ARTICLE IN PRESS

International Journal of Information Management 28 (2008) 391402 www.elsevier.com/locate/ijinfomgt

Implications of the t between organizational structure and ERP: A structural contingency theory perspective
Neil A. Morton, Qing Hu
Department of Information Technology and Operations Management, Barry Kaye College of Business, Florida Atlantic University, Boca Raton, FL 33431-0991, USA

Abstract Despite the tremendous popularity and great potential, the eld of Enterprise Resource Planning (ERP) adoption and implementation is littered with remarkable failures. Though many contributing factors have been cited in the literature, we argue that the integrated nature of ERP systems, which generally requires an organization to adopt standardized business processes reected in the design of the software, is a key factor contributing to these failures. We submit that the integration and standardization imposed by most ERP systems may not be suitable for all types of organizations and thus the t between the characteristics of the adopting organization and the standardized business process designs embedded in the adopted ERP system affects the likelihood of implementation success or failure. In this paper, we use the structural contingency theory to identify a set of dimensions of organizational structure and ERP system characteristics that can be used to gauge the degree of t, thus providing some insights into successful ERP implementations. Propositions are developed based on analyses regarding the success of ERP implementations in different types of organizations. These propositions also provide directions for future research that might lead to prescriptive guidelines for managers of organizations contemplating implementing ERP systems. r 2008 Elsevier Ltd. All rights reserved.
Keywords: ERP; ERP implementation; Contingency theory; Organizational structure

1. Introduction Enterprise Resource Planning (ERP) systems have become the backbone of IT infrastructure in most largeand medium-sized organizations across the world. ERP systems are essentially software packages that integrate information ow across business functions and unit boundaries, and even among business partners. The benets associated with ERP systems are both tangible and intangible (Murphy & Simon, 2002), and could be reected in operational, managerial, strategic, infrastructural, and organizational dimensions of a business (Shang & Seddon, 2002). When implemented successfully, ERP systems can reduce cycle time, enable faster business transactions, facilitate better management, and enable e-commerce integration (Davenport, 2000). However, the
Corresponding author. Tel.: +1 561 297 3675; fax: +1 561 297 3043.

E-mail addresses: nmorton2@fau.edu (N.A. Morton), qhu@fau.edu (Q. Hu). 0268-4012/$ - see front matter r 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.ijinfomgt.2008.01.008

successful implementation of ERP systems has proven to be a difcult task. In one estimate, over half of ERP implementations are judged to be failures (Scheer & Habermann, 2000). The implementation of an ERP system generally requires an organization to adopt the standardized business processes embedded in the software and to move away from a functionbased organizational structure in favor of an integrated, process-oriented structure (Al-Mashari, 2003; Benders, Batenburg, & van der Blonk, 2006; Davenport, 1998). While such effects may be desirable to many organizations, they may not be appropriate for all (Davenport, 1998; Davison, 2002). Researchers have long recognized that a critical determinant of an information systems success within an organization is the t between the design of the system and the organization (Markus & Robey, 1983, 1988). Accordingly, the internal structure of an ERP system is not necessarily aligned with the implementing organizations existing structure (Soh, Sia, & Boh, 2003). However, because of the promising strategic and operational benets

ARTICLE IN PRESS
392 N.A. Morton, Q. Hu / International Journal of Information Management 28 (2008) 391402

that may occur after the implementation of an ERP system, an organizations top management is often tempted to adopt ERP without understanding the consequences of a potential mist between the system and the organization. A complex technological innovation, such as ERP, may be adopted by an organization without justications of performance related benets but under the pressure of institutional forces (DiMaggio & Powell, 1983; Gosain, 2004; Teo, Wei, & Benbasat, 2003). Indeed, inappropriate adoptions of ERP have been noted in the press and academic studies, along with reports of implementation failures and signicant nancial losses. As a consequence, ERP success has been a focal point of academic research over the last decade. Multiple streams of research exist on the critical factors required for successful ERP implementation as well as the impact of ERP on various performance and effectiveness measures (AlMashari, 2003; Hitt, Wu, & Zhou, 2002; Holland & Light, 1999). Unfortunately, most of the research on critical success factors is void of a strong theoretical foundation (Robey, Ross, & Boudreau, 2002). Some of these studies have investigated the issue of t or alignment between ERP systems and organizations. One notable example is a study by Hong and Kim (2002). The authors found that the t between ERP and organization, in terms of data, process, and user interface, had a signicant effect on implementation success. Another is Soh, Sia, and Tay-Yap (2000) who argued that the mist between ERP and organization is greater in Asia, because the underlying structure of most ERP systems is inuenced by the US and European business processes which in many cases are substantially different from those prevalent in businesses in Asia. However, few researchers have addressed the critical issue of t between ERP and organizational structure with a strong theoretical lens. We argue that it is important for both researchers and practitioners to understand: (1) which types of organization may t with ERP systems and which may pose signicant challenges? (2) from what dimensions could the t between an organization and an ERP system be examined? and (3) what precautions could an organization take to mitigate the risk of mist? For this purpose, we draw from Mintzbergs (1979) work on ideal organization structural types and Donaldsons (2001) work on structural contingency theory to examine the issue of t between organizational structures and ERP systems. In doing so, we hope to ll a signicant gap in the current discussion on ERP implementation and open new paradigms for future research on this important issue. The balance of the paper is arranged as follows. We rst review the literature on the relationship between technology and organizational structure and structural contingency theory, from which we develop our research framework. We then identify contingency variables associated with ERP systems and their related structural characteristics or dimensions. Propositions concerning the relationship between ERP systems and a set of organization structural types are developed based on the framework and the

