational and statistical tests, we therefore conclude that the probability of CMV is minimal, and is unlikely to bias the findings of our study.

Structural Model

After examining the measurement validity, we employed PLS to test the structural model. The significance of the paths was determined using the T-statistic calculated with the bootstrapping technique. As discussed earlier, based on our review of prior literature, we include control variables (annual sales, number of employees, and industry) for BDA use and the organizational performance constructs. The results show that none of these control variables had a statistically significant effect on any of the dependent variables in our research model. The path coefficients for the structural model are shown in Figure 3. Table 5 presents a summary of the hypoth- esis results, which shows that all hypothesized causal paths in the research model were significant. From the results of the PLS structural model, we observe that BDA use directly

influences asset productivity and business growth, and explains approximately 8.5 percent and 9.2 percent of the variances in these organizational outcomes, respec- tively. In addition, we observe that the primary antecedent variables (technical compatibility, TMT support, expected benefits) in the nomological network collec- tively explain approximately 40.1 percent of the variance in BDA use. Furthermore, we observe that the competitive pressure and organizational readiness collectively explain approximately 25 percent of the variance in TMT support. In sum, the structural model results provide strong support for all the hypotheses in our main model (H1a, H1b, H3, H4, H5, H6, and H7). We conducted post hoc analyses to further assess the nomological network of the

research model. Specifically, we conducted mediation analyses to examine the media- tion effects of: (1) the primary antecedent on organizational performance via BDA use; and (2) competitive pressure and organizational readiness on BDA use via TMTsupport. To conduct this additional assessment, we followed the mediation analysis procedures suggested by prior literature [4, 69]. Our findings show that: (1) BDA use either fully mediates or is a required intervening variable between the primary antecedent variables and both components of organizational performance; and (2) TMT support partially mediates the influence of competitive pressure and organizational readiness on BDA

BIG DATA ANALYTICS IN SUPPLY CHAIN MANAGEMENT 25

use. Thus these findings support the structure of the posited nomological network. The next section discusses the results of tests for the proposed moderating effects.

Big Data Analytics Use

40.1%

Asset Productivity 14.9% (8.5%)

Competitive Pressure

Environmental Dynamism

Business Growth

16.8% (9.2%)

-0.248 (t=3.774)

0.303 (t=4.597)

0.300 (t=4.519)

0.337 (t=4.456)

0.291 (t=4.330)

0.421 (t=6.545)

0.286 (t=3.574)

0.174 (t=2.951)

0.257 (t=3.881)

TMT Support

25.0%

Organizational Readiness

Expected Benefits

Technology Compatibility

Technological

Organizational

Environmental

Figure 3. Research Results

Table 5. Summary of Hypothesis Tests

Hypotheses Support for hypotheses

H1a: Big Data Analytics Usage → Asset Productivity Supported** H1b: Big Data Analytics Usage → Business Growth Supported** H2a: Environmental Dynamism Positively Moderates:

Big Data Analytics Usage → Asset Productivity Significant**. But a negative

moderation effect is observed H2b: Environmental Dynamism Positively Moderates:

Big Data Analytics Usage → Asset Productivity Supported**

H3: Expected Benefits → Big Data Analytics Usage Supported** H4: Technology Compatibility → Big Data Analytics

Usage Supported**

H5: Top Management Support → Big Data Analytics Usage

Supported**

H6: Organizational Readiness → Top Management Support

Supported**

H7: Competitive Pressure → Top Management Support

Supported**

** Path coefficient is significant at 0.01.

26 CHEN, PRESTON, AND SWINK

Test of Moderation Effects

In addition to the linear aspects of our model, we examined two moderating effects: H2a and H2b, respectively, posit that environmental dynamism positively moderates the degree to which BDA use influences asset productivity and business growth. Testing moderating effects involves a comparison of a main effect model with a moderating effect model. We conducted our analyses by creating interaction variables directly within Smart PLS. Interaction terms were computed using the standardized scores, thus limiting potential issues with regard to multicollinearity between the main and interaction variables. In each of the interaction models, the interaction terms are significant with the addition of each of the interaction variables. We observe that, as hypothesized, the path coefficient for H2b is positive (0.257, significant at 0.01). However, the moderation effect of H2a is negative (rather than positive as hypothe- sized) with a path coefficient of –0.248 (significant at 0.01). To further examine each of the interaction effects, we tested whether the variance explained due to the moderated effect is significant beyond the main effects using the following F-statistic [10]:

F½dfinteraction � dfmain; N � dfinteraction � 1� ¼ ½ΔR2= dfinteraction � dfmainð Þ�= F½ð1 � R2interactionÞ= N � dfinteraction � 1ð Þ�:

The F-statistic was found to be significant for both outlined tests. Specifically, the F-statistics for H2a and H2b are 10.73 and 11.90, respectively, with both values significant at a level of 0.01. Thus, these results provide further statistical support that BDA use might have a diminished influence on asset productivity when the organization is faced with a dynamic environment; however, that BDA use will have a stronger influence on business growth when in a dynamic environment. See Table 6 for a summary of the moderation test results.

