A realist evaluation of post-implementation modifications: a context-initiative-mechanism-outcome perspective

ABSTRACT

This paper exemplifies a theoretical and methodological evaluation of IT Initiatives. It discusses post-implementation modifications (ERP-PIM) to enterprise resource planning (ERP) systems for the purpose of business process optimization (BPO) from a Context-Initiative-Mechanism-Outcome (CIMO) perspective. CIMO suggests that context and mechanisms are factors triggering causal effects responsible for outcomes derived from technology initiatives. Through conceptualization, data collection/analysis, and retroduction, the paper proposes a middle-range theory to facilitate the explanation of outcomes from diverse post-implementation initiatives. Data from multiple case study identifies situational and action-formation mechanisms, and actions as essential to realizing the desired outcomes. The proposed ERP-PIM-CIMO theory encapsulates: (i) Business Process Optimization (BPO) as context, (ii) ERP-PIM as structure, (iii) Business Process Efficiency, Effectiveness, and Flexibility as observable events, (iv) ERP capability as causal mechanism, (v) actions as necessary for the actualization of mechanism, and finally (vi) effort and cost, knowledge, and training as factors influencing actualization.

KEYWORDS:

• Critical realist evaluation
• CIMO
• business process optimization
• ERP post-implementation
• ERP capability
• mechanism

1. Introduction

In a realist evaluation perspective, this paper investigates Business Process Optimization (BPO) as the context in which post-implementation modifications are made to Enterprise Resource Planning (ERP) systems with the aim to uncover mechanism(s) that facilitate outcomes such as business process efficiency, effectiveness, and flexibility.

To avoid becoming stagnant, organizations are constantly innovating, and developing new internal capabilities through digitalization. In other words, there is a strong focus on transformation through the adoption of new business models and digital technologies. Organizations are advised to adopt continuously refine, adapt, revise, and reformulate their business model by deploying innovation strategies (Caputo et al. 2020). Research however shows that in several cases, business models and companies have crumbled as a result of disruptive innovation (Sewpersadh 2023).

Whilst organizations are undergoing transformation through the adoption of new business models and digital technologies, their existing enterprise technologies remain important (Ivanović and Marić 2021). New technologies and business models are critical to transformation. In the same vein, business optimization through the rethinking and integration of business processes remains an important component for business transformation. Asprion, Schneider, and Grimberg (2018) suggest that questions addressing the role of current enterprise technologies must be answered and clearly included in a digital transformation roadmap. For instance, will the current ERP system be consolidated or a new one implemented? Will this be a part of the proposed digital transformation initiative? In paving way for successful digital business transformation, organizations are encouraged to implement new enterprise technologies such as cloud, artificial intelligence, and robotic process automation whilst adapting existing enterprise technologies to the changing business landscape. As an organization’s ERP system embodies the organization’s business processes, it plays a major role in the organization’s innovation and transformation endeavors. Accordingly, post-implementation modifications to these implemented ERP systems are important for ongoing business process optimization.

The question of how post-implementation modifications will deliver business process optimization as a part of a digital transformation initiative is however not yet answered in the literature. Allen et al. (2013) suggest that one can only understand why information systems succeed, for whom, and in what context by focusing on the interplay of stakeholder actions and context. Such focus facilitates the generation of middle-range theories where context is taken into consideration and mechanisms (causal factors triggering observable outcomes) are identified (Henfridsson and Bygstad 2013). With a focus on context, initiative, mechanism, and outcome, the paper therefore explores an explanation for mechanisms that trigger business process efficiency, effectiveness, and flexibility when organizations conduct ERP post-implementation modifications within the context of business process optimization. A focus on the context of ERP implementations is increasing, and a recent study of ERP success factors has highlighted the value of considering contextual dimensions independently (Vargas and Comuzzi 2020). Proposing a middle-range theory (ERP-PIM-CIMO), this paper emphasizes the relevance of intermediate levels of explanation by integrating knowledge of ERP-PIM into a coherent explanation of outcomes through the identification and validation of context and mechanisms. Specifically, the theory includes explication of situational and action-formation mechanism for explaining causality and identifies ‘Actualization’ (Strong et al. 2014) as essential to the realization of expected outcomes.

The paper begins with a discussion of Critical Realism (CR) in Information Systems research. CR has been adopted as the methodological lens of the paper; first the object of investigation is presented, including the context, and expected outcomes, followed by an explanation of steps undertaken in the development and testing of the proposed ERP-PIM-CIMO theory. The paper concludes in Retroduction, a revisiting of established theory.

2. Methodological perspective

2.1 Critical realism

While CR has been around as a research philosophy since the 1980s (Bhaskar et al. 1998), it is a relatively new philosophy for IS research where it has begun to see adoption as an alternative to traditional positivist and post-modern approaches actors (Carlsson, 2003; Dobson, 2002). CR identifies three domains of stratification: ‘real’ – where social structures, natural objects, material artefacts and conceptual entities exist independently of human perception; ‘actual’ – where events and outcomes may or may not be observed; and ‘empirical’ – a subset of the ‘actual’ that is observable or experienced. The ‘real’ domain is associated with causal mechanisms responsible for generating events (also known as outcomes) which are observable in the ‘actual’. With a focus on causality and the identification of causal mechanisms, CR is unlike the traditional positivist search for causal generalizations and the traditional interpretivist’s incapacity for describing causality independent of social actors (Carlsson, 2003; Dobson, 2002).

