Bostrom surveying instruments model 5

Business process redesign project success: the role of

socio-technical theory Junlian Xiang

Ted Rogers School of Management, Ryerson University, Toronto, Canada, and

Norm Archer and Brian Detlor DeGroote School of Business, McMaster University, Hamilton, Canada

Abstract

Purpose – The purpose of this paper is to seek to advance business process redesign (BPR) project research through the generation and testing of a new research model that utilizes formative constructs to model complex BPR project implementation issues. Instead of looking at management principles, the paper examines the activities of improving business processes from the project perspective. Design/methodology/approach – A survey of 145 managers and executives from medium and large-sized USA and Canadian companies was used to validate the model. Findings – The model, based on socio-technical theory, includes three implementation components (change management, process redesign, and information and communication technology infrastructure improvement), and links the effects of these components to BPR project outcomes. The empirical findings indicated that all three implementation components had a significant impact on BPR project success, with change management having the greatest effect. Interestingly, the results also showed that productivity improvement was no longer the main focus of companies carrying out BPR projects; instead, improvement in operational and organizational quality was more important. Research limitations/implications – The main limitation of this study is its generalizability with respect to company size and organizational culture. The sample in this study was drawn from medium- and large-sized companies in Canada and the USA, but small-sized organizations were excluded from this study due to their distinct features (e.g. superior flexibility or ability to reorient themselves quickly). Also, this study controlled the variable of organizational culture by limiting respondents to Canada and US companies. It would be very interesting to investigate BPR project implementations in other countries where the organizational working culture may be different. Practical implications – Based on the findings of this study, BPR practitioners can refer to the three BPR project implementation components and then prioritize and sequence the tasks in a BPR project to achieve their preset BPR goals. Originality/value – This is the first study which utilizes formative constructs to validate the important BPR project components.

Keywords Change management, Business process management, Business process redesign, Information and communication technology infrastructure, Socio-technical theory

Paper type Research paper

1. Introduction Business processes have drawn a great deal of attention from industrial practitioners and academic researchers since the 1990s because of their great potential for improving the efficiency and effectiveness of organizations. The roots of business process focussed research can be traced back to business process reengineering, promoted by Hammer and Davenport et al. (Hammer and Champy, 1993; Davenport, 1993), that described it

The current issue and full text archive of this journal is available at www.emeraldinsight.com/1463-7154.htm

Received 26 October 2012 Revised 22 January 2013

5 April 2013 28 July 2013

19 September 2013 Accepted 21 September 2013

Business Process Management Journal

Vol. 20 No. 5, 2014 pp. 773-792

r Emerald Group Publishing Limited 1463-7154

DOI 10.1108/BPMJ-10-2012-0112

This research was supported by a grant from the Social Sciences and Humanities Research Council of Canada.

773

Business process redesign project

success

as a radical redesign of business processes resulting in a singular transformation. However, researchers soon realized that better results were obtained when organizations started with a revolutionary design phase, followed by actually implementing changes in an evolutionary manner ( Jarvenpaa and Stoddard, 1998). Later on, a more general term business process management (BPM) was widely adopted by many researchers in the business process focussed research context. BPM is defined as an integrated management philosophy and set of practices that includes incremental change and radical change in business processes, and emphasizes continuous improvement, customer satisfaction, and employee involvement (Hung, 2006; Houy et al., 2010). BPM covers concepts such as total quality management, business process reengineering, business process redesign (BPR), business process improvement, etc.

Usually BPM emphasizes different phases along its life cycle. A complete review of various BPM life cycles has been provided by Houy et al. (2010). The current study examines a specific group of business process focussed projects, i.e. BPR. Instead of looking at management principles, we examine the activities involved in improving business processes from the project implementation perspective. Therefore, the term BPR is used in this study. It is defined as a deliberate (planned) change, typically enabled by information technologies (IT) in an attempt to redesign and implement business processes to achieve performance breakthroughs in quality, speed, customer service, cost, etc. (Grover and Jeong, 1995).

