Decision Support Systems 56 (2013) 334–347
Contents lists available at ScienceDirect
Decision Support Systems journal homepage: www.elsevier.com/locate/dss
Integrating ERP and e-business: Resource complementarity in business value creation Pei-Fang Hsu ⁎ Institute of Service Science, College of Technology Management, National Tsing Hua University, No. 101, Section 2, Kuang-Fu Road, Hsinchu 30013, Taiwan
a r t i c l e
i n f o
Article history: Received 31 October 2011 Received in revised form 15 May 2013 Accepted 28 June 2013 Available online 8 July 2013 Keywords: Complementarity ERP Business integration IT integration Resource-based view (RBV) Business value
a b s t r a c t We investigate the complementary effect between ERP and e-business technologies, and the impact of such effect on business value creation. Previous studies have examined the effects of ERP and e-business technologies independently, and show positive effects on business value from their use. However, both the resource based view and microeconomic theory as well as practitioner experience suggest that the impacts from their joint and complementary use should be much greater, but this proposition has not yet been examined empirically. We use two different approaches (product term and direct measure approaches) to measure the complimentary effect. Comparing results using firm performance accounting data with self-reported survey data of 150 U.S. manufacturing firms, we provide confirming empirical evidence that the complementary effect between ERP and e-business technologies in creating business value is stronger than the main effects of ERP or e-business technologies alone. We further find that the complementary use of these IT resources to build system and business integration capabilities can extract the most complementarity value for firms. These findings provide empirical support for the theory of competitive advantage that the resource based view (RBV) proposes. Furthermore, these findings provide practical guidance to firms on how to utilize and deploy ERP and e-business technologies in a mutually reinforcing manner. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Enterprise resources planning (ERP) systems are large commercial software packages that standardize business processes and integrate business data throughout an organization [21,47,74]. These systems codify and organize an enterprise's business data into an integrated database, and transform the data into useful information that supports business decisions [68]. The ability to access information from various parts of an organization has helped firms to streamline their business processes and reduce inefficiencies [75]. In both large and medium size firms, ERP systems represent the largest portion of the application budget and about one-third of their IT budgets [41,53]. Although the benefits of ERP are considerable, traditional ERP systems that streamline and integrate internal processes improve efficiency only within the boundaries of an enterprise [21]. Because firms' value chains increasingly extend beyond their boundaries and include other firms within their business ecology, it is important to improve operational performance along the whole supply chain. Inventory turnover, asset utilization, and profitability depend on improved processes and information flows not only inside the focal firm, but also those “between” firms [50]. The full potential of an ERP system cannot be realized if its integration and coordination capabilities are confined within the walls of a firm [75]. ⁎ Tel.: +886 3 5742221. E-mail address:
[email protected]. 0167-9236/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.dss.2013.06.013
E-business technologies have exploded on the scene in the last decade, and some advocates claim that they are the ultimate solution to the information exchange problem among firms' enterprise systems. Consistent with previous studies [5,26,87], e-business technologies are defined as the Internet-based technologies, such as Extranets, Websites, and EDI communication technologies that link two firms for performing e-business functions such as online selling, online purchasing, online coordination and online information sharing. Because of their lower cost and greater ease of implementation/use, e-business technologies hold the promise of enabling information made from ERP systems to be shared among firms in the extended supply chain [4,75]. Ebusiness technologies serve to extend the original value proposition of ERP [27,36], offer an ERP-based organization the opportunity to build interactive relationships with its business partners [3,4], and bring together their previously separate information at a very low cost [55]. E-business technologies comprise the external part of the extended enterprise, and ERP comprises the internal portion [55]. From a technical point of view, Fig. 1 shows how ERP fits with ebusiness. In the middle is a focal firm's ERP system that was originally used only inside the firm. From the left hand side, with middleware software that is based on industry pre-defined standards such as RosettaNet, XML (extensive markup language) and more recently web services and service oriented architecture (SOA), information generated by ERP systems can be shared via the Internet/EDI directly with suppliers' ERP systems. If a supplier does not have an ERP system, it can still receive and exchange business information through an
P.-F. Hsu / Decision Support Systems 56 (2013) 334–347
E-Business (with Suppliers)
335
E-Business (with Customers)
Focal Firm’ sERP system
Internet Supplier without ERP system
Internet Purchasing Module
Extranet web
Internet or EDI
Material Mgmt. Module
Production Planning Module
Sales and Distribution Module
Financial Module
Human Resources Module
Extranet web
Middleware Server
Middleware Server
RosettaNet, Web Services, SOA and XML
RosettaNet, Web Services, SOA and XML
Customer without ERP system
Internet or EDI Customer
Supplier
ERP System
ERP System
Fig. 1. How ERP fits with e-business.