variables identied. Finally, we submit concluding observations and discuss potential future research directions. 2. Theoretical background 2.1. Technology and organization structure The fundamental issue in this study is the relationship between information technology and organizational structure. Leavitt and Whisler (1958) argued that information technology could have a signicant impact on the shape of organizations and the nature of managerial jobs. They predicted that entire levels of middle management would be eliminated, and centralized decision-making would reside at the highest levels of the corporate hierarchy. Subsequent literature produced mixed ndings on nearly all aspects of this relationship (Attewell & Rule, 1984). The prediction of centralization has proven true in some instances, while greater decentralization has occurred in others. Some researchers agreed with the prediction of centralization (Reif, 1968), suggesting that the accurate information produced by the use of computers promotes control by upper management because it makes information more readily available to high levels of authority. Other researchers (Markus & Robey, 1988) argued that the computers ability to take over routine low level decision-making enables lower and middle level managers of organizations to handle higher level decisions, resulting in decentralization. More recently, researchers have put forth a more complex explanation of the relationship between information technology and organizational structure, suggesting that organizational design is an emergent process, with the organizational form being the result of a complex interaction between the technological and social environments (Lucas & Baroudi, 1994; Robey & Boudreau, 1999). We submit that understanding the impact of information technology on organizational structure requires a ner framework and analysis of the structural characteristics of the focal organization and information technology. 2.2. Dimensions of organizational structure Organizational structure is more complicated than distinguishing between centralized and decentralized operations or decision-making. Other commonly cited structural dimensions include specialization, standardization, formalization, hierarchical levels, and span of control (Donaldson, 2001). Given the number of proposed structural dimensions and the variety of their denitions, identifying a denitive set of organizational dimensions is difcult without specic context and objectives (Blackburn, 1982). For the purpose of our study, we argue that the dimensions developed in the structural contingency theory literature are appropriate starting points. Within the structural contingency literature, the main organizational contingencies are specialization, formalization, structural

ARTICLE IN PRESS
N.A. Morton, Q. Hu / International Journal of Information Management 28 (2008) 391402 393

differentiation, and decentralization (Donaldson, 2001). In this study, the dimensions of formalization and specialization are merged into a single dimension of formalization, as is used in the technology-structure literature (Fry, 1982). This is because formalization and specialization are usually highly correlated, thus a distinction between them is not always made (Donaldson, 2001). In this paper, formalization is dened as the standardization of work processes and documentation (Donaldson, 2001), structural differentiation is dened as the difference in goal orientation and in the formality of the structure of the organizational units (Lawrence & Lorsch, 1967, p. 10), and decentralization is dened as the extent to which power over decision-making in the organization is dispersed among its members (Mintzberg, 1980, p. 326). In order to examine the t between organizational structures and ERP systems, it is useful to group the dimensions of structure into the common types generally found in organizational structure literature. Mintzberg (1979) conducted a synthesis of research on organization design from which he developed ve structural congurations that can be treated as a typology of ideal or pure organizational types. These ve basic congurations provide a useful framework of organizational structures for the purposes of this research. Although Mintzburg (1979) suggested four possible ways of using these prototypes, we focus on the second applicationtreating the set of congurations as ideal typesas the starting point for our discussions. The ideal types and their pertinent characteristics are presented in Table 1. 2.3. Structural contingency theory Contingency theory posits that organizational effectiveness is achieved by matching organizational characteristics to contingencies. Contingency is dened as any variable that moderates the effect of an organizational characteristic on organizational performance (Donaldson, 2001, p. 7). The structural contingency theory of organizations argues that the performance of an organization is dependent upon the t between organizational structure and contingencies. According to Donaldson (2001), three main elements form the core paradigm of structural contingency theory: (1) there is an association between contingency and the organizational structure; (2) contingency impacts the organizational structure; and (3) there is a t of some level of the structural variable to each level of the contingency, where high t leads to effectiveness and low t leads to ineffectiveness. It is important to note, however, that effectiveness in contingency theory has a wide-ranging meaning that includes efciency, protability, and worker satisfaction (Donaldson, 2001). A number of potential contingencies have been identied in the literature (e.g., technology, innovation, environmental change, size, and diversication). Donaldson (2001) argues that size, environment, and technology are the underlying contingencies in the structural contingency

Table 1 Ideal types of organizational structures (Mintzberg, 1979) Ideal type Simple form Salient characteristics

               

Small, simple Low formalization Highly centralized Unsophisticated technical systems Perform routine operating tasks Highly formalized Relatively centralized decision-making Automated and integrated technology Highly differentiated structure Standardized work processes used for coordination Operate in stable environments Regulating, non-automated technical system Decentralized decision-making Standardization of skills used for coordination Highly skilled workers who value autonomy Nonregulating, nonsophisticated technical system

Machine bureaucracy

Professional bureaucracy

Divisionalized form

 Centralized headquarters  Semiautonomous, loosely joined divisions  Little interdependence or close coordination
among divisions

 Main goal of headquarters is to coordinate goals of   

Adhocracy divisions with that of its own without sacricing autonomy Standardized outputs of divisions used for coordination Divisions are generally machine bureaucracies Technical system separated into segments, one for each division Operates as a cohesive group working together Mutual coordination and cooperation Innovative Workers are trained experts from different specialties Ad hoc project teams Low formalization Decentralized decision-making Operate in dynamic environments Sophisticated and often automated technical system (in the administrative adhocracy)

        

literature. While size is relatively straightforward, the ways researchers operationalized the environment and technology contingencies have been a source of problems (Pennings, 1975). Donaldson (2001) suggests that many contingencies, excluding size, can be divided into two aspects of organizational tasks; task uncertainty and task interdependence. Along with size, task uncertainty and task interdependence make up the underlying contingencies of the structural contingency literature. 3. Research framework and propositions Applying the structural contingency theory in the context of ERP implementation, we can see that ERP systems possess characteristics that relate to the task

ARTICLE IN PRESS
394 N.A. Morton, Q. Hu / International Journal of Information Management 28 (2008) 391402

uncertainty and task interdependence contingencies, as well as to the structural dimensions of formalization, structural differentiation, and decentralization. From the discussions of organizational types and structural contingency theory presented above, the core research framework of this study emerges. We posit that the t between organizational dimensions and characteristics of ERP systems has signicant implications for the success or failure of ERP implementation in organizations. In the context of ERP implementations, organizational structures that have low levels of business integration and relatively non-standardized work processes will encounter high resistance from within, as the ERP pushes the organizations to integrate functions and units and adopt the standardized business processes embedded in the system (Gattiker & Goodhue, 2004). An organization that already uses a cross-functional structure might still face resistance as it is forced to change its businesses processes in accordance with the ones embedded in the ERP modules to be implemented. However, such resistance could be much less signicant than the one in the previous case. Political conicts and organizational resistance to change have been identied as a critical success factor for ERP implementation success (Hong & Kim, 2002; Lee & Myers, 2004) and other IT projects (Backhouse, Hsu, & Silva, 2006; Markus, 1983). Successful implementation depends on emotions and politics as much as on rational decision-making (Markus & Robey, 1983). If the information system ts well with the organization, then the power structures, responsibilities, and job denitions will not change as signicantly (Leifer, 1988). The greater the change the system imposes, the greater the resistance; the greater the resistance, the less chance for implementation success (Markus, 1983; Markus & Robey, 1983). These ideas are presented in Fig. 1. 3.1. ERP characteristics, structural dimensions, and contingencies In this paper, for the benet of theory development and discussion, we make the strong assumption that ERP systems and their implementation in organizations are in their pure forms. In its pure form, we assume the full functionality of a single ERP package is implemented
Organization Structure