2.2 Critical realism in information systems

Information Systems research benefits from many attributes of CR.

First, CR employs a retroductive methodology which facilitates the understanding of interactions among many components of systems such as hardware, software, processes, and people, through the identification of mechanisms. Dobson’s (2002) argument for CR in IS research focuses on the utilization of technology as requiring deep investigation that favors explanation over description because of the real nature of technology and the outcomes associated with the utilization of technology. In other words, through deep questioning embedded in the methodology of retroduction, hypothetical mechanisms are proposed as a part of investigating technology, which, if in existence, would impact the outcome of technology implementations. An interesting study that discusses the use of retroduction is that of McAvoy and Butler (2018), where differences between theory-based expected mechanisms and observed mechanisms across several studies are highlighted. The study emphasizes the role of retroduction as facilitating identification of areas where misalignments exist between expectations and observations, thereby urging researcher(s) to further examine the phenomenon of interest.

Second, CR emphasizes the role of context in providing plausible explanation. Wikgren (2005), for instance, suggests that studying human information actions in context requires a distinguishing between human action and socio-cultural structure, and would benefit from a CR perspective. This suggestion is based on the idea that the properties possessed by the social and cultural norms that guide the use of information may differ from those possessed by the individuals who act as consumers and users of information.

Third, the generation of middle-range theories as plausible explanation of phenomena as facilitated by CR is one of the many benefits that IS researchers draw upon when adopting the philosophy. Lyytinen and Newman (2008), for example, make argument for a punctuated socio-technical change model as a middle-range theory for explaining information systems change. In the proposition of their theory, the authors consider socio-technical systems as acting in the real domain of mechanisms that are expected to generate events in the empirical domain, and observable by actors in the epistemic domain.

2.3 Critical realism for theory building and testing in IS research

Many CR articles present a philosophical discussion. Nevertheless, CR is now gaining momentum in IS investigations for both theory building and testing through empirical research, thereby building several contributions into the IS literature. For instance, in building a theory-based explanation of organization change mediated by technology utilization, Allen et al. (2013) adopt CR in addition to Activity theory, concluding that one can only understand why IS succeeds, for whom it succeeds, and the context in which it succeeds by focusing on the interplay of actions of stakeholders and context. Similarly, Strong and Volkoff (2010) from their study theorizing IT Associated Organizational Change Processes identify the concept of affordance as plausible explanations of causality, a key point of CR. With great emphasis on theory building, Bygstad, Munkvold, and Volkoff (2016); Strong et al. (2014) build on the concept of affordances and present a discussion on the concept of Actualisation – ‘actions taken by actors as they take advantage of one or more potentials through their use of the technology to achieve immediate outcomes in support of organisational goals’. In consideration of IS users as reflexive agents which are involved in various types of internal conversations with, or about, ICT, De Vaujany (2008), exemplifies how internal conversation theory may be helpful in integrating a social reflexivity into IS sociological frameworks.

Carlsson (2012) discusses two studies that exemplify CR-based theory building in IS; a study investigating organization-ERP fit by Volkoff, Strong, and Elmes (2007) and another by Morton (2006) seeking to explain causes of the outcomes proceeding from attempts to develop and implement strategic IS plans in organizations.

2.4 Realist evaluation

Carlsson (2012) also presents the idea of the ‘Realist Evaluation’ of information systems, where investigators seek to understand why an IS initiative works, for whom, and in what circumstances. Realist evaluation follow a realistic effectiveness cycle, starting off with theory, in this case propositions on how outcomes are generated from IS inventions introduced into pre-existing contexts. In conducting empirical CR-based IS studies, researchers focus on the identification of mechanisms through data collection and analysis. Realist evaluation include the generation of hypotheses proceeding from theoretical analysis of mechanisms, contexts, and expected outcomes, and can be undertaken using both qualitative and quantitative data collection approaches. An example of realist evaluation incorporating qualitative data is the work of Dobson, Myles, and Jackson (2007) which investigated an automated performance management system by developing a framework based on literature and conducting a qualitative case study as a way of testing the framework. Horrocks (2009); Oroviogoicoechea, Elliott, and Watson (2008) on the other hand is an example of a realist evaluation utilizing a purely quantitative approach. Through statistical analysis of open and closed ended questionnaires, the researchers investigate the perception of nurses regarding the impact of a hospital information system. Against popular opinion that CR cannot be applied as a philosophy for quantitative methodologies such as statistical analysis, Mingers (2004a, 2004b) specifically makes argument for statistical analysis to be used in the exploratory stage of CR-based IS research for the detection of patterns within data. Indeed, the point of CR is to discover underlying structures that generate patterns of events.