Discussion and Implications

Discussion

Unlike extant literature that focuses on predicting organizational IT adoption intent at an individual level, the current study investigates the value creation process of the actual usage of a new set of enterprise-wide technologies, big data analytics, at an organizational level leading to key organizational benefits. Our examination of both the antecedents and consequences of BDA use at the organizational level provides a more thorough assessment of how BDA may influence organizational performance. In addition, our study adds to the theoretical understanding of the mechanisms that support this phenomenon. In the remainder of this section, we discuss the implica- tions of the findings pertaining to our primary research questions. Our findings support the proposition that organizational BDA use generates

organizational value creation in the supply chain management domain. Specifically, we find that BDA use, an important organizational information

BIG DATA ANALYTICS IN SUPPLY CHAIN MANAGEMENT 27

processing capability, has a positive influence on both asset productivity and busi- ness growth, thus supporting our hypotheses and confirming the potential of BDA for contemporary businesses. These findings are consistent with the established literature that information processing capability is an important part of organizational management, particularly within the domain of supply chain management [27, 65]. Furthermore, we view information processing as a dynamic capability that brings temporary competitive advantage to firms [25]. Consistent with this view of dynamic capabilities, we observe that environmental dynamism moderates the way in which BDA use influences both asset productivity and business growth. Specifically, we observe that BDA use influences business growth to an even greater degree in dynamic environments, in accordance with our hypothesis. On the other hand, we observe the influence of BDA on asset productivity—

although it is a positive association, it is less pronounced in more dynamic environ- ments. This finding is in contrast to our proposed hypothesis; however, as we assess this result further, we see a potential basis for this observation. There are several reasons why BDA may not be as beneficial to asset productivity, as compared to business growth, in highly dynamic markets. First, it is likely that substantial structural changes must be made to physical assets before asset productivity improvements can be realized. Managers may be either reluctant to make such changes or they may feel that they are not empowered to do so. For example, changes to plant, property, and equipment are not made easily or quickly. Such changes typically occur only when large organizational maneuvers are made, rather than occurring incrementally, and as such, changes may take time to ramp up. Second, BDA may be less influential on asset productivity improvements in dynamic markets, as opposed to a more stable environment, because volatility creates uncertainty. Such uncertainty in turn makes appropriate asset adjustments difficult to discern. The findings of the current study enhance the explanatory power of the TOE

framework, also enriching it within the BDA and SCM context. The data analysis results show that both technological components in our model (technology

Table 6. Summary of Moderation Test Results

Hypotheses Support for hypotheses

Path Coeff (β)

F- statistic

H2a: Environmental Dynamism Positively Moderates: Big Data Analytics Usage → Asset Productivity

Contrary hypothesis supported

–0.248** 10.73**

H2b: Environmental Dynamism Positively Moderates: Big Data Analytics Usage → Business Growth

Supported 0.257** 11.90**

** Path coefficient is significant at 0.01.

28 CHEN, PRESTON, AND SWINK

compatibility and expected benefits) directly influence BDA use. These findings support our hypotheses and are also generally consistent with the TOE framework. Furthermore, these findings provide additional insight into the value creation process of the BDA usage. Specifically, we observe that it is the actual usage of the technologies (BDA), rather than the technological factors per se, that directly impacts organizational performance. The ultimate influence of big data on organiza- tional outcomes is mediated by BDA usage. As such, the TOE factors are needed antecedents to facilitate BDA usage, which will ultimately be the critical component leading to greater firm performance. In addition, our findings show that the organizational and environmental factors

indirectly influence BDA use through TMT support. We note that TMT support has previously been employed by other authors to understand factors driving organiza- tional IT practices. However, with the exception of Liang et al. [49], the majority of the existing studies treat TMT behavior as an exogenous variable. Our study includes TMT in the research model as a mediator of the influence of organizational and environmental factors on organizational BDA use. As such, we extend the TOE framework by theorizing that organizational levels of IT use and the resulting value creation will vary even for organizations facing the same TOE factors. We find support that TMT plays a critical role when it comes to BDA. The TMT is

especially important for screening and interpreting the implications of external stimuli to the deployment of strategic IS-related issues such as BDA diffusion. Building on the extant IS literature, which has found that TMT plays a critical role in strategic IS concerns such as IS strategy and alignment [12, 59], our study offers further evidence of the role of human agents in translating environmental signals into concrete organizational actions. We posit that TMT support is particu- larly important in the context of the SCM environment because BDA often requires the collection and integration of data from dispersed functional areas and also the dissemination of information across the organization. TMT support is particularly important with regard to BDA because the various functional groups within the organization may be less likely to have established protocols on whether organiza- tional members should attempt to derive insight from vast amounts of potentially ambiguous data and how they should do so.