throughout an organization. In practice, ERP vendors usually offer a certain degree of customization to clients and not all modules of an ERP package need to be installed in a particular implementation. For instance, SAP R/3 systems contain 12 standard modules that are marketed as a single package, yet organizations may choose to install only a few of them (Benders et al., 2006). However, focusing on a full ERP implementation enables us to explicate the key issues and develop more generalizable propositions. Before we discuss the contingencies of ERP implementation success, a clear denition of success is in order. Information system success is a complex and multidimensional variable in IS research that has been the focus of much discussion and debate (DeLone & McLean, 2003). Accordingly, numerous frameworks of IS success have been presented in the extant literature. The DeLone and McLean (1992) IS success is one particular framework that has been referenced in numerous studies. This framework identies six different ways in which IS research has conceptualized IS success: information quality, system quality, information use, user satisfaction, individual impact, and organizational impact. In this study, we examine success in terms of the implementation project. Therefore, we dene ERP success from an implementation project perspective. Implementation project success is often described in terms of the achievement of some predetermined goals, such as time, cost, and function (Markus & Tanis, 2000). ERP implementations that do not meet project goals in terms of expected cost, time, and system performance are considered unsuccessful. The cross-functional and integrated nature of ERP systems enable unied view of rms and force organizations to move away from function-based system towards process-oriented integration (Al-Mashari, 2003). Such integration imposed by the implementation of ERP increases interdependencies among functional units. Scholars have long argued that information systems increase interdepartmental coordination (Saunders, 1981), and this phenomenon is relatively more pronounced for ERP systems due to the strong requirement for integration by the very nature of ERP systems. Interdependence is dened as a condition where actions taken within one unit affect the actions and work outcomes of another unit

Structure ERP Fit

ERP Implementation Success

ERP System

Fig. 1. Framework for organizational t and ERP implementation success.

ARTICLE IN PRESS
N.A. Morton, Q. Hu / International Journal of Information Management 28 (2008) 391402 395

(McCann & Ferry, 1979). Interdependence is clearly a result in even the most basic implementations of ERP systems. The cross-functional and integrated nature of ERP has signicant impacts on business processes. Standardization of data and business processes is necessary to enable ERP software to handle transactions and data storage across the various functions and units of an organization (Gattiker & Goodhue, 2004). It is well known that the implementation of an ERP system generally requires an organization to modify its internal processes to accommodate the standardized processes embedded in the ERP software (AlMudimigh, Zairi, & Al-Mashari, 2001; Benders et al., 2006; Hong & Kim, 2002). It should be noted that some organizations may prefer to keep their current business processes and customize ERP software to match their existing business processes, thus avoiding the need to make signicant changes in their work environment (Swan, Newell, & Robertson, 1999). However, this could be a costly and risky proposition. Researchers often recommend that organizations change business processes to match the standard business processes embedded in the ERP system in order to increase the likelihood of ERP implementation success (Benders et al., 2006; Davison, 2002; Hong & Kim, 2002). Both characteristics of ERP systems identied, standardization of business processes and integration of business functions, t the denition of a contingency as dened in the structural contingency literature. These two characteristics correspond to the key contingency aspects of organizational task as dened in Donaldson (2001): task uncertainty and task interdependence. Task uncertainty refers to the difference between information needed to complete a task and the amount of information already presents (Galbraith, 1977). By standardizing and integrating information ows in real time throughout the organization, ERP essentially reduces task uncertainty. The low task uncertainty associated with ERP is predicted to be a good t with organizational structures that have a high degree of formalization and low levels of decentralization (Donaldson, 2001). On the other hand, task interdependence refers to the degree that activities in an organization are connected with each other. ERP requires interdependence, close coordination, and cooperation among the functional departments and units of an organization in order to efciently and effectively perform tasks (Stefanou, 2001). Sharma and Yetton (2003) have noted that ERP systems are characterized by high levels of task interdependence. The high degree of task interdependence imposed by ERP is predicted by structural contingency theory to be a good t with organizational structures characterized by low formalization, low structural differentiation, and low levels of decentralization. The prediction that ERP will t relatively unformalized structures is contradictory with both the ERP literature and the task uncertainty contingency. However, task uncertainty is considered to be a stronger contingency

factor than is task interdependence (Donaldson, 2001). Furthermore, the standardized processes of ERP are by nature highly formalized since the standardization of processes necessitates the formalization of many organizational rules, policies, and procedures (Umble, Haft, & Umble, 2003). The argument that ERP is associated with low levels of decentralization is also supported by the literature. Sia, Tang, Soh, and Boh (2002) suggest that while ERP may enable both empowerment and centralization, it is the technology that perpetuates management power. Benders et al. (2006) also note the link between ERP implementation and increased centralization. One case in point, Dell Computer attempted to implement an ERP system but found that it would not t with its decentralized management model. Dell eventually decided to drop the project (Davenport, 1998). In summary, the literature and theory suggest that due to the task interdependence and task uncertainty contingencies, ERP systems will be a better t with organizations possessing the structural dimensions of high formalization, low structural differentiation, and low decentralization. Organizations that possess these structural dimensions would be a good t with ERP systems, should face relatively low resistance to change, and would have relatively high likelihood of implementation success. On the other hand, organizations that do not possess these structural dimensions could face relatively high resistance to change due to the poor t between the system and the organizational structure, and thus would have a relatively low likelihood of implementation success. These arguments are summarized in Table 2 and developed more fully in the next section. 3.2. Fit between ERP systems and organizational types In this section, we elaborate on the contingency t between ERP and organization types presented in Table 2. A set of propositions are developed regarding the likelihood of ERP implementation success in different types of organizations. To provide support for our propositions, we use case examples of ERP implementations reported in the business press and peer reviewed journals. Admittedly, these cases are ad hoc evidence at best, but they do serve as the starting points for more rigorous validation with larger sample sizes in the future. 3.2.1. Machine bureaucracy The organizational structure that most closely represents the dimensions of high levels of formalization, low levels of decentralization, and low levels of structural differentiation is the machine bureaucracy in Mintzburgs ideal forms. Machine bureaucracies are often found in mature, large, mass production companies residing in stable environments. The machine bureaucracy satises the criteria of formalization because it is already characterized by high level of formalization. The machine bureaucracy is