Many IS studies applying CR have focused on the investigation of mechanisms following two different approaches; (i)Archer’s Morphogentic approach (Archer et al. 2013; Archer 2020), which aims to set out the conditions under which change in an information system’s given state or structure is understood and (ii) the Context-Initiative-Mechanism-Outcome (CIMO) approach, where Information Technology initiatives and outcomes are viewed from the perspective of contexts and mechanisms.

Utilizing a morphogenetic approach through an exploratory case methodology, Horrocks (2009) investigates the interrelationship between the structure(s), culture(s) and agents of a specific IS. The morphogenetic approach is helpful for identifying possibilities for change in a social setting.

The Context-Initiative-Mechanism-Outcome (CIMO) approach is an application of CR known as a realist evaluation, specifically designed for evaluating IT initiatives. In this approach, the explanation of intended and unintended outcomes observed within any phenomenon is succinctly captured in a heuristic context + mechanism = outcome (Wong 2017). The CMO heuristic (more widely referred to as CIMO with I representing an ‘Intervention’ or ‘Initiative’) identifies a problem in its context (C), introduces an intervention/initiative (I), and analyses the generative mechanisms (M) through which the initiative produces verified outcomes (O) (Costa, Soares, and de Sousa 2018). The CIMO configuration facilitates the development of a middle-range theory through theoretical understanding of how the working of a particular mechanism leads to the generation of outcomes. In other words, links between the context and outcome of an IS initiative are established ‘through’ mechanisms as the focus shifts from the question of what works to ‘what is it about the system that works for whom in what circumstances?’ (Dobson, 2002).

Whilst a useful philosophy for IS research, the main challenge of conducting CR-based IS studies is the ‘how’. To address this, some authors offer specific guidelines to assist researchers. For instance, Wynn and Williams (2012) outline principles to guide the use of case studies in CR-based IS research.

2.5 CIMO-informed research design

Incorporating methodological guidelines offered by Wynn and Williams (2012) to support CR research, the paper follows a three-step design (Figure 1).

Figure 1. Three-step methodological design.

The three steps cover the explication of events, structure and context, empirical corroboration of hypothesized mechanisms, adoption of triangulation, and Retroduction of mechanisms.

3. STEP 1: ERP-PIM-CIMO theory conceptualization

The role of an organization’s ERP systems has continued to evolve, supporting change and business transformation. With a continuous desire for competitive advantage, business transformation is becoming a key focus alongside continued utilization and expansion of existing technologies for example, achieving a single instance of a global ERP solution (Stam and Theijssen 2014). While ERP systems continue to be critical organizational resources, the most organizations can obtain from ERP systems would be operational efficiency which at most provides a temporary competitive advantage (Ng and Chang 2009). For long-term competitive advantage however, post-implementation modifications that concern further alignments with business objectives, innovative enhancements and functional upgrades are recommended. Hence, any sustained competitive advantage to be derived from an ERP system will proceed from interdependent development and use by organizations (Kalling 2003). Post-implementation modifications therefore become important as they reveal how organizations develop and build the capability of their ERP systems. Such understanding is supported by the resource-based view of the firm, suggesting that organizations compete against one another based on their resources, specifically assets and capabilities (Mata, Fuerst, and Barney 1995). The resource-based view offers guidance on how ERP can bring about differential value to firms (Radhakrishnan, Zu, and Grover 2008), and serves as a perspective from which the effect of IT systems on business processes can be observed (Mooney, Gurbaxani, and Kraemer 1996). In this view, given the variety of post-implementation modifications that can be undertaken, the impact of post-implementation modifications on business processes can be expected to vary. For instance, some may provide new or improved automation and integration of business functions to support business process efficiency while others may simply permit innovation and differentiation in business processes (Karimi, Somers, and Bhattacherjee 2007a,b; Uwizeyemungu and Raymond 2012).

To explain the optimization of business processes through ERP post-implementation modifications, a middle-range theory is proposed, combining the CIMO methodological perspective and the Resource-based view. The use of multiple perspectives, while challenging in terms of theoretical convergence, provides an opportunity to achieve conceptual fit. The theory includes (i) ERP-PIM-CIMO configuration, where the relevant contexts, initiatives, mechanisms, and outcomes are identified, and (ii) ERP-PIM-CIMO propositions that suggest how mechanisms trigger the generation of outcomes from modifications.

3.1 ERP-PIM-CIMO configuration

Informed by Carlsson’s CIMO configuration, we propose the ERP-PIM-CIMO configuration.

The configuration (Figure 2b) offers explanation for business process optimization in terms of efficiency, effectiveness and flexibility gained from ERP post-implementation modifications through an identification of the relevant context, initiative, mechanism, and outcomes. Context – utilization of ERP systems within organizations for business process optimization. Initiative – structures of interest, in this case ERP post-implementation modifications – maintenance, upgrades, enhancement packs, and further alignments of the ERP system with organizational goals. Outcome – observed events, in this case business processes being optimized so they become more efficient, more effective, and (or) more flexible. In other words, efficiency, effectiveness, and flexibility are indicators of business processes optimized through ERP-PIM initiatives. Mechanism – ERP capability is identified as the causal mechanism that triggers BPO; automational, informational and transformational.