Although extensive research has been carried out in the BPR area, a recent article has shown a continuing increase in interest in this field (Houy et al., 2010). One of the reasons for this interest is probably that the failure rate of BPR projects still remains high (Žabjek and Štemberger, 2009). Furthermore, the increased interest in this field implies that there are limitations in existing research results. For example, studies that do exist tend to be case study reports, making it difficult to generalize findings that yield consistent results (Bradley, 2008). Many previous studies have carefully examined the readiness of an organization to embark on a BPR project (Fenelon, 2002; Abdolvand et al., 2008; Aghdasi et al., 2010). Some have surveyed the best practices of BPR project implementation (Mansar and Reijers, 2007). A few others have started to measure outcomes related to BPR and IT portfolios (Ramirez et al., 2010; Ozcelik, 2010). However, as indicated by a recent literature review, “[y] most approaches concentrate on what needs to be done before and after the improvement act, but the act of improving itself still seems to be a black box” (Zellner, 2011, p. 1). Our observations are similar to the results of Zellner’s study; i.e. few studies have focussed on measuring and modeling what organizations should do during project implementation. Therefore, there is a need for more rigorous research in this area, specifically with regard to BPR project implementation, to determine and better understand how BPR project success rates can be improved. Lastly, few published studies have been able to address the complexities of the BPR problem in one research model. One reason is that it is almost impossible to use traditional reflective constructs to model such a complex problem. The appropriate technique to use in these situations is the application of formative constructs, but this has been neglected in the past ( Jarvis et al., 2003; Diamantopoulos et al., 2008; Petter et al., 2007).

The goal of this investigation is to model important BPR project implementation components, as well as to test the impact of these components on BPR project success. Various facets of BPR project outcomes were used in order to examine how BPR project implementation components would bring performance improvements.

774

BPMJ 20,5

To achieve this goal, the researchers began with a review of the BPR literature on project success factors, and then identified one important BPR enabler and three implementation components, using socio-technical theory (STT) as a guide. These formed the basis of the study’s research model. The paper continues with the development of the theoretical model and hypotheses. After carefully examining the causal relationships between the BPR project implementation components and their measures, formative constructs were judged to be the appropriate way to model those components. The research method involved validating the model, through a survey administered to 145 managers and executives from medium and large US and Canadian companies; responses are analyzed using structural equation modeling techniques, involving partial least squares (PLS). Next, the research results are presented and analyzed. Lastly, the implications of the research are discussed.

2. Theoretical model and hypotheses 2.1 STT and BPR project implementation components From the early 1990s to the mid 2000s, few theories were used to inform or guide BPR project research. Rather, most of the literature reported during this period merely listed BPR project success factors identified through case study research (Grover et al., 1998). However, since the mid 2000s, BPR researchers have started to borrow theories from other areas of research and apply these to the BPR context (Newell et al., 2000; Sarker and Lee, 2002; Sarker et al., 2006). As a result, initial frameworks (Grover and Jeong, 1995), models (Ifinedo and Nahar, 2009; Law and Ngai, 2007), and constructs (Guha and Grover, 1997) have been developed. Of these, STT appears to be promising.

STT views the organization as a work system with two interrelated subsystems: the technical system and the social system (Bostrom and Heinen, 1977). The technical system is concerned with the processes, tasks, and technology needed to transform inputs such as raw materials to outputs such as products; the social system is concerned with the relationships among people and their attitudes, skills, and values. The outputs of a work system are a result of the joint interaction between these two subsystems.

A critical important enabler for BPR project success, regarding the social aspect of STT, is top management support (TMS). Top management plays an important role in BPR projects as suggested by STT (Markus, 1983). “This type (supportive) of leadership offers a vision of what could be and gives a sense of purpose and meaning to those who would share that vision. It builds commitment, enthusiasm, and excitement. It creates a hope in the future and a belief that the world is knowable, understandable, and manageable. The collective energy that transforming leadership generates, empowers those who participate in the process [y]” (Roberts, 1984, as cited in Lashway, 2006, p. 91). As claimed in project management research, “top management support is the most important critical success factor for project success and is not simply one of many factors” (Young and Jordan, 2008, p. 1). Numerous surveys and case studies have shown that TMS is one of the most highly ranked success factors in BPR project practice (Fui-Hoon Nah et al., 2003; Grover and Jeong, 1995; Herzog et al., 2009). Grover and Jeong (1995) identified the lack of TMS as a serious problem impeding the success of business process reengineering implementations. Strong support from top management is also necessary to resolve any conflicts of interest among the various parties involved (Ahadi, 2004). Therefore, this study hypothesizes that TMS would positively impact BPR project success. However, this study takes a further step to explore what are the important components in the BPR project

775

Business process redesign project

success

implementation and to evaluate how TMS affects BPR project success through the project implementation components. This brings us to our first hypothesis:

H1. Stronger TMS will result in a better BPR project implementation procedure.