Limitations, Implications, and Future Research

We note several potential limitations of the study. First, the theories we employed use causal terms to describe the relationships included in the research model; however, the cross-sectional research design employed does not fully allow firm conclusions of causality. Longitudinal research should be employed to provide additional support for causal relationships. For future research, a longitudinal study might provide additional insight into the temporal nature of the way BDA use influences aspects of organizational performance specifically with the interplay of the dynamic environment. A longitudinal study would also provide a more

BIG DATA ANALYTICS IN SUPPLY CHAIN MANAGEMENT 29

granular understanding of how the TOE factors influence BDA use and also how these factors possibly interact with each other sequentially. As such, a longitudinal study will work well to potentially complement the observations found in the current study. Second, our study focuses on BDA usage in a supply chain management context. BDA can also likely play a key role in other functional areas such as accounting and most notably in the marketing domain. Also, the current study examined BDA use only in the context of physical product-centric organizations. Future studies could examine BDA in other organizational/industry contexts (i.e., banking, health care, etc.). In these ways, future studies should examine the general- izability of our findings. Although the current sample is not expected to be idiosyn- cratic and should be representative of organizations in general, future studies should be conducted to examine this phenomenon more comprehensively and within other functional and industry contexts. The organizations in the current study showed an appropriately substantial level and variance of BDA (mean = 2.5/5.0 and std. dev. = 1.04), thus providing evidence that DBA use is relevant in our sample of organiza- tions. However, it would be interesting to examine BDA when this phenomenon becomes more mature, as it is currently in its early stages of development as an organizational asset. Despite these limitations, this study is the first to use a large-scale field survey

approach to test an integrated theoretical model that examines both the antecedents and consequences of BDA use. The study extends the existing body of knowledge on organizational-level IT adoption. Specifically, the research model, which is cohesively grounded in theoretical lenses that include dynamic capabilities theory in conjunction with the TOE framework, is readily applicable and generalizable to future studies on the value creation process of enterprise-level usage of knowledge creation systems. We focus on BDA use as a dynamic organizational information processing capability that can influence key aspects of organizational performance. Future research should examine the influence of the firm-level employment of BDA (or other knowledge systems) on other aspects of organizational performance, beyond the finite components examined for the present study. In addition, we provide more insight into the nature of the relationship between BDA use and organizations performance levels as dictated by the volatility of the external envir- onment. Although our findings on the differential effects of environmental volatility on internal (i.e., productivity) and external (i.e., growth) performance are very interesting, future research should replicate our study to determine whether such differential effects of environment dynamism would hold using different perfor- mance metrics. Future studies could also assess other contextual factors that could influence the degree to which the use of BDA (or other knowledge creation systems) influences organizational outcomes (e.g., technology maturity, BDA adoption phase, etc.). In addition, the current study has helped to develop a bridge between TOE factors

and organizational outcomes via BDA use; however, additional development of the nomological network is warranted in future research to provide an even richer understanding of this phenomenon. Specifically, future research should examine

30 CHEN, PRESTON, AND SWINK

potential intervening variables that may exist between organizational IT practices and performance outcomes. For instance, the alignment of BDA usage with business unit strategies is a potentially important phenomenon that could be a key variable linking BDA use to organizational performance. The study also extends the extant body of knowledge by examining how the TOE

factors influence organizational IT use. Our examination is intensive, in that it examined the influence of these factors at a more granular level through the integration of dynamic capabilities theory in conjunction with the TOE framework. Consistent with the TOE framework, we find that the technological factors have a direct effect on BDA usage. We also find as an extension of this framework that organizational and environmental factors have an indirect influence on BDA usage via TMT support. Our findings thus provide richer insight into the prior TOE literature that generally posited similar influence of the TOE factors without con- sidering differential paths that may exist due to the nature of the phenomenon of interest (i.e., BDA in the case of the current study). The integration of dynamic capabilities theory with the TOE framework provides theoretical positioning linking the primary antecedents to organizational outcomes via BDA usage as the driving force. Future research should integrate additional complementary theoretical per- spectives, where applicable, into the framework that has been developed for this study. The results of this study also have several important implications for managerial

practice. First, we provide support that BDA use can directly influence organiza- tional performance. As we mentioned, there has been a massive amount of discus- sion within industry regarding the potential promise of big data analytics. However, many organizations are still reluctant to make such commitments, based on lack of knowledge about how to effectively proceed and thus uncertainty of the payoff. Our findings should provide some “proof of concept” that BDA use can indeed pay off for the organization. Furthermore, our findings provide insights for managers on how to assess the potential impact of environmental factors such as volatility on particular performance outcomes. We find that BDA use positively influences business growth, and that this influence is amplified in a dynamic environment. Although we find that BDA also influences asset productivity, this effect is dam- pened in a highly dynamic environment. As such, organizational decision makers should be aware that environment context may play a key role in the degree to which BDA may influence specific organizational outcomes. The current study also iden- tifies key levers through which management can influence BDA use. As we have noted, many organizations could be considered “big data laggards” despite their wishes to engage in BDA initiatives, because they do not comprehend the appro- priate route for potential success. Organizational managers should be aware that technological factors are important and are more straightforward with respect to the link to BDA use. Our study also contends that TMT support is a key part of the equation. The concept that TMT support is fundamental with regard to strategic decisions, such as BDA use, is sometimes overlooked. In particular, organizations must understand that TMT is especially important to BDA use with regard to

BIG DATA ANALYTICS IN SUPPLY CHAIN MANAGEMENT 31

organizational factors and at the interface of the environment with the organization. Therefore, having a technical requirement in place for BDA is only part of the picture. BDA championship by top management is essential to bridging the organi- zational and environmental factors into actionable usage of BDA.