Extranet website. Therefore, useful information such as inventory levels, production planning and materials purchasing can be exchanged between the focal firm and suppliers via ERP and e-business technologies, as referred to business to business integration. Similarly, on the right hand side, the focal firm can exchange valuable business data with its customers, such as order status, invoice, and online order fulfillment. As more and more established organizations realize that they need to form alliances with their customers and suppliers over electronic networks, integrating e-business technologies with ERP systems become a critical issue [3,4,47]. Several IS researchers have identified ERP and e-business integration, as one of the most important IS areas for future research [11,32,75]. Others indicate that reconfiguring and integrating ERP systems with front-end web-based systems to support e-business initiatives should be at the top of the list for IS executives [67,73]. However, the extant literature investigating the value of ERP focuses on its internal integration capability only (see literature review below), and neglects the potential huge value from external integration enabled by the e-business technologies. Thus, this study brings a more complete assessment of ERP value by focusing on both internal and external integration. In other words, although existing studies have already addressed the importance and examined the benefits of using ERP and e-business technology individually, they are limited in consideration of integration between the two technologies as an important factor for firms to fully extract the benefits of IT. Our study investigates the complementary effect between ERP and e-business technologies. In summary, there are two gaps in our understanding that need more research efforts. First, how to complementarily integrate the two technologies is not well understood. We need a theoretically rigorous framework and an empirically validated measure to calibrate the complementary level of the two technologies. Based on resource based view (RBV) and microeconomic theory that provide theoretical rationale for complementarity, this study develops a theoretical framework and proposes two different approaches to measure the complementary effect—the product term approach and direct measure approach. Second, whether the complementarity between ERP and e-business technologies contributes to business value has not been assessed empirically. Whether ERP's full potential can be better realized in the e-business era has not been confirmed. Our study intends to use large scale data to validate the theoretical framework and confirm the benefits of integrating ERP and e-business technologies. 2. Literature review In this section, we first review two streams of existing studies that build our knowledge: (1) Business value of ERP and (2) business
value of e-business technologies. We then draw from resource based view and microeconomic theory to develop a theoretical framework and hypotheses for understanding the complementary effect of ERP and e-business technology.
2.1. Business value of ERP The Business value of ERP can be categorized into three facets, intangible, operational, and financial. At intangible level, Mabert et al. [42,43] reported that the most improvements after using ERP were in intangible areas such as increased interaction across the enterprise, quicker response time for information, integration of business process, and availability and quality of information. Gattiker and Goodhue [25] showed that ERP can deliver intangible benefits to firms including better information, more efficient internal business process, and better coordination between different units of a firm. At operational level, Banker et al. [5] found that ERP systems have positive impact on plant performance including product quality, product time to market, and plant efficiency, while Mabert et al. [42,43] indicated there were also operational improvements in order management, on-time deliveries, and customer interaction. Cotteleer and Bendoly [20] used longitudinal data from an ERP implemented firm to show that order fulfillment lead-time was significantly improved after ERP system deployment. Lastly, Karimi et al. [33,34] found that ERP implementation is associated with process efficiency, effectiveness, and flexibility. At financial level, mixed results are found in previous studies. Hitt et al. [31] compared data of 350 ERP adopters and non adopters and found that ERP adopters showed positive but not consistent performance results on productivity, profitability, and market value measures. They found that while ERP adopters showed a better performance on productivity, Return on Assets (ROA), inventory turnover, and profit margin, they have a significant negative performance on Return on Equity (ROE). They also found some evidence of a decline in productivity and business performance shortly after completion of the implementation. Partially replicating Hitt et al.'s work, Aral et al. [2] collected financial data of 623 US firms that were ERP adopters over a 7-year-period (1998 ~ 2005) to investigate the business value of ERP. Their results showed that using ERP systems improves productivity, inventory turnover, and asset utilization, but had no association with ROA, ROE, and Profit margin. Poston and Grabki [59] compared 54 ERP adopters and non adopters and found that ERP implementation was associated with an unexpectedly significant cost increase — Cost of Goods Sold (COGS) and Selling, General and Administrative Expenses (SG&A) — one year after implementation, and no association was found with income changes. Ranganathan and Brown [62] found that firms implement greater ERP functional scope or greater physical scope receive greater
336
P.-F. Hsu / Decision Support Systems 56 (2013) 334–347
shareholder returns. Appendix A shows more details about these studies. In summary, while the existing studies have significantly expanded our understanding of ERP business value, the results are mixed with some indicating improved value and others not. A review of IT business value research [50] suggests that in addition to looking at the IT within the firm, we should also look at interfirm IT linkages because some performance improvements such as inventory turnover, better asset utilization, or profitability depend on improved processes and information flows “between” firms. Jacobs and Bendoly [32] also pointed out that most existing ERP research focuses on the impact of an ERP system itself, but not on the much richer area of ERP extendibility. They argue that with the growing popularity of B2B and B2C e-commerce systems, there should be a strong interest in assessing how to best integrate the functionality of these systems with ERP systems to provide competitive advantage for firms. 2.2. Business value of e-business technologies Several studies have examined the relationship between e-business technologies and firm performance. Using self-reported survey data, Lederer et al. [37], Barua et al. [9] and Zhu et al. [87,89] found a positive and significant relationship between e-business use and firm performance. However, they stressed the limitation in their studies that the subjective performance measures could potentially induce biases, and therefore, firm level accounting data is needed to confirm these findings. Yet, using objective accounting data, Zhu et al.'s [86,88] studies showed that e-business use is significantly associated with Cost of Goods Sold and inventory turnover, but it had weak or no association with Return on Assets (ROA) and Gross Margin. Appendix B shows a summary of these studies. IS researchers indicate that the limitation of current e-business studies is neglecting the important role of ERP in e-business settings, and encourage future studies on e-business to focus on more specific questions about how a firm integrates the Internet with its existing internal IS such as ERP systems [10,87]. Jacobs and Bendoly [32] pointed out that while buzzwords like “B2B”, “B2C”, and just about anything else preceded by an “e-” seem to have taken center stage, yet ironically, each of these new terms at their most basic levels represent extensions of ERP systems to the customers and to the suppliers. A true e-business enabled firm needs the support from a well-tuned ERP system, since ERP is the core to fulfill the promises made on the web pages. Without clean internal processes and data that are provided by ERP systems, e-business may be just flashy web pages with no real substance behind them [55]. Overall, the foregoing studies of ERP and e-business technologies indicate that they each could contribute to the business value of IT. However, none of the previous studies examined the role of ERP systems linked with e-business technologies in interfirm or interorganizational systems. Melville et al. [50] and Mukhopadhyay and Kekre [54] suggest that such interorganizational linkages might produce greater benefits than either technology alone. Consequently we propose to investigate both the independent and the complementary effect of ERP systems and e-business technologies on business performance. In order to have a solid theoretical framework to guide the research, we draw from the resource-based theory and microeconomic theory to develop theoretical propositions.