Figure 2. (a) CIMO Configuration (Carlsson, 2012)/ (b) ERP-PIM-CIMO configuration.

In summary, the ERP-PIM-CIMO configuration suggests business process optimization through the utilization of ERP systems when post-implementation modifications are undertaken that enhance ERP capability. Each component is explicated:

3.1.1 Context – business process optimization

Business process optimization is defined as fundamental revision of processes to reach spectacular improvements in measurements of efficiency, such as costs, quality, service and quickness (Hammer and Champy 1993). Supported by the resource-based view, business process optimisation permits process-oriented assessment of the business value of ERP post-implementation modifications. Given organizations’ continued need for improvements in business processes, BPO becomes an important consideration for the initiation of post-implementation modifications. Yet not all modifications are initiated with a focus on business processes optimization. The fact however remains that significant value can only be achieved from ERP implementations only as organisations continue to optimise business and management processes (Harris and Davenport 2006). Thus, business process optimization forms the context in which this paper investigates the impact of ERP post-implementation modifications on business processes.

3.1.2 Initiative – ERP-PIM

Optimization of business processes is a major driver for initial implementation of ERP systems, systems which become key organizational resources. As organizations strive to achieve this, they continue to undertake post-implementation modifications once the implementation of the system has been achieved (Harris and Davenport 2006; Karimi, Somers, and Bhattacherjee 2007b). Post-implementation modifications take different forms, for example minor upgrades, major upgrades, technical/maintenance upgrades, and functional upgrades (Ng and Gable 2010), and are driven by different organizational needs such as compliance or the desire for new functionality (Ng 2001; Scheckenbach et al. 2014). These modifications form the structures of interest for this paper, and include initiatives such as Maintenance, Upgrades, Enhancement packs, and further modifications/configurations to the ERP system.

3.1.3 Outcome – efficiency, effectiveness, flexibility

The resource-based view suggests that the effect of organisational resources such as IT investments can be observed at an operational level (Mooney, Gurbaxani, and Kraemer 1996). Business processes provide an avenue for such operational-level observations, being defined as a sequence of activities for the creation of goods and services by the conversion of input to output (Dutta and Roy 2004). Therefore, the use of IT systems in executing business processes is expected to result in business process optimisation and improved overall firm performance (Mueller et al. 2010). It is therefore no new finding that ERP systems add value to organizations by improving decision quality, empowering employees, and decreasing resource utilization, thereby enabling business process efficiency, effectiveness, and flexibility (Karimi, Somers, and Bhattacherjee 2007a). It is also not surprising to find that organizations that can commit resources to modifications to enhance business processes are more likely to optimize business processes, improving business process efficiency, effectiveness, and flexibility (Harris and Davenport 2006). Business process efficiency, the conversion of input to output in the shortest possible time with the lowest utilization of resources, is achieved when operational costs are reduced. Business process effectiveness, the satisfaction of one or more business objectives while meeting or exceeding the recipient stakeholder’s needs, is achieved through better and timely access to corporate data and higher levels of enterprise-wide data integration. Business process flexibility, the ability to adjust quickly and easily to changes in internal constraints or stakeholder requirements, is gained with new ways of customizing processes and responding quickly to business changes (Karimi, Somers, and Bhattacherjee 2007a; Trinh, Molla, and Peszynski 2012; Trischler 1996).

In this paper, these three (business process efficiency, effectiveness, and flexibility) form the outcomes that are expected when ERP post-implementation modifications are initiated within the context of business process optimization.

3.1.4 Mechanism

The main goal of this paper is to provide an explanation of business process efficiency, effectiveness, and flexibility as outcomes of ERP post-implementation modifications when initiated within the context of business process optimization. This explanation is provided through the identification and testing of mechanisms. Since the introduction of mechanisms, there have been preliminary attempts to group mechanisms into common categories. Hedström and Swedberg (1998) classified mechanisms into three interrelated types: situational, action-formation and transformational. Situational mechanisms describe how specific social situations or events influence the beliefs, desires, and opportunities of individual actors; action-formation mechanisms describe how individual choices and actions are influenced by specific combinations of opportunities; and transformational mechanisms describe where collective individuals’ actions and interactions generate both intended and unintended outcomes. The literature also documents discussions around organizational structure and technical structure related mechanisms, affordances as a subset of mechanisms, and specifically related to this paper, causal mechanisms explaining empirical observations about the uses of new IT in organizations (Wynn Jr et al. 2013).

3.1.5 Identification of mechanisms – a resource-based view

To explain business process efficiency, effectiveness, and flexibility from ERP post-implementation initiatives, we again draw on our overarching theoretical perspective, the resource-based view. As organizations utilize their resources (ERP systems in this case), they develop capabilities (Ng 2006) which then forms a basis of their ability to perform competitively with other organizations (Mata, Fuerst, and Barney 1995). This paper emphasizes the notion of capability as rooted in processes and business routines.