Conclusion

Our study provides a nomological network that examines the influence of BDA on organizational outcomes, as well as the key factors that facilitate BDA use. Collectively, the findings provide a theory-based understanding of BDA usage, while also providing guidance regarding what managers should expect from using this rapidly emerging technology. In summary, our findings show that: (1) organizational- level BDA use has significant impacts on two types of supply chain value creation: asset productivity and business growth; (2) the impacts of BDA use to supply chain value creation are moderated by environmental dynamism; (3) technological factors (expected benefits and technological compatibility) have a direct influence on organi- zational BDA use; and (4) organizational (i.e., organizational readiness) and environ- mental (i.e., competitive pressure) factors have an indirect influence on organizational BDA use through top management support.

NOTES

1. For more details, see www.pwc.com/us/bigdata. 2. One exabyte is 1 billion gigabytes. 3. For example, whereas traditional business intelligence (BI) tools are database manage-

ment systems (or DBMS)-based (i.e., to process structured data collected through legacy systems); BDA tools are web and mobile based (i.e., to process unstructured data collected from various sources). Furthermore, using traditional BI tools, data are moved into the software, whereas using BDA tools, data software is being moved to the data. As such, BDA creates new demand for workers, and education for new skills.

4. We have observed that, although a wide range of business analytics tools exist in the marketplace, the features of these BDA technologies can be categorized as data management (e.g., massively parallel-processing databases), open-source programming (e.g., Hadoop, MapReduce), statistical analysis (e.g., sentiment analysis, time-series analysis), visualization tools that help structure and connect data to uncover hidden patterns, anomalies, unknown correlations, and other actionable insights, and in-memory computing (IMC) (e.g., SAP’s HANA).

5. These practitioners include executives and editors at CSCMP, SCMR, a couple of other supply chain practice journals, and CSC partners in global regions.

REFERENCES 1. Afuah, A. Dynamic boundaries of the firm: Are firms better off being vertically

integrated in the face of a technological change? Academy of Management Journal, 44, 6 (2001), 1211–1228.

2. Armstrong, C.P., and Sambamurthy, V. Information technology assimilation in firms: The influence of senior leadership and IT infrastructures. Information Systems Research, 10, 4 (1999), 304–328.

32 CHEN, PRESTON, AND SWINK

3. Armstrong, J.S., and Overton, T.S. Estimating non-response bias in mail surveys. Journal of Marketing Research, 14 (August 1977), 396–402.

4. Baron, R.M., and Kenny, D.A. The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical concerns. Journal of Personality and Social Psychology, 51 (1986), 1173–1182.

5. Barton, D., and Court, D. Making advanced analytics work for you. Harvard Business Review, 90 (2012), 79–83.

6. Brynjolfsson, E., and McAfee, A. Race against the Machine: How the Digital Revolution is Accelerating Innovation, Driving Productivity, and Irreversibly Transforming Employment and the Economy. Lexington, MA: Digital Frontier Press, 2011.

7. Cannella, A.A.; Park, J.-H.; and Lee, H.-U. Top management team functional back- ground diversity and firm performance: Examining the roles of team member colocation and environmental uncertainty. Academy of Management Journal, 51, 4 (2008), 768–784.

8. Cao, Q.; Jones, D.R.; and Sheng, H. Contained nomadic information environments: Technology,vorganization, and environment influences on adoption of hospital RFID patient tracking. Information and Management, 51, 2 (2014), 225–239.

9. Carmeli, A.; Schaubroeck, J.; and Tishler, A. How CEO empowering leadership shapes top management team processes: Implications for firm performance. Leadership Quarterly, 22, 2 (2011), 399–411. 10. Carte, T.A., and Russell, C.J. In pursuit of moderation: Nine common errors and their

solutions. MIS Quarterly, 27, 3 (2003), 479–501. 11. Chatterjee, D.; Grewal, R.; and Sambamurthy, V. Shaping up for e-commerce:

Institutional enablers of the organizational assimilation of web technologies. MIS Quarterly, 26, 2 (2002), 65–89. 12. Chen, D.Q.; Mocker, M.; Preston, D.S.; and Teubner, A. Information systems

strategy: Reconceptualization, measurement, and implications. MIS Quarterly, 34, 2 (2010), 233–259. 13. Chen, H.; Chiang, R.H.L.; and Storey, V.C. Business intelligence and analytics: From

big data to big impact. MIS Quarterly, 36, 4 (2012), 1165–1188. 14. Chen, I.J., and Paulraj, A. Towards a theory of supply chain management: The con-

structs and measurements. Journal of Operations Management, 22 (2004), 119–150. 15. Chin, W. The partial least squares approach to structural equation modeling. In G.A.