survive imitation because of isolating mechanisms such as history dependence, causal ambiguity, and social complexity [8]. The RBV (or its variations) has been applied by information system (IS) researchers to analyze the business value of IT (see Wade and Hulland [81] for a review). Resources and capabilities are two terms that have been frequently used in the RBV theory. This paper distinguishes between “resources” and “capabilities” based on the definitions in RBV literature [1,28,29,44,81]. Resources are inputs into a firm's production process, such as capital, equipment, information systems, and individual employees. Capabilities, in contrast, refer to a firm's capacity to deploy resources using organizational processes. Capabilities can be viewed as the capacity of a bundle of resources to perform some task or activity. Through continued use, capabilities become more difficult for competitors to understand and imitate [28,29,63,81]. Most previous studies based on the RBV posit a direct relationship between IT resources and firm performance [12,49,65]. More recently, some researchers have emphasized that the IT resource is likely to affect firm performance only when it is deployed to create unique “complementarities” with other IT or other firm resources [13,60,61,63,76]. Complementarity represents an enhancement of resource value and arises when a resource produces greater returns in the presence of another resource than by itself [51]. Resources rarely act alone in creating or sustaining competitive advantage, and this is particularly true of IT resources that, in almost all cases, act in conjunction with other firm resources to provide strategic benefits [81]. IT-based success rests on the ability to “fit the pieces together” complementarily [63]. The importance of complementarity is not only argued by the RBV researchers, but also by microeconomic scholars. The idea of optimization in microeconomic theory [51] provides an approach to model complementarity formally. Using microeconomic theory, Milgrom et al. [51,52] provide a simple example in the following to explain why integrating complements (in our paper, ERP and e-business systems) is difficult, slow, hard to be duplicated, and if they are successfully integrated, it may bring firms more profits. Inside a firm, two managers in different departments control inputs x and y separately. Both of them seek to maximize profits as given by the entries in the Table 1. The payoff function f(x, y) depends on the parameter Θ. If Θ increases from 0 to 2, the optimal levels of x and y rise from (low, low) to (high, high). However, since we usually do not know how Θ changes (increase or decrease, and the speed of increase or decrease as a function of x and y), it's not easy to find the optimal level of x and y to maximize the firm's total profits. Furthermore, if the two managers make their decisions separately on how much x and y should be invested, (i.e. x and y are not fully coordinated or successfully integrated), suboptimal decisions such as (low, high) or (high, low) will happen. Milgrom et al. [51] argue that, “Suppose Θ increases from some value Θ′ b 1 to a value Θ″ N 1. No amount of individual, uncoordinated search will find an improvement, and the system can get stuck at the suboptimal position. This example illustrates how strong complementarities make it more likely that individual adaptations will fail to converge upon optimal results and why change in a system marked by complementarities may be difficult. Changing only a few of the system elements at a time to their optimal values may not come at all close to achieving all the benefits that are available through a fully coordinated move, and may even have negative payoffs.” Borrowing the same idea from the example, we understand why practitioners often claim it is not easy to integrate two complements
2.3. Theoretical background: Resource complementarity in RBV and microeconomic theory The Resource-Based View (RBV) argues that firm resources are heterogeneously distributed across firms. When the firm resources are valuable, rare, imperfectly imitable, and nonsubstitutable, they could create competitive advantages, which in turn could explain the differences in firm performance [8,28,84]. Moreover, resources tend to
Table 1 Milgrom et al.'s example. Payoff function: f(x, y)
Low y
High y
Low x High x
5 3
4 4+Θ
P.-F. Hsu / Decision Support Systems 56 (2013) 334–347
(ERP and e-business systems) at their optimal levels to receive the highest total value. It takes time, effort, and experience to find the characteristics of Θ. The know-how of extracting the most complementarity value (Θ) is firm specific. In summary, why does complementarity between IT resources make it more possible for firms to achieve competitive advantage than one resource alone? Based on the RBV literature and microeconomic theory, we propose two explanations for the role of complementarity in IS context. First, in general, physical technology such as a complex information system, by itself is typically imitable. If one firm can purchase these physical tools of production, then other firms should also be able to purchase these physical tools, and thus such tools should not be a source of sustained competitive advantage [8]. On the other hand, if a firm can exploit physical technology involving the use of socially complex firm resources, the synergies among them are far more difficult to imitate [9,88]. Several firms may all possess the same physical technology, but only some of these firms may possess the social relations, culture, traditions, etc. to fully exploit this technology in implementing strategies [8]. Second, complementarity means not only the co-presence of the two resources as indicated above, but also that the two resources are used in a mutually reinforcing manner. How effective a firm is in using two ITs in a reinforcing manner to support and enhance its business core competencies is difficult. Therefore, complementarily leveraging resources is considered a firm-specific capability [29,44]. Numerous studies have commented that integrating IT resources (systems) to build a flexible and sophisticated IT infrastructure requires both considerable time and expertise [6,47,61,63]. Although the individual components that go into the infrastructure are commodity-like, the process of integrating the components to develop an infrastructure tailored to a firm's strategic context is complex and imperfectly understood [12].