3.1.6 ERP capability as mechanism

Though capabilities are important if ERP systems are to be exploited for further organizational benefits (Beard and Sumner 2004; Kalling 2003; Ram, Wu, and Tagg 2013), very little attention has been given to the role of the capabilities, particularly an organization’s ERP capability. Nonetheless, benefits from ERP systems reflect what organizations can achieve through routines embedded in the ERP system for business tasks and functions. These routines build ERP capability and in this paper are explored for explaining the observable outcomes derived when post-implementation modifications are initiated within the context of business process optimization.

Not a commodity to be implemented but rather something that must be built, ERP capability is viewed from the perspective of the resource-based view to cover (i) reach, functional scope, geographic scope (Karimi, Somers, and Bhattacherjee 2007b), and (ii) integration, flexibility, and transversality (Uwizeyemungu and Raymond 2012). While different, these two perspectives of capabilities impact on operational and managerial processes, thus enabling business value from ERP use through automational, informational and transformational effects. ERP capability as discussed above refer to the realization of value from ERP systems through automational, informational and transformation impacts on business processes; impacts which result in three separate but related outcomes. Drawing on the above, we suggest a new view of ERP capability that facilitates its use as a possible explanation for business process efficiency, effectiveness, and flexibility. First, automational impacts concern the underlying automation of business processes enabled by the ERP system; suggested as Automational ERP capability – the ability of an organization’s ERP system to automate and integrate its existing business processes, thus creating value by substituting capital asset for labor and reducing cost, leading to business process efficiency. Second, informational effects concern underlying availability and processing of data brought about by the ERP system; suggested as Informational ERP capability – the ability of an organization’s ERP system to collect, store, process and disseminate information, leading to business process effectiveness. Third, transformational effect concerns the innovation and differentiation of business processes afforded by the ERP system; suggested as Transformational ERP capability – the ability of the ERP system to facilitate and support business process innovation, differentiation, and transformation, leading to business process flexibility (Mooney, Gurbaxani, and Kraemer 1996).

3.2 ERP-PIM-CIMO propositions

The ERP-PIM-CIMO configuration and propositions are derived from a consideration of what ERP-PIM must look like to produce BPO and suggests that post-implementation modifications do not on their own lead to BPO but rather activate a mechanism that is capable of triggering BPO. The propositions are testable descriptions of mechanisms that require triggering for business process optimization to be generated when modifications are undertaken and are specifically necessary for validating the explanation. The theory includes three sets of propositions (Table 1), each covering an element of structure (ERP-PIM), and focusing on the presence of mechanisms (automational, informational, and transformational ERP capability) as explanation for observable events (efficiency, effectiveness, flexibility).

Table 1. ERP-PIM proposition.

Proposition		Proposition Statement
P1		Business Process Efficiency
P1	A	ERP post-implementation modifications that enhance automational ERP capability improve process efficiency.
P1	B	ERP post-implementation modifications that do not enhance automational ERP capability do not improve process efficiency.
Note on P1		Due to differences in modification types, it may be that not all modifications are able to enhance automational ERP capability that helps increase business process efficiency. For instance, Maintenance is not expected to have a direct impact on efficiency.
P2		Business process effectiveness
P2	a	ERP post-implementation modifications that enhance informational ERP capability improve process effectiveness
P2	b	ERP post-implementation modifications that do not enhance informational ERP capability do not improve process effectiveness.
Note on P2		Not all modifications are likely to improve informational ERP capability. For instance, although maintenance modifications may eradicate bugs, they are incapable of delivering technologies/functionalities that enhance ERP-based process effectiveness.
P3		Business process flexibility
P3	a	ERP post-implementation modifications that enhance transformational ERP capability improve process flexibility.
P3	b	ERP post-implementation modifications that do not enhance transformational ERP capability do not improve process flexibility.
Note on P3		Modifications that do not include sensing and responding capabilities that help organizations to quickly and effectively adapt to changes may not provide any substantial improvement in transformative capability and as such are likely to be incapable of improving business process flexibility.

For a valid explanation of the causal mechanisms at work when an ERP-PIM does (or does not) result in business process optimization, the above three sets of propositions suggest the circumstances under which business process efficiency, effectiveness, and flexibility can be expected from modification initiatives.

Step 2 of the methodological design corroborates our conceptualization with empirical data.

4. STEP 2: embeded multiple case study

A realist multiple case study was conducted for empirical corroboration of the ERP-PIM-CIMO model, allowing the examination of ERP-PIM initiatives in a theory-testing pattern. Data obtained from case study interviews revealed the workings of mechanisms which were then compared to those proposed in the ERP-PIM-CIMO model. Two large organizations with multiple years of ERP utilization were purposefully selected. Multiple conversations with key IT and business personnel revealed intents and outcomes of modification initiatives. Data was collected using a protocol specifying indicators of business process efficiency, effectiveness, and transformation. Interviews were analyzed using a pattern-matching logic (Yin, 2009), where constructs identified from the data were matched to constructs defined in the ERP-PIM-CIMO Model. With several modifications identified in each case organization, multiple case study was analyzed as an embedded case, with each modification initiative as the unit of analysis and as such acting as a case within a case.