ARTICLE IN PRESS
396 N.A. Morton, Q. Hu / International Journal of Information Management 28 (2008) 391402 Table 2 A framework of contingency t between ERP and organizational types Organizational type Structural dimensions Formalization Structural differentiation Medium High Medium High High Medium Decentralization Degree of t and likelihood of ERP success

Machine bureaucracy Professional bureaucracy Professional bureaucracy support staff component Divisionalized form Adhocracy Administrative adhocracy operating component

High Low High Medium Low High

Low High Low High High Low

High Low High Low Low High

relatively centralized because it has limited horizontal decentralization, so it ts the decentralization contingency as well. The machine bureaucracy ts moderately well with structural differentiation. The organizational structure of a machine bureaucracy is only moderately differentiated because there is an emphasis on following formal programs and plans, and integration is affected mainly through these programs (Miller, 1986). Still, the functional responsibilities are only brought together at the highest levels of the organization (Mintzberg, 1980), so ERP will likely push towards lower levels of structural differentiation, causing moderate but not severe resistance to change. This leads to: Proposition 1. The organizational structure of machine bureaucracies has a higher degree of t with ERP characteristics and thus is likely to have a positive impact on ERP implementation success. A signicant amount of case evidence exists that supports this proposition. The ERP implementation at Motts North America Inc. is one of these cases. Motts is a large producer of juice and applesauce. The industry is mature and relatively stable. As indicated above, large production rms in mature, stable industries are typical machine bureaucracies (Mintzberg, 1979). The company implemented a US$10 million SAP R/3 project to cut IT support costs, reduce head count, and move to a single corporate system. Ofcials reported that the implementation was a success and had signicantly contributed to organizational efciency (Stedman, 1999). Guide Corp. is another exemplar of a successful ERP implementation in a machine bureaucracy. Guide Corporation is an automobile lighting systems manufacturer with manufacturing operations in Indiana, Louisiana, and Mexico. It is an old company in a mature industry, and has about 35,000 employees. The company supplies a limited number of products. These attributes suggest a machine bureaucratic structure. Guide implemented ERP software from QAD Inc. to handle everything from shipping to operations on the shop oor. The company says the system is a complete success (Fox, 2001).

3.2.2. Professional bureaucracy As a whole, professional bureaucracies do not t well with ERP systems. They have low levels of formalization and low levels of task interdependence. Moreover, professional bureaucracies have high levels of structural differentiation because professionals work relatively independently of their colleagues and are more or less free to set their own goals and to structure their work as they see t. Typical examples of such an organization include law rms, accounting rms, and universities. Imposing formalization on a professional bureaucracy would likely draw signicant resistance from professionals. In contrast to the utilization of formalized procedures, the standards established by professionals are formalized by their professions, not the organizations by which they are employed (Mintzberg, 1979). Attempts to increase the interdependence among professionals would meet with signicant resistance. Highly trained professionals demand considerable autonomy in their work. Professionals work relatively freely not only from other divisions, but also from their own colleagues (Mintzberg, 1979). Thus, professional bureaucracies are a poor t with ERP in all dimensions, and professional bureaucracies that attempt to implement ERP are likely to encounter political conicts and signicant resistance. This leads to our next proposition: Proposition 2. The organizational structure of a professional bureaucracy has a lower degree of t with ERP characteristics and thus is likely to have a negative impact on ERP implementation success. Typical cases of professional bureaucracy are found in the education, nance, law, and healthcare industries. An article describing the implementation of an ERP system at an Ivy League university (Wagner, 2003) supports this proposition. In 1999, the university adopted an ERP system from Oracle. Implementation of the system at this university was met with signicant resistance because it infringed on the autonomy of many faculty researchers. Resistance was also attributed to fact that the system shifted administrative efforts to academic departments.

ARTICLE IN PRESS
N.A. Morton, Q. Hu / International Journal of Information Management 28 (2008) 391402 397

The system managed to temporarily squash the valued academic freedom of many faculty researchers despite their power within the wider university community (Wagner, 2003). The faculty ultimately rejected the ERP system, resulting in the need for radical design modications to create a workable system. The creation of the workable system was costly in both monetary and human terms, leading the authors to question the appropriateness of ERP for a university or higher education organizations in general (Wagner, 2003). However, as indicated above, the support staff of the professional bureaucracy is characterized by structural dimensions distinct from the rest of the organization. In many professional bureaucracies, such as universities and large consulting rms, the support staffs tend to work in a machine bureaucratic structure side by side to its professional bureaucracy counterpart (Mintzberg, 1980). Since the support staff component of a professional bureaucracy is essentially a machine bureaucratic component within the professional bureaucracy, based on Proposition 1, the support staff component is likely to be a good t with ERP systems. Thus we suggest: Proposition 2a. The support staff component of professional bureaucracies has a higher degree of t with ERP characteristics and thus is likely to have a positive impact on ERP implementation success. The University of Basels ERP implementation is a case in point. The university implemented an ERP system from SAP to facilitate the processes of student administration, fee collection, applications distribution, and data exchange (SAP, 2003). Acceptance of the system was high, and is reported as a success. While this professional bureaucracys implementation was a success, none of the cited benets are directly concerned with the core operating work of the professionals. Rather, the benets are primarily related to administrative and support tasks performed primarily by the non-professional support staff. 3.2.3. Adhocracy Adhocracies have very little formalization, high structural differentiation, and high decentralization. There is very little formalization of behavior or standardization of work processes in adhocracies to encourage integration. Imposing formalization on an adhocracy would result in signicant resistance. Adhocracies have high levels of structural differentiation because people with vastly different goals and time horizons work together (Miller, 1986). Reducing this structural differentiation would likely draw signicant resistance from members of an adhocracy. Imposing centralization on an adhocracy would also draw signicant resistance because the power in an adhocracy is decentralized (Miller, 1986). Thus, we propose: Proposition 3. The organizational structure of adhocracies has a lower degree of t with ERP characteristics and thus is