When exploring BPR project implementation components, this study considers both the technical and the social dimensions, and their interactions. As Grover and Jeong (1995) point out, STT emphasizes the impact of the changes to technical and human resources on altered tasks or processes. As such, technical resources, human resources, and altered tasks/processes are reflected in this study as three implementation components: information and communication technology infrastructure (ICTI) improvement, change management, and process redesign.

ICTI is the technology foundation dealing with the IT capabilities on which processes and humans rely, such as networks, databases, data inter-exchange, etc. for transforming inputs to outputs (Law and Ngai, 2007). The capabilities of ICTI consist of a wide spectrum of components, including ICTI platforms, standards, policies, and different types of service arrangements.

Change management includes the methods through which attitudes, skills, and values of the people in the system are managed and transformed (Huq et al., 2006). Change management is the soft part of the change process since it addresses human resource problems within organizations, such as employee resistance and structural adjustments. It requires effectively balancing forces in favor of change over forces of resistance from organizations, groups, and individuals (Guha and Grover, 1997; Markus, 1983).

Process redesign can be thought of as the interaction of the social and technical aspects of STT. For example, people (such as employees, customers, etc.) work to get expected outputs by following company procedures that implement specific business processes, supported mainly by IT.

When applying STT to BPR project implementations, it is important to understand that, whether companies have fixed process goals (e.g. BPR projects that follow external rules enforced for ERP systems implementations) or they design processes that fit their own requirements, new processes cannot work well unless people issues are resolved and IT support is suitable. In this way, the effects of ICTI improvement and change management are reflected through the redesigned processes. Business processes can never work without both operators (people) and carriers (technical implementation) (Davenport, 1993).

Given the above discussion, our first hypothesis can be decomposed into three sub- hypotheses. That is, TMS is hypothesized to positively affect the three implementation components. Hence:

H1 (a, b, c). Stronger TMS will result in better levels of change management, BPR, and ICTI improvement, respectively.

2.2 BPR Project outcomes and success BPR project success is defined as the advantageous outcomes that a BPR project achieves for an organization. It is improper to use a single financial criterion (e.g. cost reduction itself) to evaluate BPR project outcomes. Grover used two different perspectives: perceived level of success and goal fulfillment, to evaluate BPR success (Grover and Jeong, 1995). The perceived level of success seeks to assess the degree of

776

BPMJ 20,5

attainment in relation to the targets, and the goal fulfillment perspective determines success by attainment of a normative state (Hamilton and Chervany, 1981).

The second perspective, goal fulfillment, is based on the commonly emphasized goals of BPR projects. Four categories of outcomes were adopted in our study (Raymond et al., 1998): first, operational quality improvement, or improved quality in goods and services (e.g. customer service and satisfaction); second, organizational quality improvement, or improved quality of organizational coordination and communication (e.g. lessened managerial hierarchy, improved task enrichment, and reduced bureaucracy); third, cost savings, involving administrative and production cost savings (in terms of return on investment, personnel costs, operational costs, and profits); and fourth, productivity improvements that result from increased productivity of workers and managers (more units produced, fewer delays).

In order to examine the effect of the three implementation components, their impact was tested on each of the four categories of outcomes, hypothesized as follows:

H2 (a, b, c, d). A higher level of change management will result in a higher level of operational quality improvement, organizational quality improvement, cost savings, and productivity, respectively.

H3 (a, b, c, d). A higher level of BPR will result in a higher level of operational quality improvement, organizational quality improvement, cost savings, and productivity, respectively.

H4 (a, b, c, d). A higher level of ICTI improvement will result in a higher level of operational quality, organizational quality, cost savings, and productivity, respectively.

The researchers were also interested in how these categories of outcomes would contribute to the perceived level of project success. Different BPR projects may target different fulfillment goals. It is important to understand what company goals are focussed on through BPR projects because different goals may trigger different emphases in their BPR project implementation components. Thus:

H5 (a, b, c, d). A higher level of operational quality improvement (organizational quality improvement, cost savings, or productivity, respectively) from a BPR project will result in a higher level of perceived BPR project success.

The research model built from this study is summarized in the path model in Figure 1. The model examines how TMS affects BPR project implementation components, which in this study includes change management, process redesign, and ICT infrastructure improvement; the model further examines how these BPR project implementation components would affect BPR project outcomes (operational quality improvement,

H2 H3 H4

H5 H1

BPR Project Outcomes - Operational quality improvement - Organizational quality improvement - Cost savings - Productivity

BPR Project Implementation Components

- Change management - Process redesign - ICTI improvement

Top Management

Support

BPR Project

Success Figure 1. Research model

777

Business process redesign project

success

organizational quality improvement, cost savings, and productivity) and ultimately BPR project success.