Marcoulides (ed.), Modern Methods for Business Research (pp. 295–336). Mahwah, NJ: Lawrence Erlbaum, 1998. 16. Christopher, M.; Lowson, R.; and Peck, H. Creating agile supply chains in the fashion

industry. International Journal of Retail and Distribution Management, 32, 8 (2004), 367– 376. 17. Chwelos, P.; Benbasat, I.; and Dexter, A.S. Research report: Empirical test of an EDI

adoption model. Information Systems Research, 12, 3 (2001), 304–321. 18. Cooper, R.B., and Zmud, R.W. Information technology implementation research: A

technological diffusion approach. Management Science, 36, 2 (1990), 123–139. 19. Craighead, C.W.; Ketchen, D.J.; Dunn, K.S.; and Hult, G.T.M. Addressing common

method variance: Guidelines for survey research on information technology, operations, and supply chain management. IEEE Transactions on Engineering Management, 58, 3 (2011), 578–588. 20. Davis, F.D. Perceived usefulness, perceived ease of use, and user acceptance of infor-

mation technology. MIS Quarterly, 13, 3 (1989), 319–340. 21. Dehning, B.; Richardson, V.J.; and Zmud, R.W. The financial performance effects of IT-

based supply chain management systems in manufacturing firms. Journal of Operations Management, 25 (2007), 806–824. 22. Dove, R. Agile enterprise cornerstones: knowledge, values, and response ability. In R.

Baskerville (ed.), Business Agility and Information Technology Diffusion (pp. 313–330). Berlin: Springer, 2005. 23. Droge, C.; Jayaram, J.; and Vickery, S.K. The effects of internal versus external

integration practices on time-based performance and overall firm performance. Journal of Operations Management, 22 (2004), 557–573.

BIG DATA ANALYTICS IN SUPPLY CHAIN MANAGEMENT 33

24. Eisenhardt, K.M. Making fast strategic decisions in high-velocity environments. Academy of Management Journal, 32, 3 (1989), 543–576. 25. Eisenhardt, K.M., and Martin, J.A. Dynamic capabilities: What are they? Strategic

Management Journal, 21 (2000), 1105–1121. 26. Fornell, C., and Larcker, D.F. Evaluating structural equations models with unobservable

variables and measurement error. Journal of Marketing Research, 18, 1 (1981), 39–50. 27. Galbraith, J.R. Designing Complex Organizations. Reading, MA: Addison-Wesley,

1973. 28. Galbraith, J.R. Organization design: An information processing view. Interfaces, 4, 3

(1974), 28–36. 29. Gillon, K.; Aral, S.; Ching-Yung, L.; Mithas, S.; and Zozulia, M. Business analytics:

Radical shift or incremental change? Communications of the Association for Information Systems, 34 (2014), 287–296. 30. Goodhue, D.L., and Thompson, R.L. Task-technology fit and individual performance.

MIS Quarterly, 19, 2 (1995), 213–236. 31. Grant, R.M. Prospering in dynamically competitive environments: Organization cap-

ability as knowledge integration. Organization Science, 7, 4 (1996), 375–387. 32. Hambrick, D., and Mason, P. Upper echelons: The organization as a reflection of its top

managers. Academy of Management Review, 9, 1 (1984), 193–206. 33. Harman, H.H. Modern Factor Analysis. Chicago: University of Chicago Press, 1967. 34. Helfat, C.; Finkelstein, S.; Mitchell, W.; Peteraf, M.A.; Singh, H.; Teece, D.J.; and

Winter, S.G. Dynamic Capabilities: Understanding Strategic Change in Organizations. Oxford: Blackwell, 2007. 35. Holsapple, C., and Jones, K. Exploring secondary activities of the knowledge chain.

Knowledge and Process Management, 12, 1 (2005), 3–31. 36. Hult, G.T.M.; Ketchen, D.J.; and Nichols, E.L. Organizational learning as a strate-

gic resource in supply management. Journal of Operations Management, 21 (2003), 541– 556. 37. Hult, G.T.M.; Ketchen, D.J.; Cavusgil, T.; and Calantone, R.J. Knowledge as a strategic

resource in supply chains. Journal of Operations Management, 24 (2006), 458–475. 38. Iacovou, C.L.; Benbasat, I.; and Dexter, A.S. Electronic data interchange and small

organizations: Adoption and impact of technology. MIS Quarterly, 19, 4 (1995), 465–485. 39. Jarvenpaa, S.L., and Ives, B. Executive involvement and participation in the manage-