likely to have a negative impact on ERP implementation success. This proposition is supported by the case of Agilent Technologys ERP implementation. Agilent is an innovative high tech company that was spun off by HP. It is a relatively young company operating in a complex, dynamic high tech environment. The above attributes are all indicative of an adhocracy. Agilent both develops and manufactures its high tech products to serve its customers, suggesting an administrative adhocracy in which the manufacturing work of the operating core is largely truncated from the more innovative component of the organization. Agilent implemented an ERP package from Oracle. The implementation was problematic and costly (Songini, 2002). The company says the ERP implementation resulted in shortfalls of US$105 million in revenue, US$70 million in operating prots, and huge additional programming costs. The company spent an unexpected US$35 million in programming costs related to the problems. Agilent says the problem was not with the quality of the Oracle software. Rather, the case highlights issues concerning people, processes, policies, and the companys culture as the culprits leading to implementation difculties. Such issues are indicative of a signicant mist between the system and the design of the organization. Another case is Microchip Technology Inc., a maker of embedded microcontrollers that installed a Baan IV ERP suite. It is a young, small technology company. Small technology companies are typically adhocracies (Mintzberg, 1979). The implementation of the ERP system revealed a mist between the organization and the system. As noted by Microchips manager of business operations, ythe functionality and business logic of the modules didnt always t the companys needs (Gill, 1999). While the company was able to implement the system, it relied heavily on a phased implementation approach and customization of modules due to the mist between the system and the organization. There are, however, administrative adhocracies in which the operating component is separated from the rest of the organization. This component, which usually has the structure of the machine bureaucratic form, is isolated so as to allow the administrative component to function freely as an adhocracy. Since the operating component of the administrative adhocracy is essentially a machine bureaucratic component within the administrative adhocracy, based on Proposition 1, the operating component of the administrative adhocracy is likely to be a good t with ERP systems. Thus, we propose: Proposition 3a. The organizational structure of the operating component of administrative adhocracies has a higher degree of t with ERP characteristics and thus is likely to have a positive impact on ERP implementation success.

ARTICLE IN PRESS
398 N.A. Morton, Q. Hu / International Journal of Information Management 28 (2008) 391402

The case of ERP implementation at Thermacore is an illustrative example for this proposition. Thermacore is a small company involved in the design and production of innovative thermal management solutions for electronics OEMs worldwide. It develops extremely complex, innovative products in a dynamic environment, indicating an adhocracy. Although the company is based in Lancaster, PA, its manufacturing plant is located in Mexico. The companys design and manufacturing functions are sharply distinct, suggesting an administrative adhocracy. The company implemented an ERP from SAP. According to the company, the system was implemented to get products to the right place at the right time and price. The systems implementation was a success, and the company was quick to realize payoffs (Engler, 1999). Thermacore did not let the project exceed its scope. It was careful to refrain from implementing the system in areas where implementation was not crucial. By focusing the system on getting products to customers accurately and efciently, the effects of the system on the more innovative tasks performed by the highly trained project groups are mitigated. Another example is the ERP implementation by Novartis Pharma K.K. Incorporated in 1997, Novartis Pharma K.K. is a subsidiary of Novartis Pharmaceuticals, the fth ranked company in the global pharmaceuticals market. The adhocracy form is often found in companies developing new pharmaceuticals, as they must rely on highly trained scientists to innovate in highly complex, dynamic environments. The drug development process requires highly trained experts to work together and to coordinate via mutual adjustment. The company develops its products to serve its clients, indicating an administrative adhocracy, in which there is a sharp distinction between the research activities and the day to day work of the operating core. The company successfully implemented SAP R/3 for nancial accounting, controlling, sales and distribution, material management, production planning, and warehouse management. The company notes that the ERP system did not impose on its drug development process. An ofcial from the company stated: Our current SAP solution does not affect our drug development process. (SAP, 2004) This essentially mitigates the t issues otherwise posited for ERP implementations in an adhocracy. 3.2.4. Divisionalized form The focus thus far has been on organizational structures at the business unit level. However, the divisionalized form is more of a superimposition of one structure on others rather than a complete structure unto itself (Mintzberg, 1980). For this reason, the focus for the analysis of the divisionalized form is the structure at the corporate level. The divisions within a divisionalized organization generally serve different markets and thus receive relatively little close coordination and a fair amount of autonomy (Mintzberg, 1980). However, despite granting a fair amount of autonomy to individual divisions, the head

ofce of a divisionalized organization seeks to formalize the behavior of divisions through standardized procedures and methods to improve control over divisions whenever possible (Miller, 1986). The result is a moderate level of formalization. Structural differentiation is high, and the divisional form is relatively decentralized. Increased formalization may be an attractive option to headquarters, but will likely draw erce resistance from divisional managers who control the divisions. Similarly, decreased structural differentiation and decreased decentralization will not be opposed by headquarters, but will be opposed signicantly by the divisional managers. In fact, headquarters is likely to view favorably the opportunity to increase centralization and reduce structural differentiation. A primary task of headquarters is coordinating the goals of the divisions with those of its own (Mintzberg, 1979). Moreover, an important problem for headquarters is getting accurate information so that it may make better decisions. Increasing centralization and reducing structural differentiation are attractive ways to accomplish this. However, the resistance by divisional managers combined with the relatively poor t of ERP with the structural dimensions of divisionalized organizations will make successful implementation difcult. Therefore, we suggest: Proposition 4. The organizational structure of the divisionalized form has a lower degree of t with ERP characteristics and thus is likely to have a negative impact on ERP implementation success. There are a number of cases that provide support for this argument. An ERP implementation at Westinghouse Electric Corp. is one case in point. Westinghouse Electric Corp. is a multidivisional electric and electronics conglomerate. In 1994, the company began the installation of an SAP R/3 system encompassing six of its divisions. Westinghouse was organized in a highly decentralized fashion, allowing each of its divisions to control its own nances and resources. Divisional managers severely resisted the system upon realizing it would infringe on their control of their respective divisions. Divisional managers haggled over system standards and system processes. When the ERP project team needed information, divisional managers were of little help and slow to respond. The result was a drawn out implementation exceeding the planned budget by about US$4 million (Schneider, 1999). Another case is an ERP implementation by an Atlantabased forest products manufacturer. The company adopted an ERP system from SAP to integrate each of its 12 divisions. Eleven months into the implementation, divisional VPs revolted against the system upon realizing it would infringe on their autonomy. The project was halted and the company had to write off US$120 million in software development costs (Schneider, 2000). However, there is an important caveat that must be made concerning the proposition for the divisionalized