3. Research methods and data analysis 3.1 Construct operationalization Three constructs for modeling BPR project implementation were created: change management, process redesign, and ICTI improvement. Each was modeled as a second-order construct and each of their first-order dimensions was operationalized as a latent variable. Researchers (e.g. Jarvis et al., 2003; Diamantopoulos et al., 2008; Petter et al., 2007) advocate examining constructs carefully before classifying them into reflective or formative categories. Jarvis et al. (2003) and Petter et al. (2007) proposed criteria and decision rules on how to determine if a construct should be modeled as formative or reflective. All three of these constructs were identified as formative second-order and reflective first order constructs, based on their recommended approach.

The construct of change management aims to assess the extent to which a BPR project utilizes change management practices. Three measures, as suggested in the literature, were developed to assess this construct: first, change management at the organizational level, to assess the impact of change management on organizational culture change and human resources system change (Al-Mashari and Zairi, 1999); second, change management at the employee level, to assess the impact of change management such as empowerment, communication and training on employee resistance (Grover and Jeong, 1995; Al-Mashari and Zairi, 1999); and third, change management at the stakeholder level, to assess the impact of change management on stakeholder resistance and commitment (Oakland and Tanner, 2007; Paper and Chang, 2005). The items for measuring the three dimensions were adapted from previous studies (Ramirez et al., 2010; Al-Mashari and Zairi, 1999; Grover and Jeong, 1995; Oakland and Tanner, 2007; Paper and Chang, 2005), and their dimensions were used as indicators to create the super-ordinate construct. The change management scale is shown in Table I.

The process redesign construct assesses the extent to which process redesign practices were used in a BPR project. This study adopted the two dimensions of process redesign (Attaran, 2003): technical redesign and social redesign. Technical redesign (PR_T) is used to redesign the allocation of the process workload; social redesign (PR_S) is used to redesign the allocation of personnel workloads. The items for measuring the two dimensions were adapted from Mansar and Reijers’ studies (Mansar and Reijers, 2005). The process redesign scale is shown in Table II.

Dimensions Item ID Measurement

Organizational level CM_OL1 Reward/motivation and compensation systems CM_OL2 Human resource policies CM_OL3 Organization’s receptivity to change CM_OL4 An effective culture for organizational change

Employee level CM_EL1 Communicate the reasons for change to employees CM_EL2 Empower relevant employees for change CM_EL3 Provide adequate training for relevant employees

Stakeholder level CM_SL1 Communicate vision to stakeholders CM_SL2 Solicit feedback from stakeholders

Table I. Change management construct and measurements

778

BPMJ 20,5

In this study, ICTI improvement assessed the extent to which a company’s ICTI capabilities have been improved through a BPR project. The four dimensions of ICTI improvement (network communications, data integration, training, and, facilities and management) and their corresponding items as proposed by Law and Ngai (2007) were adopted to create the super-ordinate construct. The ICTI improvement scale is shown in Table III.

TMS, a well developed construct in the IS literature, was operationalized as the extent to which top management provided support and commitment to a BPR project. Five items were chosen and minor modifications were made to fit the context of BPR projects (see Appendix 1), including top management’s understanding and support of the BPR project, as well as top management’s funding and communication support (Grover and Jeong, 1995).

BPR project success was measured from two perspectives. The first perspective was an overall perception of success (Grover and Jeong, 1995). A two-item scale was used to assess this perspective of success regarding to the achievements from a BPR project (see Appendix 1). The second was a goal-specific perception (Altinkemer, 2011; Raymond et al., 1998). The four facets of BPR project outcomes adopted from (Raymond et al., 1998) were used to measure goal-specific success in this study because it is relevant to this study. These included operational quality improvement (OpQI), organizational quality improvement (OrQI), cost savings (CS), and productivity (PROD).