4.1 Case organizations: Bexplo and Bchem

To examine post-implementation modifications, we captured different contexts in which modifications were initiated by evaluating business process efficiency, effectiveness, and flexibility resulting from several modifications across two organizations: Bexplo and Bchem. Both organizations are large providers of energy; Bexplo, manufactures commercial explosives and blasting systems to the mining and infrastructure markets, and Bchem manufactures industrial chemicals and chlor-alkali products. Bexplo implemented SAP R/2 in 1995, and Bchem, SAP ERP (R/3 Version 3.0D) in 1997. Following initial ERP implementations, both organisations have undertaken several post-implementation modifications as narrated in Figure 3.

Figure 3. Longitudinal view of ERP-PIMs at Bexplo and Bchem.

4.2 Multiple case study analysis

As expected, given the timelines from the first implementation of the ERP system, both Bexplo and Bchem conducted a variety of modifications, some as initiatives for business process optimization and others solely as routine activities. The ERP-PIM-CIMO propositions were tested for each ERP-PIM initiative, with the aim to prove or disprove ERP capability as causal mechanism for facilitating business process efficiency, effectiveness, and flexibility, and in so doing, offer a good enough explanation.

4.2.1 ERP-PIMs as routine activities

Data indicated that certain ERP-PIMs such as Support Stack (I2) conducted at Bexplo was implemented to fix bugs and maintain an up-to-date system as confirmed by the IT Director:

‘our implementation of support stacks are not benefit-oriented. we basically do them as they get released to keep updated’ (IT director)

The data reveals that generally, support stacks have no impact on business process efficiency, effectiveness, or flexibility. At Bchem, OSS notes (I2), Support Stack (I5), and GUI upgrade (I6) were a means of bringing the system to the latest required patch level. Further supported by the IT Executive, these initiatives had no business impact:

‘We have to implement OSS notes … or upgrade to the new GUI version before any upgrade of business functions … the GUI upgrade by itself has no business value’ (IT Executive)

Also true for Support stack (I5), undertaken regularly to fix bugs. Not implemented in response to a specific business need, I5 was technically driven without inputs from business units. As a result, the support stack initiative would not impact on business operations; although it enabled the ERP system to run more quickly.

‘ … the support stack we have implemented is more on system performance so business users might notice that the system runs a bit quicker … other than that, it should not have any impact on the users’ work’ (IT Executive)

The above ERP-PIM initiatives support Propositions P1a, P2a, and P3a, that not all PIMs will optimize business processes. It is observed that Business process efficiency, effectiveness, and flexibility were not gained as the above-discussed initiatives fixed technical faults and updated the ERP system, without any effect on time or cost savings, decision quality, resource utilization, and process agility.

4.2.2 ERP-PIMs for business process optimization

Data uncovers a few instances where ERP-PIMs were conducted within the context of business process optimization. Ongoing extension of plant maintenance (I1) at Bexplo, was beneficial for automating the process of reordering spare parts. Stated by the Senior Business Analyst, extending Plant Maintenance (PM) facilitated information visibility and consistency across the entire organization:

We've been gradually extending SAP plant maintenance into different manufacturing plants to provide a consistency of approach … it requires in the order of thousands, pieces of master data to be collected and organized and put into the system’ … The benefit there is quantifiable in terms of safety performance and in reliability of the equipment. (Senior Business Analyst)

Facilitating better integration and automation of tasks, ongoing extension of PM improved visibility of information as Bexplo now had access to specific knowledge of maintenance costs. This alludes to the enhancement of informational ERP capability (i.e. the ability to collect, store, process and disseminate information), and Bexplo achieved timely access to corporate data, as well as higher levels of enterprise-wide data integration, both indicators of business process effectiveness. However, although automational capability has also been enhanced as the extension of PM allowed for further integration of tasks, the data did not indicate that business process efficiency or flexibility were also achieved. Automating the process of reordering spares implies that the process would require less time and human effort, thus implying efficiency and supporting P1a. Similarly, P2a, which suggests that PIMs that improve informational ERP capability will bring about business process effectiveness, is supported. Proposition P3b is likewise supported, with no evidence of transformational ERP capability nor business process flexibility. In summary, Propositions P1a, P2a and P3b are supported.

Sales and pricing configuration (I3) at Bexplo was initiated to address problems associated with calculating product pricing, and resulted in cost savings and increase in information accuracy. This is reflected by the Senior Business Analyst:

‘There became a requirement to develop a system that would be much more frequent and automated. SAP did not have and still does not have the capability to do this kind of calculations automatically … We’ve had some cost savings and its really improved accuracy of information on which decisions can be then taken … so decision support is really important’ (Senior Business Analyst)

Business process efficiency was observed in the form of cost savings, and business process effectiveness in the form of improved information accuracy for better decision-making, whilst business process flexibility was not observed. Business process efficiency and effectiveness as outcomes of configuring sales and pricing were enabled by both the automational and informational ERP capability delivered by the initiative; automational ERP capability was observed as the PIM activated automation of sales and pricing, and informational ERP capability as previously un-integrated components had now become integrated. The configuration of sales and pricing (I3) at Bexplo supports P1a and P2a as automational and informational ERP capability were enhanced, consequently delivering business process efficiency and effectiveness. However, I3 did not enhance transformational ERP capability and delivered no business process flexibility, thus supporting Proposition P3b. In summary, P1a, P2a, and P3b are supported.