ARTICLE IN PRESS
N.A. Morton, Q. Hu / International Journal of Information Management 28 (2008) 391402 399

form. This proposition suggests that organizations of the divisionalized form do not t well with ERP systems. It should be emphasized that this does not suggest that individual divisions will have difculty in implementing ERP. In fact, many of the large multidivisional rms have chosen approaches to ERP implementation that do not fully integrate the data and processes across the entire enterprise (Markus, Tanis, & Fenema, 2000). Perhaps realizing that the one size ts all approach to ERP is a poor t with their organizational structures, some rms have turned to less comprehensive ERP solutions. For example, Kraft Foods planned to install a single ERP system at each of its 53 manufacturing sites, but found that the operations of its eight product divisions were too diverse for the one size ts all ERP system. Kraft elected to implement a single ERP nancial module, but implemented different ERP congurations for each of its divisions to allow for diversity in business unit operations (Stedman, 1999a). When multidivisional rms implement ERP at the divisional level, the t and likelihood of implementation success are dependent on the characteristics of the divisions, which could be classied into any of the preceding three organizational types. 4. Discussion 4.1. Dynamics of organizational structure In the preceding discussions about the t between ERP and organizational types, for theoretical and methodological clarities, our treatment of organizational structure and ERP implementation is static and ideal in two primary aspects. First, in our application of Mintzbergs (1979) congurations, we have treated the congurations as a typology of ideal types, and thus the dynamic aspects concerning these congurations have been addressed only tangentially. However, dynamic forces are an important element of these congurations, particularly in the other applications of Mintzberg framework: (1) as a set of ve pulls, (2) as a set of ve pure types that reect the structures of many organizations, (3) as the basis for describing hybrid structures, and (4) as the basis for describing structural transitions (Mintzberg, 1979). In this study, we have focused on the second of the four suggested applications only for the purpose of theoretical clarity and simplicity. In reality, however, the dynamic aspects of organizational conguration could become the major complicating factors in ERP implementation success. While a comprehensive treatment of the dynamic aspects is beyond the scope of our study, some key points should be noted here. In particular, the structural congurations can be conceptualized as a set of ve forces that pull organizations in different structural directions: (1) the pull exerted by the strategic apex to centralize and to coordinate by direct supervision in order to retain control over decisionmaking; (2) the pull exerted by the techno-structure to

coordinate by standardization; (3) the pull exerted by the operators to professionalize; (4) the pull exerted by middle managers to Balkanize, or to divide the structure into market-based units; and (5) the pull exerted by the support staff (and by the operators as well in the operating adhocracy) for collaboration in decision-making. In theory, most organizations will be dominated by one of these pulls, and consequently will design structures close to one of the congurations. The fact that the types may remain ideals that are never fully realized in practice need not diminish their orienting role, as they provide a framework for human thinking about organizations. Thus, in this sense, the propositions based on the ideal types form a conceptual guide for management when considering implementing ERP systems in an organization, but not necessarily a prescriptive guide for predicting the nal outcome. 4.2. The duality of technology In this study, we have treated organizational structures primarily as social facts that possess an objective existence independent of ongoing social relationships. This stems from our guiding theoretical lens, structural contingency theory, which is primarily a static view (Barley, 1990). Alternative theories take more dynamic views in which organizational structure and human action are inextricably linked. One such theory, the theory of structuration (Giddens, 1984), has been a particularly inuential theoretical lens for viewing the link between social structures and human actions. The theory views social reality as constituted by both subjective human actors and by objective institutional properties. Giddens proposes the duality of structure concept, which sees the structure or institutional properties of social systems as created by human action, which subsequently serves to shape future human action. Structuration describes a continuous tension between conscious human action and institutional structures. Human action is seen as both forming and being formed by institutional properties. Drawing from the theory of structuration, Orlikowski (1992) describes what she refers to as the underlying duality of information technology. This duality conceptualizes information technology as both an antecedent and a consequence of organizational action. More specically, information technology has a constituted nature (it is a product of social interaction within a specic structural and cultural context), and also plays a constitutive role (it is simultaneously an objective set of rules and resources involved in mediating human action and hence contributing to the creation, recreation, and transformation of these contexts). In essence, information technology both shapes and is shaped by organizational action. The duality view of technology is especially silent in the context of ERP implementation. During the ERP implementation process, organizations not only can and often do customize the modules to t their business processes, but

ARTICLE IN PRESS
400 N.A. Morton, Q. Hu / International Journal of Information Management 28 (2008) 391402

also recongure and redesign some business processes to accommodate and take advantage of the best practices embedded in the software. In this sense, the propositions developed in this study based on a static view of the organizational structure and ERP software may not have strong predictive power. However, they indeed call the attention of managers and vendors to the potential pitfalls of mist between organizational structure and system characteristics and serve as warnings regarding what could happen if no appropriate customization or process redesigning or reengineering takes place. Furthermore, these propositions form a foundation for future research that takes into the consideration of the dynamic and interactive view of technology and organizational structure.

5. Conclusion In this paper, we have developed a set of propositions about the relationships between the characteristics of ERP systems and the dimensions of organizational structure based on structural contingency theory and Mintzbergs (1979) ideal structural types of organization. It is suggested that ERP systems are a good t with some organization types, but a poor t with others. Organizations whose structures are a better t with ERP systems are likely to have greater chances of successful implementations. Organizations whose structures are a poor t with ERP systems are likely to face organizational resistance to the systems and thus increase the chances of unsuccessful implementation. Notwithstanding the caveat of static and ideal views of organizational structure and ERP systems, this analysis provides a baseline for ERP planning and implementations in various types of organizations. It lls a gap in the research literature and contributes to the knowledge of ERP implementation in organizations in both management practice and academic research. From a practical perspective, this study directs managements attention to the importance of recognizing and addressing the issue of t between organizational structure and ERP package. Our ndings suggest that organizations implementing ERP must consider the t with their structures, the consequences of changing their business processes, and the potential resistance from within. Such organizations must recognize that the implementation of ERP systems is likely to induce conicts within their organizations and consequently impose difculties and even result in failure in the implementation process if an initial mist exists. Managers within an organization must be able to determine if a proposed ERP system is a good t with their organizational structure, or if it may only be a good t with certain parts of the organization, or if signicant customization and process redesigning or reengineering will be required. While we do not recommend specic practical prescriptions for managers, we believe our framework is relevant in that it is a novel view