Dimensions Item ID Measurement

Technical redesign PR_T1 Eliminating unnecessary tasks PR_T2 Combining or dividing tasks PR_T3 Re-sequencing tasks in processes PR_T4 Paralleling tasks PR_T5 Integrating business processes

Social redesign PR_S1 Empowering workers with more decision-making authority PR_S2 Assigning workers to perform as many steps as possible for

single orders PR_S3 Making human resources more specialized or more generalized PR_S4 Minimizing the number of departments, groups, and persons

Table II. Process redesign construct

and measurements

Dimensions Item ID Measurement

Network communications ICTII_NC1 Improving networks linked with suppliers ICTII_NC2 Improving networks linked with customers

Data integration ICTII_DI1 Improving data sharing across the company ICTII_DI2 Reducing/eliminating data duplication ICTII_DI3 Improving standardization of data element definitions

Training ICTII_TR1 Improving IT training programs ICTII_TR2 Improving user training ICTII_TR3 Improving IT personnel training

IT facilities and management ICTII_FM1 Increasing capacity of servers ICTII_FM2 Reducing regular preventive maintenance down time ICTII_FM3 Increasing expertise to manage IT facilities ICTII_FM4 Increasing satisfaction with IT services ICTII_FM5 Improving IT administration standards and procedures

Table III. ICTI improvement

construct and measurements

779

Business process redesign project

success

The measurements for TMS and BPR project success, as well as for the three BPR project implementation components, are listed in the questionnaire that was used (see Appendix 1).

3.2 Content validity test before data collection Before data collection, two steps were taken to ensure content validity. The first step was a Q-Sorting test on the three formative constructs. The goal of Q-sorting was to verify the dimensions or categories of the items that were drawn from the literature. Hence, a one-round Q-Sort was sufficient for this purpose. Five participants knowledgeable in the IS area, but with no prior knowledge of this study, were asked to examine a series of descriptive items that would be used for each of the constructs and to place each of them into one of several categories composed of the formative constructs. The measures and constructs theoretically identified by the researchers fully matched the results of the Q-sort, after clarifying a misunderstanding of the only unmatched category (social-based process redesign).

The second step used an expert panel. The instrument was examined by one academic researcher and two industrial experts for its content validity and was improved according to their opinions.

3.3 Data collection methods The data collection process was approved by the Research Ethics Board of a major university. Considering that small-sized companies may not have well-defined business processes or their business processes may not heavily rely on IT or ICTI solutions, data collection was limited to medium and large companies (each with 100 or more employees) in the USA and Canada that had undertaken a BPR project in the past three years. All respondents had participated in at least one such BPR project.

The researchers used an internet panel of potential respondents, chosen from the existing database of a commercial survey agency. In total, 21 percent of the targeted business professionals (294 out of 1,481) qualified under the requirements indicated above. Among the 294 qualified business professionals, 155 completed the survey, yielding an overall response rate of 52 percent. Ten of the 155 cases were dropped because participants failed to complete a number of the scales. For missing data in the remaining 145 cases, Little’s MCAR test (Little and Rubin, 2002) showed that these values were missing completely at random (Little’s MCAR test: w2¼7.358, Sig.¼ 0.393). The missing data were replaced by mean values.

Table IV provides demographics of industry distributions and respondent department distributions.

3.4 Analysis method SmartPLS 2.0 (Ringle et al., 2005) was used to analyze the data, as it allows latent constructs to be modeled either as formative or reflective indicators as was the case with this study’s data, and it is appropriate for exploratory studies (Chin, 1998). Because the model contains three second-order variables, the researchers created super-ordinate second-order constructs using factor scores for the first-order construct (Chin et al., 2003).

PLS works well with smaller sample sizes. The most frequently used rule for minimum sample size in PLS was proposed by Chin (1998, p. 311): “one simply has to look at the arrow scheme and find the largest of two possibilities: (a) the block with the largest number of formative indicators (i.e. largest measurement equation); or (b) the

780

BPMJ 20,5

dependent LV with the largest number of independent LVs impacting it (i.e. largest structural equation) [y] the sample size requirement would be 10 times either (a) or (b), whichever is the greater.”

The research model in this study involves three formative constructs: change management, process redesign, and ICTI improvement, which have 3, 2, and 4 formative indicators, respectively. As for the structural equation, the dependent LV with the largest number of independent LVs impacting it is overall success, which has four paths leading into it. Therefore, the minimum sample size requirement for this study is 10 � 4, or 40.

Another rule that should be considered for sample size is that there is a need for doing a principal components factor analysis on the indicators for all the constructs for an exploratory study such as this. Everitt (1975) recommended that the proper case-to-indicator ratio range for PCA should be at least 10. The largest construct considered, i.e. ICTI improvement, has 13 indicators, although they are divided into four dimensions. Therefore, according to this rule, the minimum sample size requirement is 13 � 10, or 130. The number of valid cases collected was145, meeting this criterion.