Direct customer interfacing configuration (I4) at Bexplo was undertaken to facilitate more effective management of customer contracts. The SAP Global Support Lead explained that the result of undertaking this configuration was obvious in Bexplo’s ability to better consolidate and report on customers’ requirements:

‘At the end of the period when the customer wants a daily or weekly report, the report is available and the system can be triggered to put out a pdf report that gives the information to the customer. So we are now able to harmonize customer needs and can create reports through the system rather than it being done on excel spreadsheet’ (SAP Global Support Lead).

Explained above, the configuration was designed to enhance the capability of Bexplo’s ERP system to make information more available and accessible to customers:

‘We had to create new tables and new information within the system because the kind of information that the customer wanted was not totally available in SAP at the time’ (SAP Global Support Lead)

The direct customer interfacing PIM at Bexplo offers support for Proposition P2a as informational ERP capability was enhanced, leading to business process effectiveness. However, without evidence of automational ERP capability for business process efficiency nor transformational ERP capability for business process flexibility, I4 at Bexplo supports Propositions P1b and P3b.

Report configuration, (I1) at Bchem, undertaken because the ERP system did not meet business requirements in terms of interfacing with customers, actually improved reporting. According to the Senior Business Analyst, it was necessary to automate payment transactions to facilitate better payment reports:

‘Our developments improve the reporting requirement … we've done a lot of developments with our banking partners as we don’t do in-house cheque processing … we use SAP standard transactions to process the payment and then we create a file and just one file layout handles all payments whether it is a local or foreign currency. It is all an automatic process now’.(Senior Business Analyst)

The above reflects that the automational ERP capability was improved by the configuration, as the system had been modified to automate the processing of payments. However, while time savings may be inherent because the process of payment had been automated and therefore previously manual tasks had been eliminated, there was no specific mention of cost or time savings. Hence, Initiative I1 at Case C provides only implicit evidence to suggest that the company improved the efficiency of the payment process by undertaking the initiative. However, this initiative clearly supports P2b, as no informational ERP capability was enhanced and as a result no business process effectiveness was delivered. Proposition P3b is also supported, as there was no evidence of transformational ERP capability or business process flexibility.

Language upgrade, (I2) at Bchem allowed Chinese characters to be printed in the Asian region, but did not improve business process efficiency, effectiveness or flexibility. This is indicated in the SAP Basis Administrator’s remark:

‘The language upgrade allowed us to replace texts that do not exist … ’ (SAP Basis Administrator)

No more than facilitating the use of languages other than English and German, which are standard in the ERP system, I2 had no impact on ERP capability, therefore supporting Propositions P1b and P2b, which suggest that Enhancement PIMs do not impact business process efficiency or effectiveness if ERP capability is not improved. In addition, with no impact on business process flexibility, I2 provides support for P3b.

Installation of EHP6 (I5) at Bexplo, though conducted with an aim to transform the organization by providing access to new IT capability did not result in BPO. The IT Director remarked:

‘We installed the components of the pack (EHP6) to have the new technical capability. To get the right level of information, you need to put in the right information and you need to transact into the system’ (IT Director)

In addition to installing EHP6, Bexplo still needed to enhance its informational ERP capability (i.e. its ability to collect, store, process and disseminate information) by feeding information into the ERP system. Only after that would it become possible for Bexplo to transform data into context-rich information and knowledge capable of supporting business analysis and decision-making. This can then be expected to lead to improvement in decision quality and resource utilization for enhancing the effectiveness of business processes. However, as informational ERP capability had not been enhanced, Bexplo gained no business process effectiveness from the initiative. Rather, having an enhanced technical platform that could possibly support enhanced business functionality if utilized correctly was the key outcome of I5 at Bexplo. The EHP6 installation at Bexplo therefore supports Proposition P2b, which suggests that only PIMs that enhance informational ERP capability may result in business process effectiveness.

Following the installation of EHP6, Bexplo sought new functionalities through the Activation of technical usages for Enhancement pack 4 (EHP4-I6). As this activation had just been initiated at the time of data collection, benefits had not yet been gained from the initiative. Notwithstanding, according to the Business Analyst, activating technical usages would enhance automational and informational ERP capability:

‘There has been some changes to business processes because we have gone through the process of inputting the right level of information … we can now see the true cost to the company when a product is sold … we have the information and now transact into the system … some activities that were done on the excel spreadsheet outside the system are now done in the SAP system so we have the information recorded and stored … we can report better now and also are much quicker is responding to our customers because we have clearer cost information’ (Senior Business Analyst)

By setting relevant information into the ERP system to facilitate the use of functionality provided in EHP4, informational ERP capability has been enhanced. Likewise, eliminating tasks which had previously been undertaken using spreadsheets and activating functionality to enable the tasks to be performed in the ERP system demonstrate the enhancement of automational ERP capability. With automated pricing came better reporting and visibility of inherent cost, and led to quicker customer response. This is indicative of enhanced transformational ERP capability and increase in business process flexibility. All three business process outcomes: efficiency, effectiveness, and flexibility are evident in this initiative, and can be seen to be a result of enhanced automational, informational, and transformational ERP capabilities. Thus, Propositions P1b, P2b, and P3b are all supported.