3. Model and hypotheses Since the above theories provide two different rationales to explain complementarity, we propose two corresponding research models to investigate ERP, e-business technologies, and their complementary effect on business value. Model 1 represents the first rationale that the theories provide: Firm resources are considered complementary when the presence of one resource enhances the effect of another resource. This interaction perspective of complementarity is typically operationalized using multiplicative terms in statistical analysis [63]. Hence, we measure the level of complementarity by the product of ERP and e-business technologies variables (Fig. 2).
337
Model 2 represents the second rationale that the theories provide: Resource complementarity is based on how resources are utilized and deployed; complementaries arise when resources are used in a mutually reinforcing manner [63]. RBV and microeconomic theory argue that complementarily deploying and utilizing resources is a firmspecific capability. Model 2 assesses complementarity by measuring how ERP and e-business technologies are integrated and utilized at two levels (Fig. 3). The first level measures how the two technologies are integrated at the information system level. The second measures how the integrated ERP and e-business system is utilized at the business process coordination level (left-most bubbles in Fig. 3). The reason we gage the complementarity at two levels is based on Markus's distinction between information system integration and business process coordination [48], which has also been proposed by many IS researchers [6,24,61,62]. System integration refers to the creation of tighter linkages between different computer-based information systems and databases. Business process coordination represents the extent to which the business process of two firms are tightly coordinated and standardized through information systems [6,78]. System integration is required to achieve business coordination; however, even when system integration is achieved, the goals of business coordination may not be realized. Due to the concerns about information leakage, firms usually are reluctant to exchange business information with their business partners [16]. Two firms might both achieve a high level of system integration, but their business coordination level might vary. System integration is viewed as a prerequisite and facilitator of business coordination, but does not guarantee a firm's willingness to achieve a higher level of business coordination [48,61]. Rai et al.'s study [61] on supply chain integration capabilities distinguishes “IT infrastructure integration” and “supply chain process integration”. They argue that lower-order IT integration capability enables higher-order process integration capabilities. We adopt the same view and measure both system integration and business process coordination in this study. In addition to the theoretical arguments, there are two other methodological advantages that motivate us to use the two different approaches: First, guidelines for research on complementary effect (interaction effect or fit) suggest that researchers should compare the utility of different statistical techniques using the same data set because each technique has implied biases [77]. Studies should be designed to permit comparative evaluation of as many forms of “fit” as possible [80]. Second, using the directly measured and observable items to assess the complementarity construct, we can unpack the complementarity effect and thereby provide practical guidelines to practitioners about how to use the two technologies to achieve complementarity.
Complementary Effect between ERP and E-Business Technologies (Product Term Approach) Cost Efficinecy
ERP modules H1 H2
Business Value
Differentiation
H3
E-Business technologies for communication
Intangible ERP*EB Interaction Control Variables: Firm Size, Industry, Year, ERP Vendor
Fig. 2. Research model 1.
338
P.-F. Hsu / Decision Support Systems 56 (2013) 334–347
Complementary Effect between ERP and E-Business Technologies (Direct Measure Approach)
ERP modules
Cost Efficinecy H 1
E-Business technologies for communication
Business Value
Differentiation
H
3
ERP and EB System Integration
H2
Intangible
ERP and EB Complementaries ERP and EB Business Process Coordination
Control Variables: Firm Size, Industry, Year, ERP Vendor
Fig. 3. Research model 2.