of ERP-structure t that allows managers to enhance or restructure the mental models they use in practice. From a theoretical perspective, Robey et al. (2002) note the lack in the research literature of theoretical frameworks in the area of ERP implementations in general and the critical success factors in particular. This study partially lls this void as it draws from established theory to develop such a framework for one critical success factor, namely the t between organizational structure and ERP system. Furthermore, the proposed framework indirectly suggests that different critical success factors may be more or less important depending on the type of organizational structure in which the ERP system is implemented, an issue which has not attracted much deserved attention in the research literature. In general, our work serves as a basis for future empirical and theoretical work in this area. Some caution should be used when observing these ndings. The proposed relationships between organizational structures and ERP are not empirically tested, even though they are theoretically based and supported by anecdotal case evidences. Empirically testing the proposed relationships will be an important subject of future research, and should add signicantly to the validity of the propositions. As mentioned above, this paper studied ERP in its pure form, and compared it to Mintzbergs structural types in their ideal forms, despite evidence that such pure and ideal forms do not always exist in reality. With this caveat, the proposed relationships between ERP systems and organizational structures should provide guidance for future research, and provide insight as to which types of organizations are likely to benet from ERP, on what scale ERP should be implemented, and how much resistance organizations may expect from ERP implementation. References
Al-Mashari, M. (2003). A process change-oriented model for ERP application. International Journal of HumanComputer Interaction, 16(1), 3955. Al-Mudimigh, A., Zairi, M., & Al-Mashari, M. (2001). ERP software implementation: An integrative framework. European Journal of Information Systems, 10(4), 216226. Attewell, P., & Rule, J. (1984). Computing and organizations: What we know and what we dont know. Communications of the ACM, 27(12), 11841217. Backhouse, J., Hsu, C. W., & Silva, L. (2006). Circuits of power in creating de jure standards: Shaping an international information systems security standard. MIS Quarterly, 30(5), 413438. Barley, S. R. (1990). The alignment of technology and structure through roles and networks. Administrative Science Quarterly, 35(1), 61103. Benders, J., Batenburg, R., & van der Blonk, H. (2006). Sticking to standards: Technical and other isomorphic pressures in deploying ERP-systems. Information and Management, 43(2), 194203. Blackburn, R. S. (1982). Dimensions of structure: A review and reappraisal. The Academy of Management Review, 7(1), 5966. Davenport, T. H. (1998). Putting the enterprise into the enterprise system. Harvard Business Review, 76(4), 121131. Davenport, T. H. (2000). Mission critical: Realizing the promise of enterprise systems. Boston, MA: Harvard Business School Press.