ment of information technology. MIS Quarterly, 15, 2 (1991), 204–227. 40. Johnson, P.F.; Klassen, R.D.; Leenders, M.R.; and Awaysheh, A. Utilizing e-business

technologies in supply chains:The impact of firm characteristics and teams. Journal of Operations Management, 25 (2007), 1255–1274. 41. Karahanna, E., and Preston, D.S. The effect of social capital of the relationship between

the CIO and top management team on firm performance. Journal of Management Information Systems, 30, 1 (2013), 15–56. 42. Karimi, J., and Walter, Z. The role of dynamic capabilities in responding to digital

disruption: A factor-based study of the newspaper industry. Journal of Management Information Systems, 32, 1 (2015), 39–81. 43. Kearns, G.S., and Sabherwal, R. Strategic alignment between business and information

technology: A knowledge-based view of behaviors, outcome, and consequences. Journal of Management Information Systems, 23, 3 (2007), 129–162. 44. Klein, R., and Rai, A. Inter-firm strategic information flows in supply chain logistics

relationships. MIS Quarterly, 33, 4 (2009), 735–762. 45. Kohli, A.K., and Jaworski, B.J. Market orientation: The construct, research proposi-

tions, and managerial implications. Journal of Marketing, 54 (1990), 1–18. 46. Lengnick-Hall, C.A., and Wolff, J.A. Similarities and contradictions in the core logic

of three strategy research streams. Strategic Management Journal, 20, 12 (1999), 1109– 1132. 47. Li, X. Managerial entrenchment with strategic information technology: A dynamic

perspective. Journal of Management Information Systems, 25, 4 (2009), 183–204.

34 CHEN, PRESTON, AND SWINK

48. Li, X.; Chung, C.; Goldsby, T.J.; and Holsapple, C.W. A unified model of supply chain agility: The work-design perspective. International Journal of Logistics Management, 19, 3 (2008), 408–435. 49. Liang, H.; Saraf, N.; Hu, Q.; and Xue, Y. Assimilation of enterprise systems: The effect

of institutional pressures and the mediating role of top management. MIS Quarterly, 31, 1 (2007), 59–87. 50. Lim, J.-H.; Stratopoulos, T.C.; and Wirjanto, T.S. Sustainability of a firm’s reputation

for information technology capability: The role of senior IT executives. Journal of Management Information Systems, 30, 1 (2013), 57–96. 51. Lindell, M., and Whitney, D. Accounting for common method variance in cross-

sectional research designs. Journal of Applied Psychology, 86 (2001), 114–121. 52. Lummus, R.R.; Vokurka, R.J.; and Duclos, L.K. Delphi study on supply chain flex-

ibility. International Journal of Production Research, 43, 13 (2005), 2687–2708. 53. Malhotra, N.K.; Kim, S.S.; and Patil, A. Common method variance in IS research: A

comparison of alternative approaches and a reanalysis of past research. Management Science, 52 (2006), 1865–1883. 54. March, J.G., and Simon, H.A. Organizations. New York: Wiley, 1958. 55. McCarthy, I.P.; Lawrence, T.B.; Wixted, B.; and Gordon, B.R. A multidimensional

conceptualization of environmental velocity. Academy of Management Review, 35, 4 (2010), 604–626. 56. Nair, R., and Narayana, A. Getting Results from Big Data: A Capabilities-Driven Approach

to the Strategic Use of Unstructured Information. Florham Park, NJ: Booz & Co., 2012. 57. Podsakoff, P.M., and Organ, D.W. Self-reports in organizational research: Problems and

prospects. Journal of Management, 12 (1986), 531–544. 58. Podsakoff, P.M.; Scott, B.M.; Lee, J.-Y.; and Podsakoff, N.P. Common method biases in

behavioral research: a critical review of the literature and recommended remedies. Journal of Applied Psychology, 88 (2003), 879–903. 59. Preston, D., and Karahanna, E. The antecedents of IS strategic alignment: A nomolo-

gical network. Information Systems Research, 20, 2 (2009), 159–179. 60. PwC. Capitalizing on the promise of big data. 2013. 61. Qian, C.; Cao, Q.; and Takeuchi, R. Top management team functional diversity and

organizational innovation in China: The moderating effects of environment. Strategic Management Journal, 34, 1 (2013), 110–120. 62. Rai, A.; Patnayakuni, R.; and Patnayakuni, N. Firm performance impacts of digitally

enabled supply chain integration capabilities. MIS Quarterly, 29, 4 (2006), 625–652. 63. Rogers, E.M. Diffusion of Innovations New York: Free Press 1983. 64. Sanders, N.R. Pattern of information technology use: The impact on buyer–suppler

coordination and performance Journal of Operations Management, 26, 3 (2008), 349–367 65. Schoenherr, T., and Swink, M. Revisiting the arcs of integration: Cross-validations and

extensions. Journal of Operations Management, 30, 1–2 (2012), 99–115. 66. Siemsen, E.; Roth, A.; and Oliveira, P. Common method bias in regression models with

linear, quadratic, and interaction effects. Organizational Research Methods, 13, 3 (2010), 456–476. 67. Simon, H.A. Models of Man: Social and Rational-Mathematical Essays on Rational