3.1. Hypothesis development Based on the theoretical foundation discussed above, ERP systems focus on internal process efficiency and effectiveness, and can coordinate information across different departments within a company. ERP systems are expected to affect internal firm operations by decreasing internal coordination costs [6,59]. On the other hand, e-business technologies are focused on external, cross-enterprise process efficiency and effectiveness. They can reduce external coordination costs and reap the benefits of supply chain integration [40,59]. Therefore, we propose the following hypotheses: H1. A firm with more complete ERP modules to provide cross-functional integration is more likely to gain business value. H2. A firm with more complete e-business technologies to provide inter-firm integration is more likely to gain business value. Integrating ERP and e-business is extremely complex. First, when implementing an ERP system alone, it comes in numerous configuration tables that must be customized to suit a firm's business needs, and the customization is a challenging task [47]. ERP implementation also requires substantive changes in business processes, routines, and roles, as referred to business process reengineering (BRP), which is also a difficult project [6,47,61,63]. Going a further step to integrate ERP and e-business technologies to build a digitalized platform that link business partners to perform business processes electronically is even more complicated. Business processes such as procurement and fulfillment are inherently complex, since it involves not only the focal firm itself, but also its business partners. As firms integrating ERP and e-business systems to link key suppliers and customers, they have to reengineer business processes and IT systems not only the focal firm itself, but also all the business partners in the supply chain [6,12]. The new business processes that are supported by a well integrated ERP and e-business system are like dominoes in a row; each new transaction sets off a cascade of new events [15,31]. Such complexity of integration requires firms' specific knowledge and capacity, and is not easily imitated by others [66]. Therefore, integrating ERP systems and e-business technologies to enable business processes electronically transferred among focal firm and its business partners is considered hard to be substituted from RBV's perspective, and is proposed as a source of competitive advantage. Bendloy et al. [11] argue that once entire value chains are acting as formidable entities, using inter-firm ITs such as ERP and e-business technologies to cooperate, the structures of these partnered communities are hard to duplicate. The idiosyncrasy strengthens the sustainability of the competitive advantages [11]. Although the individual IT itself could contribute to firm performance, the process of integrating the
individual IT components to develop a platform tailored to a firm's strategic context is much more complex, imperfectly understood, and could contribute more business value to firms [12]. Accordingly, we hypothesize that H3. The complementary effect between ERP and e-business technologies is stronger than the main effects of ERP or e-business technologies alone. Lastly, since there are two perspectives on complementarity argued in the theory: co-presence of two resources (model 1), and how the two resources are utilized and deployed in a firm (model 2), we want to test the different perspectives on complementarity. This is to respond to the call that researchers should compare the utility of different statistical techniques and bring as many forms of “fit” as possible using the same data set to understand complementarity [77]. As we think that how two resources are utilized, deployed, and used in a mutually reinforcing manner is much more difficult to be imitated by competitors than co-presence of the two resources, we hypothesize that H4. The complementarity based on how the two resources are utilized and deployed in a firm could contribute more in creating business value than the complementarity based on IT resource-copresence. 4. Methodology 4.1. Data To test our research model, a questionnaire was designed to collect data on each of the variables in the model. We contracted SRBI Inc. (Schulman, Ronca & Bucuvalus), which is a professional survey firm that specializes in large-scale survey research, to conduct a telephone survey. The sampling was selected randomly within U.S. manufacturing industry. Interviews were conducted only with those companies that make use of ERP in conducting their business. Eligible respondents to the survey were the individuals who are considered the most knowledgeable about ERP and e-business use in their companies, such as a CIO or IS manager to have the best quality of data. Our target completes were 150 interviews. In total, 1813 potential respondents were contacted via telephone, with 226 firms finishing the survey. Among the 226 firms, 76 firms were not ERP users; the remaining 150 firms provided usable data. The responsible rate is comparable to previous large-scale ERP and e-business surveys as reported in [5,25,42,43]. Table 2 shows the sample characteristics. The distribution of firm size measured by employee numbers reflects a balance of large and small firms. Computer, electronic products, and instruments are the top three ERP using industries. Around half (46.7%) of the sampling firms have 5–10 year ERP using experience. The majority (80%) of
P.-F. Hsu / Decision Support Systems 56 (2013) 334–347 Table 2 Sample characteristics (N = 150).
339
Table 3 Descriptive statistics for ERP and e-business technologies.
Employees
#
%
Industry sector
#
%
ERP Modules:
b100 100–199 200–499 500–999 1000–4999 N4999
23 35 40 14 25 13
15.3% 23.3% 26.7% 9.3% 16.7% 8.7%
Food Apparel and fabric Furniture and fixtures Chemicals and allied products Rubber and plastics Leather Metal and fabricate metal Computer Electronic products Transportation equipment Measuring, controlling & medical instruments Miscellaneous manufacturing
5 3 4 9 5 4 4 21 40 5 35
3.3% 2.0% 2.7% 6.0% 3.3% 2.7% 2.7% 14% 26.7% 3.3% 13.3%
Transactional modules:
15
10% %
Years of ERP use
#
%
Respondents
#
b1 year 1–5 years 5–10 years 11–15 years N15 years
3 22 70 28 27
2.0% 14.7% 46.7% 18.7% 18.0%
CIO IS manager Network/database administrator CFO/accounting manager Othersa
74 46 12 6 12
49.3% 30.7% 8.0% 4.0% 8.0%
a Others include vice president, general manager, COO (operation), and customer service/sales manager.
the respondents are CIOs or IS managers. We compared the profile of the responding firms with non-responding firms on demographic variables such as firm size and revenue using Chi-square analysis. The results indicated no significant response bias. We also examined common method bias that may potentially occur in survey data. Using Harman's single-factor test [57], we found that one general factor cannot account for the data variance, which indicates there is no significant common method bias in our dataset.1 Although ERP systems are designed with many modules to provide integrated support, firms seem selective when choosing modules to implement (Table 3). Transactional-oriented ERP modules, such as inventory, purchasing, production planning, sales/order entry, and finance modules, are used by more than 90% of the reported cases. In contrast, analytical modules such as advanced planning and scheduling (APS), data warehouse, and CRM modules that analyze data provided by the transactional modules are used less often. As for e-business technologies, websites are used by more firms (97%), while interorganizational networks such as Extranet and EDI technologies are used by fewer firms (37–51%). 4.2. Operationalization of constructs Constructs and measurement items used in this research are adapted from previously validated measures, or are developed on the basis of literature review. The process of operationalization of constructs as well as prior research support, are discussed below and summarized in Appendix C. ERP modules refer to the scope of ERP system functions a firm chooses to implement. Based on Porter's value chain concepts, ERP modules can be of two types: value-chain based transactional modules and enterprise support analytical modules [62]. Transactional modules are aimed at integrating a firm's value chain activities: from upstream material purchasing and inventory control modules, to the focal firm's manufacturing module, to the downstream customer facing sales/ order entry module. In contrast, analytical modules focus on analyzing the raw data collected by the transactional modules to support business decision making. These modules are often called ERP add-on modules or Bolt-Ons. We include three analytical modules suggested to have 1 Due to page limit, we do not provide details of the test results here. We will provide the full results upon request.