ARTICLE IN PRESS
N.A. Morton, Q. Hu / International Journal of Information Management 28 (2008) 391402 Davison, R. (2002). Cultural complications of ERP. Communications of the ACM, 45(7), 109111. DeLone, W. H., & McLean, E. R. (1992). Information systems success: The quest for the dependent variable. Information Systems Research, 3(1), 6095. DeLone, W. H., & McLean, E. R. (2003). The DeLone and McLean model of information systems success: A ten-year update. Journal of Management Information Systems, 19(4), 930. DiMaggio, P. J., & Powell, W. W. (1983). The Iron cage revisited: Institutional isomorphism and collective rationality in organizational elds. American Sociological Review, 48(2), 147160. Donaldson, L. (2001). The contingency theory of organizations. Thousand Oaks, CA: Sage Publications. Engler, N. (1999). Push the comfort zone. InformationWeek, 731, 151. Fox, P. (2001). Auto supplier cuts IT costs by $12 million. Computerworld. Retrieved March 2006, from: /http://www.computerworld.com/ managementtopics/roi/story/0,10801,64597,00.htmlS. Fry, L. W. (1982). Technology-structure research: Three critical issues. The Academy of Management Journal, 25(3), 532552. Galbraith, J. R. (1977). Organizational design. Reading, MA: AddisonWesley. Gattiker, T. F., & Goodhue, D. L. (2004). Understanding the local-level costs and benets of ERP through organizational information processing theory. Information and Management, 41(4), 431443. Giddens, A. (1984). The constitution of society: Outline of the theory of structuration. Berkeley, CA: University of California Press. Gill, P. J. (1999). ERP: Keep it simple. Information Week Online. Retrieved March 20, 2007, from: /http://www.informationweek.com/ 747/47aderp.htmS. Gosain, S. (2004). Enterprise information systems as objects and carriers of institutional forces: The new iron cage? Journal of the Association of Information Systems, 5(4), 151182. Hitt, L. M., Wu, D. J., & Zhou, X. (2002). Investment in enterprise resource planning: Business impact and productivity measures. Journal of Management Information Systems, 19(1), 7198. Holland, C., & Light, B. (1999). A critical success factors model for ERP implementation. IEEE Software, 16(3), 3036. Hong, K.-K., & Kim, Y.-G. (2002). The critical success factors for ERP implementation: An organizational t perspective. Information and Management, 40(1), 2540. Lawrence, P. R., & Lorsch, J. W. (1967). Organization and environment: Managing differentiation and integration. Boston, MA: Division of Research Graduate School of Business Administration Harvard University. Leavitt, H., & Whistler, T. (1958). Management in the 1980s. Harvard Business Review, 36(6), 4148. Lee, J. C., & Myers, M. D. (2004). Dominant actors, political agendas, and strategic shifts over time: A critical ethnography of an enterprise systems implementation. The Journal of Strategic Information Systems, 13(4), 355374. Leifer (1988). Matching computer-based information systems with organizational structures. MIS Quarterly, 12, 6373. Lucas, H. C., Jr., & Baroudi, J. (1994). The role of information technology in organization design. Journal of Management Information Systems, 10(4), 923. Markus, M. L. (1983). Power, politics, and MIS implementation. Communications of the ACM, 26(6), 430444. Markus, M. L., & Robey, D. (1983). The organizational validity of management information systems. Human Relations, 36(3), 203226. Markus, M. L., & Robey, D. (1988). Information technology and organizational change: Causal structure in theory and research. Management Science, 34(5), 583598. Markus, M. L., & Tanis, C. (2000). The enterprise systems experience From adoption to success. In R. W. Zmud (Ed.), Framing the domains of IT Research: Glimpsing the future through the past (pp. 173207). Cincinnati, OH: Pinnaex Educational Resources Inc. Markus, M. L., Tanis, C., & Fenema, P. V. (2000). Multisite ERP implementations. Communications of the ACM, 43(4), 4246. 401 McCann, J. E., & Ferry, D. L. (1979). An approach for assessing and managing inter-unit interdependence. Academy of Management Review, 4(1), 113120. Miller, D. (1986). Congurations of strategy and structure: Towards a synthesis. Strategic Management Journal, 7(3), 233249. Mintzberg, H. (1979). The structuring of organizations: A synthesis of the research. Englewood Cliffs, NJ: Prentice-Hall. Mintzberg, H. (1980). Structure in 5s: A synthesis of the research on organization design. Management Science, 26(3), 322341. Murphy, K. E., & Simon, S. J. (2002). Intangible benets valuation in ERP projects. Information Systems Journal, 12(4), 301320. Orlikowski, W. J. (1992). The duality of technology: Rethinking the concept of technology in organizations. Organization Science, 3(3), 398427. Pennings (1975). The relevance of the structural-contingency model for organizational effectiveness. Administrative Science Quarterly, 20(3), 393410. Reif, W. E. (1968). Computer technology and management organization. Iowa City, IA: Bureau of Business and Economic Research, College of Business Administration, University of Iowa. Robey, D., & Boudreau, M.-C. (1999). Accounting for the contradictory organizational consequences of information technology: Theoretical directions and methodological implications. Information Systems Research, 10(2), 167185. Robey, D., Ross, J. W., & Boudreau, M.-C. (2002). Learning to implement enterprise systems: An exploratory study of the dialectics of change. Journal of Management Information Systems, 19(1), 1746. SAP. (2003). The University of Basel, Switzerland [electronic version]. SAP.com. Retrieved March 20, 2007, from: /http://www.sap.com/ industries/highered/pdf/CS_University_of_Basel.pdfS. SAP. (2004). Novartis Pharma K.K. [electronic version]. SAP.com. Retrieved March 20, 2007, from: /http://www.sap.com/usa/services/ consulting/pdf/CS_Novartis_Pharma.pdfS. Saunders, C. S. (1981). Management information systems, communications, and departmental power: An integrative model. The Academy of Management Review, 6(3), 431442. Scheer, A. W., & Habermann, F. (2000). Making ERP a success. Communications of the ACM, 43(4), 5761. Schneider, P. (1999). Human touch sorely needed in ERP [electronic version]. CNN.com. Retrieved March 20, 2007, from: /http:// www.cnn.com/TECH/computing/9903/02/erpeople.ent.idg/index.htmlS. Schneider, P. (2000). Another trip to hell. CIO, 13(9), 71. Shang, S., & Seddon, P. B. (2002). Assessing and managing the benets of enterprise systems: The business managers perspective. Information Systems Journal, 12(4), 271299. Sharma, R., & Yetton, P. (2003). The contingent effects of management support and task interdependence on successful information systems implementation. MIS Quarterly, 27(4), 533555. Sia, S. K., Tang, M., Soh, C., & Boh, W. F. (2002). Enterprise resource planning (ERP) systems as a technology of power: Empowerment or panoptic control? The Data Base for Advances in Information Systems, 33(1), 2337. Soh, C., Sia, S. K., & Boh, M. (2003). Misalignments in ERP implementation: A dialectic perspective. International Journal of HumanComputer Interaction, 16(1), 81100. Soh, C., Sia, S. K., & Tay-Yap, J. (2000). Cultural ts and mists: Is ERP a universal solution? Association for Computing Machinery. Communications of the ACM, 43(4), 4751. Songini, M. L. (2002). ERP effort sinks Agilent revenue. Computerworld, 36(35), 1. Stedman, C. (1999). Motts turns ERP up a notch. Computerworld, 33(14), 1. Stefanou, C. J. (2001). Organizational key success factors for implementing SCM/ERP systems to support decision making. Journal of Decision Systems, 10(1), 4964. Swan, J., Newell, S., & Robertson, M. (1999). The illusion of best practice in information systems for operations management. European Journal of Information Systems, 8(4), 284293.

ARTICLE IN PRESS
402 N.A. Morton, Q. Hu / International Journal of Information Management 28 (2008) 391402 organizational strategy and structure. He has published manuscripts in Journal of International Technology & Information Management and Structural Equation Modeling, and presented his research at Americas Conference on Information Systems, the Academy of Management Meeting, and the Decision Sciences Institute Meeting. Teo, H. H., Wei, K. K., & Benbasat, I. (2003). Predicting intention to adopt interorganizational linkages: An institutional perspective. MIS Quarterly, 27(1), 1949. Umble, E. J., Haft, R. R., & Umble, M. M. (2003). Enterprise resource planning: Implementation procedures and critical success factors. European Journal of Operational Research, 146(2), 241257. Wagner, E. L. (2003). Narrating an organizational matter of fact: Negotiating with enterprise resource planning technology to achieve order within a traditional academic administration. Unpublished PhD dissertation, University of London.

Neil Andrew Morton is a Ph.D. candidate in the Department of Information Technology & Operations Management, Barry Kaye College of Business at Florida Atlantic University. He received an M.B.A. from the University of Montana. His research interests include the adoption and use of information technology by individuals and groups. He is particularly interested in the adoption and use of healthcare information systems that allow individuals a greater role in the healthcare process. His research also looks at the impact of information technology on

Qing Hu is Professor of Information Systems in the Department of Information Technology & Operations Management, Barry Kaye College of Business at Florida Atlantic University. He earned his Ph.D. in Computer Information Systems from the University of Miami, Florida, USA. His research interests include electronic commerce, economics of information technology (IT), information security, and IT management strategies, especially in the areas of the impact of information technology on organizational strategy, structure, and performance. His research work has been published in leading academic journals including MIS Quarterly, Information Systems Research, Journal of the AIS, Communications of the ACM, California Management Review, Journal of Management Information Systems, IEEE Transactions on Software Engineering, Journal of Strategic Information Systems, and other international academic publications.