Human Behavior in a Social Setting. New York: Wiley, 1957. 68. Snijders, C.; Matzat, U.; and Reips, U.-D. “Big Data”: Big gaps of knowledge in the

field of Internet. International Journal of Internet Science, 7 (2012), 1–5. 69. Sobel, M.E. Asymptotic confidence intervals for indirect effects in structural equation

models. In S. Leinhardt (ed.), Social Methodology. San Francisco: Jossey-Bass, 1982, pp. 290–312. 70. Son, J.-Y., and Benbasat, I. Organizational buyers’ adoption and use of B2B electronic

marketplaces: Efficiency- and legitimacy-oriented perspectives. Journal of Management Information Systems, 24, 1 (2007), 55–99. 71. Song, M.; Droge, C.; Hanvanich, S.; and Calantone, R. Marketing and technology

resource complementarity: An analysis of their interaction effect in two environmental con- texts. Strategic Management Journal, 26 (2005), 259–276.

BIG DATA ANALYTICS IN SUPPLY CHAIN MANAGEMENT 35

72. Spector, P.E., and Brannick, M.T. Common method issues: An introduction to the feature topic in organizational research methods. Organizational Research Methods, 13, 3 (2010), 403–406. 73. Stewart, G. Supply‐chain operations reference model (SCOR): The first cross‐industry

framework for integrated supply‐chain management. Logistics Information Management, 10, 2 (1997), 62–67. 74. Tallon, P.P. A process-oriented perspective on the alignment of information technology

and business strategy. Journal of Management Information Systems, 24, 3 (2007), 227–268. 75. Teece, D.J.; Pisano, G.; and Shuen, A. Dynamic capabilities and strategic management.

Strategic Management Journal, 18, 7 (1997), 509–533. 76. Teo, H.H.; Wei, K.K.; and Benbasat, I. Predicting intention to adopt interorganizational

linkages: An institutional perspective. MIS Quarterly, 27, 1 (2003), 19–49. 77. Thompson, J.D. Organizations in Action. New York: McGraw-Hill, 1967. 78. Thong, J.Y.L. An integrated model of information systems adoption in small businesses.

Journal of Management Information Systems, 15, 4 (1999), 187–214. 79. Tornatsky, L.G., and Klein, K.J. Innovation characteristics and innovation adoption-

implementation: A metaanalysis of findings. IEEE Transactions on Engineering Management, 29, 1 (1982), 28–45. 80. Tornatzky, L.G., and Fleischer, M. The Processes of Technological Innovation.

Lexington, MA: Lexington Books, 1990. 81. Venkatesh, V., and Bala, H. Adoption and impacts of interorganizational business

process standards: Role of partnering synergy. Information Systems Research, 23, 4 (2012), 1131–1157. 82. Venkatesh, V.; Morris, M.G.; Davis, G.B.; and Davis, F.D. User acceptance of informa-

tion technology: Toward a unified view. MIS Quarterly, 27, 3 (2003), 425–478. 83. Wall, T.B.; Jonathan, M.; Patterson, M.; Wood, S.J.; Sheehan, M.; Clegg, C.W.; and

West, M. On the validity of subjective measures of company performance. Personnel Psychology, 57, 1 (2004), 95–118. 84. Waller, M.A., and Fawcett, S.E. Data science, predictive analytics, and big data: A

revolution that will transform supply chain design and management. Journal of Business Logistics, 34, 2 (2013), 77–84. 85. Wang, E.T.G., and Wei, H.-L. Interorganizational governance value creation: coordinat-

ing for information visibility and flexibility in supply chains. Decision Sciences, 38, 4 (2007), 647–674. 86. Watson, H.J. Business analytics insight: Hype or here to stay? Business Intelligence

Journal, 16, 1 (2011), 4–8. 87. Williams, B.; Roh, J.; Tokar, T.; and Swink, M. Leveraging supply chain visibility for

operational responsiveness: the moderating role of internal integration. Journal of Operations Management, 31, 7/8 (2013), 543–554. 88. Yu, C.S., and Tao, Y.H. Understanding business-level innovation technology adoption.

Technovation, 29, 2 (2009), 92–109. 89. Zhu, K., and Kraemer, K.L. Post-adoption variations in usage and value of e-business

by organizations: Cross-country evidence from the retail industry. Information Systems Research, 16, 1 (2005), 61–84. 90. Zhu, K.; Kraemer, K.L.; and Xu, S. The process of innovation assimilation by firms in

different countries: A technology diffusion perspective on e-business. Management Science, 52, 10 (2006), 1557–1576. 91. Zhu, K.; Kraemer, K.L.; Xu, S.; and Dedrick, J. Information technology payoff in

e-business environments: An international perspective onvalue creation of e-business in the financial services industry. Journal of Management Information Systems, 21, 1 (2004), 17–54.