Analytical modules:
E-business technologies for communication:
Adoption Rate Inventory module Purchasing module Manufacturing module Sales/order entry module Financial and accounting module Advanced planning and scheduling Data warehouse/business intelligence Customer relationship management Website Extranet EDI
96% 95% 90% 95% 97% 59% 40% 36% 97% 37% 51%
significant impacts on firm performance: advanced planning and scheduling, data warehouse/business intelligence, and customer relationship management. The way we measure the construct “ERP modules” is similar to previous studies [5,31,42,43,75]. E-business technologies for communication refer to Internet-based technologies that serve as channels for communication between two firms (inter-organization) or two departments within a firm (intraorganization) for performing e-business functions [26,87]. Following Banker et al. [5], Geoffrion and Krishnan [26], and Zhu et al. [87], we include three e-business technologies: extranet, websites, and EDI into the research. Extranets are members-only networks run by individual organizations to directly link two parties, and are implemented as virtual private networks on the public Internet. Websites are usually publicly assessable. They serve as the most economical and basic means for performing e-business functions. Although websites usually have limited e-business functionalities, they are the most affordable electronic channels, and are included for their popularity in e-business. EDI is included as an e-business technology for two reasons. First, traditional EDI or VAN-based (Value Added Networks) EDI predates the Internet, and has gradually moved toward Internet- based technology, called Internet-based EDI. The EDI technology has been used for more than two decades, and is still used by many firms for performing B2B transactions. It is very common that both EDI and Internet technologies co-exist in a firm. In addition, IS researchers usually include EDI transactions as part of e-business [5,26,50,71,87], and we adopt the same approach to measure the e-business construct. System integration represents the extent to which different information systems are interconnected and can talk to one another [6,45]. It is conceptualized to include the extent to which information systems are integrated internally (both across functions, and between e-commerce and traditional activities) and externally (along the value chain from suppliers to end customers) [6,7,78,85]. Drawing upon this argument, we measure ERP and e-business system integration by three items: internally, (1) the extent a firm's ERP system is integrated with its own front-end e-business systems; externally, (2) the extent a firm's ERP system is directly integrated with its business partners' information systems, and (3) the extent a firm's ERP system is assessable by its business partners via electronic networks. Following both supply chain literature and IS literature, business process coordination is defined as the extent to which important operational information is shared or conducted via electronic networks/systems between a focal firm and its business partners [6,19,48,61,69,70]. Specially, we consider the sharing of inventory information, demand information, production planning information, customer orders and services as indicators of business process coordination. The five kinds of information are selected because they are suggested to have significant influence on firm performance in the supply chain literature (see discussion below), are validated items used in a previous study [13,23,61], and are identified as information that firms are more willing to share in practice when we examine content validity with consultants and industry experts.
340
P.-F. Hsu / Decision Support Systems 56 (2013) 334–347
Lee et al.'s [38] study indicates that sharing of inventory information can reduce total inventory holding level in a firm and in a supply chain thereby decreasing inventory costs. Sharing of demand related information improves forecasting and replenishment [61,75]. In contrast, with no sharing of demand related information, distorted demand signals would be amplified and transferred upstream across the supply chain to cause bullwhip effect [38]. Sharing of production planning information can enhance operational efficiencies through improved coordination of allocated resources, activities, and roles across the supply chain [39,61]. Customer orders and services conducted online improve customer satisfaction and increase efficiency [61]. Our study specifically measures the level that a firm uses electronic networks (e-business technologies) rather than traditional means (face to face, telephone, fax etc.) to share operational information stored in ERP systems. The reason is that only real-time information sharing enabled by e-business technologies allows firms in supply chains to timely synchronize production, coordinate inventory-related decisions, and develop a shared understanding of performance bottlenecks [14,40]. While system integration is concerned with a firm's technical capability to link information systems along the supply chain, business process coordination measures the information that actually is shared or conducted among supply chain partners. This view echoes Markus's propositions [45], and is consistent with prior studies [6,61]. Business value attributable to ERP is measured by three subconstructs — cost efficiency, differentiation, and intangible benefits — the three aspects of ERP related performance that concern firms and that the IS literature often uses [61,63]. They are recognized as resources for creating competitive advantages, and are important performance impacts after implementing ERP [20]. While Porter [58] argues that firms must balance cost efficiency and differentiation, and may choose either one to master, the ERP literature argues that ERP systems have the ability to improve all the three dimensions (see literature review of ERP discussed above). Thus, business value is conceptualized as a second-order construct, manifested in the three dimensions.2 Whereas subjective measures of firm performance from the survey were used in the study, we also validated these measures using objective accountingbased firm performance data. We collected inventory costs (INVT), Cost of Goods Sold (COGS), and Selling, General and Administrative Expense (SG&A) data for firms in our sample, following previous studies [31,59,86]. Firm size, industry, year, and ERP vendor are used as control variables in the model. Literature suggests that companies of different sizes tend to perceive different ERP benefits [43]. Thus we include firm size measured by the number of employees into the model. Furthermore, Melville et al.'s review study of IT business value [50] concludes that certain industries attain higher IT business impacts and greater cost reduction than others. They argue that in time-sensitive industries such as personal computers, there is an ample opportunity to apply IT to reduce cycle times, better manage inventory, and improve customer satisfaction. In our sample, we also found that high-tech manufacturing industries (computer, electronic product, and instruments) adopt ERP systems more often than traditional manufacturing industries. We then add an industry dummy variable (high tech vs. traditional) to investigate the industry effect. Finally, firms are eager to know when their investments on ERP systems can be paid back and which ERP vendors' product may be better. We include the year and ERP vendor dummy variables to investigate the issues.