36 CHEN, PRESTON, AND SWINK

Appendix A: Construct Scales and Items

Definitions provided at the beginning of the survey

“Big data” refers to large structured and unstructured data sets that require new forms of processing capability to enable better decision making. Examples include, sales data, process operating data and other information captured by sen- sors, web server logs, Internet clickstream data, social media activity reports, mobile-phone call records, etc. “Big data analytics” is the process of examining big data using advanced

technologies. These include data management (e.g., massively parallel-processing databases), open-source programming (e.g., Hadoop, MapReduce), statistical analy- sis (e.g., sentiment analysis, time-series analysis), visualization tools that help structure and connect data to uncover hidden patterns, anomalies, unknown correla- tions, and other actionable insights, and in-memory computing (IMC) (e.g., SAP’s HANA)

Big Data Usage (sources: [14, 75], Expert Interview) To what extent has your organization implemented Big Data Analytics in each area? (1 =

little or no usage . . . 3 = moderate usage . . . 5 = heavy usage) 1. Sourcing analysis 2. Purchasing spend analytics 3. CRM/customer/patient analysis 4. Network design/optimization 5. Warehouse operations improvements 6. Process/equipment monitoring 7. Production run optimization 8. Logistics improvements 9. Forecasting/demand management – S&OP

10. Inventory optimization Asset Productivity (source: [14, 73]) Please indicate your business unit’s level of performance relative to your industry average

for the most recent reporting year. (1 = in the bottom 20% . . . 3 = in the middle 20% . . . 5 = in the top 20% of industry

performers) 1. Cash-to-cash cycle time (receivables + inventory – payables) 2. Inventory turnover (sales/inventory) 3. Asset turnover (sales/total assets) 4. Return on asset (ROA) Business Growth (sources: [14, 73]) Please indicate your business unit’s level of performance relative to your industry average

for the most recent reporting year. (continues)

BIG DATA ANALYTICS IN SUPPLY CHAIN MANAGEMENT 37

Continued

(1 = in the bottom 20% . . . 3 = in the middle 20% . . . 5 = in the top 20% of industry performers)

1. Average year on year sales growth for the last 3 years 2. Market expansion (percentage of growth coming from new markets entered in the last 3

years) 3. Market share growth Environmental Dynamism (source: [45, 91]) What is the rate of change (volatility) in your business unit’s competitive environment

relative to change in other industries? (1 = very stable . . . 3 = about average for all industries . . . 5 = very volatile)

1. The rate at which your customers’ product/service needs change. 2. The rate at which your suppliers’ skills/capabilities change. 3. The rate at which your competitors’ products/services change. 4. The rate at which your firm’s products/services change. Expected Benefits (source: [17, 78]) Using Big Data Analytics enables/will enable our organization to: (1 = strongly disagree . . .

3 = neutral . . . 5 = strongly agree) 1. Improve the quality of work 2. Make work more efficient 3. Lower costs 4. Improve customer service/patient care 5. Grow sales to new customers or new markets 6. Identify new product/service opportunities Technology Compatibility (Sources: [18, 78, 79]) Please indicate your level of agreement with the following statements: (1 = strongly

disagree . . . 3 = neutral . . . 5 = strongly agree) 1. Using Big Data Analytics is consistent with our business practices 2. Using Big Data Analytics fits our organizational culture 3. Overall, it is/will be easy to incorporate Big Data Analytics into our SCM practices Top Management Support (Sources: [2, 49]) Does the Top Management Team (TMT, e.g., CIO, COO, CSCO) support the use of Big

Data Analytics? (1 = not at all . . . 3 = somewhat . . . 5 = it is among the highest priorities) 1. To what extent does the TMT promote the use of Big Data Analytics in your

organization? 2. To what extent does the TMT create support for Big Data Analytics initiatives within

your organization? 3. To what extent has the TMT promoted Big Data Analytics as a strategic priority within

your organization? Organizational Readiness (reverse coded) (Sources: [17, 91]) To what extent are the following factors preventing your business unit from fully exploiting

Big Data Analytics? (1 = not at all . . . 3 = somewhat . . . 5 = to a great extent) 1. Lacking capital/financial resources 2. Lacking needed IT infrastructure 3. Lacking analytics capability 4. Lacking skilled resources

(continues)

38 CHEN, PRESTON, AND SWINK

Continued

Competitive Pressure (Sources: [17, 49, 76]) How many others in your industry have been implementing Big Data Analytics? (0 = don’t

know, 1 = none have adopted . . . 3 = some have adopted . . . 5 = almost all have adopted)

1. To what extent have your competitors implemented Big Data Analytics? 2. To what extent have your suppliers implemented Big Data Analytics? 3. To what extent have your customers implemented Big Data Analytics? (leave blank if

you sell to individual consumers)

BIG DATA ANALYTICS IN SUPPLY CHAIN MANAGEMENT 39

Copyright of Journal of Management Information Systems is the property of Taylor & Francis Ltd and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.