ability of PLS to model both formative and reflective constructs makes it appropriate for conducting our research. For reflective constructs, we examined convergent validity, construct reliability, and discriminant validity. The measurement properties are provided in Table 4. Construct reliability measures the stability of the scale based on an assessment of the internal consistency of the items measuring the construct. In Table 4, all the reflective constructs have a composite reliability over the cutoff of 0.70, as suggested by Straub [72]. Convergent validity is verified through the t-statistic for each factor loading. As shown in Table 4, all factor loadings are greater than the typical cutoff value of 0.5 and significant at the p b 0.01 level. Discriminant validity measures the extent to which different constructs diverge from one another. In Table 5, the diagonal elements represent the square root of average variance extracted (AVE), providing a measure of the variance shared between a construct and its indicators. The square root of AVE is required to be larger than the correlations between constructs, i.e., the off-diagonal elements to meet discriminant validity [22,30]. The reflective constructs used in the model meet the criterion. For formative constructs, we check the four criteria suggest by Rai et al. [61] and Petter et al. [56],3 and found that the two constructs used in our study (ERP modules and e-Business technologies) should be modeled formatively. The weights of measurement items in the two constructs are all significant (Table 6), which suggest that our formative constructs have good quality [17,56]. Two second-order constructs are used in the model: “ERP and e-business complementarity” and “business value”. We use the secondorder construct because it is a general, more global factor that explains all the covariation among the first order factors [17,64]. A second order construct is modeled as being at a higher level of abstraction, and is used commonly in IS literature [17,61]. “ERP and e-business complementarity” is composed of two first order factors (SI and BPC), based on IS literature's distinction between information system integration and business process coordination [6,48]. System integration (SI) measures how the two technologies are integrated at the information system level, while business process cooperation (BPC) measures how the integrated ERP and e-business system is utilized at the business process coordination level. System integration is viewed as a prerequisite and facilitator of business coordination [48]. Rai et al.'s study [61] on supply chain integration capabilities distinguishes “IT infrastructure integration” and “supply chain process integration”. We adopt the same view and measure “ERP and e-business complementarity” both at system integration level and business process coordination level. “Business value” is measured by three sub-constructs — cost efficiency, differentiation, and intangible benefits, since the ERP literature argues that ERP systems have the ability to improve all the three dimensions. Validity of the second-order constructs is shown in Table 7. The paths from the second-order construct to the first-order factors are significant and of high magnitude, greater than the suggested cutoff of 0.7 [17].
4.3. Instrument validation
Model 1 uses the most common approach of testing a complementary effect — the product terms approach. Fig. 4 shows that both ERP
To validate the instruments, we conducted a confirmatory factor analysis using partial least squares (PLS-Graph version 3.00). The 2 Please note these firm measures do not reflect the overall performance of the firm, but only the improvements that are related to ERP implementation.
5. Empirical results We tested the two research models using PLS-Graph version 3.00. Both models were run using standardized construct values, and the standardized path coefficients can be interpreted and compared directly. 5.1. Empirical results of model 1
3 Whether a construct should be modeled as formative or reflective could be judged by four criteria (Rai et al. 2006): (1) direction of causality from construct to indicators, (2) interchangeablity of indictors, (3) covariation among indicators, and (4) nomological net of construct indicators.
P.-F. Hsu / Decision Support Systems 56 (2013) 334–347 Table 4 Reflective constructs: reliability, and convergent validity. Reflective constructs
Indicators Loading 0.779⁎⁎⁎ 0.815⁎⁎⁎ 0.822⁎⁎⁎ 0.877⁎⁎⁎ 0.822⁎⁎⁎ 0.735⁎⁎⁎ 0.744⁎⁎⁎ 0.708⁎⁎⁎ 0.757⁎⁎⁎ 0.833⁎⁎⁎ 0.788⁎⁎⁎ 0.808⁎⁎⁎ 0.748⁎⁎⁎ 0.723⁎⁎⁎ 0.871⁎⁎⁎ 0.859⁎⁎⁎
ERP and EB system integration SI1 SI2 SI3 Business process coordination BPC1 BPC2 BPC3 BPC4 BPC5 Cost efficiency CE1 CE2 CE3 Differentiation DF1 DF2 DF3 Intangible benefits IN1 IN2
Table 6 Formative constructs.
Convergent Reliability validity (t-stat) 17.87 21.60 23.28 10.84 7.13 5.79 6.79 4.17 9.90 32.77 17.29 17.77 13.21 9.33 61.67 22.61
Formative constructs
Indicators
Weight
ERP modules
TRA ANA EB1 EB2 EB3
0.439⁎⁎⁎ 0.769⁎⁎⁎ 0.447⁎⁎⁎ 0.369⁎⁎⁎ 0.643⁎⁎⁎
0.847 E-business technologies for communication 0.819 ⁎⁎⁎ p b 0.01.
0.836
0.805
0.858
⁎⁎⁎ p b 0.01.
modules and e-business technologies have positive and significant affects on business value (0.266*** and 0.247*** respectively). The results indicate that the more comprehensive ERP modules and e-business technologies a firm implements, the higher business value it may receive. Hypotheses H1 and H2 are supported. The interaction effect between ERP and e-business technologies is strong and significant (0.158**). The result shows that the presence of one resource enhances the value of another resource. One unit investment in e-business technologies could increase ERP's contribution to business value by 0.158 units. The model can explain 18.4% of the variance in firm performance. We further break down the ERP construct into two items (transactional and analytical modules) and compare the contribution of each of the items. This comparison helped us understand which ERP submodules are more important than others in terms of their effects on final firm performance. The factor weight of analytical modules is 0.712, greater than that of the transactional modules (0.510), which indicates that the analytical modules might contribute more to business value. Recalling that the analytical modules are used less by firms compared to the transactional modules, here we provide evidence to support RBV theory's argument: when a resource is rarer and valuable, it can make a greater contribution to generating rents. The results also suggest that firms, which have invested much in the traditional ERP modules that focus on collecting transactional data, should now invest in the analytical modules that can further analyze transactional data to support business decision making. Control variables provide some further interesting findings. First, firms are eager to know when their investments on ERP systems can be paid back. Using year dummy variables, we found that firms start to feel some business value one year after their ERP implementation (0.165*), before that, the business value is marginal (0.066). The most significant business value increment happens after five years of using ERP systems (0.285***). Industry experts predict a four to five year Table 5 Reflective constructs: Discriminant validity of instruments.
learning time for ERP implementation [35,82], and our data confirms their prediction. Firms need to adjust their business processes to best fit the ERP systems, and it therefore takes some time to see the performance improvement. Second, firm size is negatively associated with ERP business value, which suggests that smaller firms perceive more performance improvement. One possible reason provided by a previous study [43] as well as our survey results is that smaller firms are more likely to change their business processes to fit ERP system, which can minimize complicated and costly system modifications and may result in higher perceived performance improvement. On the other hand, larger firms usually have complex operations and organizational structures, and customization is usually unavoidable [45,46]. Mabert and Soni [43] find that there is significant difference between small and large firms in terms of their ERP customization level; large firms customize more. Third, industry and ERP vendor control variables are insignificant, which suggests that high-tech and traditional manufacturing industries, and firms that use different ERP products, perceive no difference in business value creation. 5.2. Empirical results of model 2 Fig. 5 shows the results of model 2. ERP modules and e-business technologies again show positive and significant impacts on business value (0.202*** and 0.147***). More important, the complementary effect between ERP and e-business technologies is stronger than the main effects of ERP or e-business technologies alone (H3 is supported), and the magnitude (0.321***) is also greater than that in the model 1 (0.158**). The result further shows that while the co-presence of two technologies is likely to be a source of competitive advantage, resource complementarity based on how the two resources are utilized and deployed in a firm could contribute more in creating business value for firms. Thus, hypothesis 4 is supported. 5.2.1. Robustness test of the complementary effects While we can directly compare the path coefficients of the two complementary constructs in model 1 and model 2, we also calculated effective size of the two interaction terms to understand their relative strength. Effective size provides a more robust estimation of the degree to which a phenomenon such as an interaction effect or a complementarity effect exists in a population [18]. The standard approach for determining the strength of an interaction effect involves contrasting the
Table 7 Measurement model: Second-order constructs.
Constructs
(1)
(2)
(3)
(1) (2) (3) (4) (5)
0.649 0.253 0.046 0.215 0.042
0.703 0.205 0.346 0.092
0.630 0.195 0.289
ERP and EB system integration Business process coordination Cost efficiency Differentiation Intangible benefits
341
(4)
0.579 0.460
(5)
0.865
Note: Diagonal elements are the square root of average variance extracted (AVE), which, for discriminant validity, should be larger than interconstruct correlations (off-diagonal elements).
Second order constructs
First order constructs
ERP and e-business System integration (SI) complementarity Business process coordination (BPC) Business value Cost efficiency Differentiation Intangible benefits ⁎⁎⁎ p b 0.01.
Loadings t-stat
Composite reliability
0.803⁎⁎⁎ 9.75 0.843 0.877⁎⁎⁎ 19.48 0.705⁎⁎⁎ 7.36 0.805 0.772⁎⁎⁎ 8.21 0.787⁎⁎⁎ 17.17
P.-F. Hsu / Decision Support Systems 56 (2013) 334–347
Analytical
0 .5
0.
10 *
Cost Efficinecy
ERP modules
0.
*** 71 2
6*
R2=18.4%
0.247***
Business Value
0.770***
0. 15 Y 0-1
Y1-5
0.1 65 * 0.285** *
*
0.
6 06
Y 5-10
Intangible 0 .0
** 94 .1 -0 1* 0.10
ERP*EB Interaction
*** p