In the Enhancement pack 3 initiative (EHP3-I4) at Bchem (Case C), the SAP Basis Administrator reported that the implementation of EHP3 at Bchem extended the sales and distribution functionality in the ERP system, thus making business users happy. According to the basis administrator:

‘ … The set of functionality on the standard SAP is quite limited. With the enhancement pack there’s more the business users can do. The enhancement pack 3 was basically part of the business requirement … the business specifically decided to explore the enhanced sales and distribution functionality included in EHP3 to see if our sales and distribution process could be improved. Basically the business users are happy … there are certain things not provided in standard SAP … for instance the sales and distribution set of functionality on the standard SAP is quite limited. With EHP3, there’s more they can do’ (SAP Basis Administrator)

The fact that business users at Bchem are able to execute more business tasks because of the implementation of EHP3 indicates that the initiative possesses inherent capabilities that may or may not be exploited by users. In other words, while the capability may be available in the enhancement pack, users are required to specifically enhance the firm’s ERP capability. However, the satisfaction experienced by business users provides no indication of ERP capability or business process efficiency, effectiveness, or flexibility from the implementation of EHP3 at Bchem.

Whilst each of the seven ERP-PIMs undertaken for business transformation across the two case organizations support the ERP-PIM-CIMO propositions, the paper aims not just to test propositions but through the testing of propositions, uncover underlying mechanisms. Thus, to further develop the ERP-PIM-CIMO theory, we conduct a theoretical review of the findings from the evaluation of propositions through retroduction.

5. Retroduction

Retroduction is a method whereby hypothetical mechanisms are proposed as causal factors for explaining a phenomenon of interest (Mingers, Mutch, and Willcocks 2013; Zachariadis, Scott, and Barrett 2013). Starting off with the identification of a few potential mechanisms and selection of the one most plausible in the specific context (Tsang 2014), retroduction allows the discovery underlying mechanisms to explain observed regularities, and it involves working back from data to possible explanations (Blaikie, 2009). With the enhancement of ERP capability indicating the attainment of BPO from modification initiatives, the ERP-PIM-CIMO theory proposes ERP capability as mechanisms that facilitate BPO from ERP-PIM if initiated within the specific context of ERP-PIM as a tool for business transformation. The propositions appear valid as the data indicated that only initiatives that were conducted within the context of business transformation could enhance ERP capability which in turn permitted the realization of BPO. A great insight from the data is that although some of the initiatives could enhance capabilities, organizational actors such as managers and business analysts were required to undertake actions to actualize the capability provided by the modification. Working back from data point to: (i) actions as necessary to actualize the capabilities embedded in the modification initiatives, and (ii) contextual factors that influence the actualization of ERP capability. Based on these findings, we return to theory and revise our conceptualization of mechanisms to account not only for situational mechanisms but also for action-formation mechanisms.

5.1 Return to theory – actualization

The implicit assumption in the ERP literature is that the implementation of a particular technology is automatically accompanied by the actualization of the potential provided by the technology. Actualization is defined specifically as ‘actions taken by actors as they take advantage of one or more potentials through their use of the technology to achieve immediate outcomes in support of organizational goals’ (Strong et al. 2014). Actualization can be perceived as actions directed at making a potential real or giving a potential the appearance of reality and from a resource-based perspective, considered to be essential to the realization of potential value (Johnson, Melin, and Whittington 2003). In the IS literature, the term ‘actualization’ refers to the achievement of a goal, and is used interchangeably with such terms as instantiation (Bygstad 2010) and realization (Bygstad, Munkvold, and Volkoff 2016). For this reason, the conceptualization of automational, informational and transformational capabilities as mechanisms facilitating business process optimization when ERP-PIMs are initiated is consistent with the idea of benefit realization. Particularly, the ERP-PIM-CIMO theory’s suggestion of automational, informational and transformational capabilities is indicative of situational mechanisms, which describe how specific situation or events influence the desires, and opportunities of individual actors. From data however, we observed that actions were required to activate the potential of ERP-PIM initiatives. This alludes to action-formation mechanisms, which describe how individual choices and actions are influenced by specific combinations of desires, beliefs, and opportunities. In other words, the data emphasizes that while the importance of technology cannot be over-stated, the actions undertaken by goal-oriented actors in utilizing the potential provided by the technology is much more important (Strong et al. 2014). The data showed that the actualization of ERP capability indeed leads to business process optimization. The relevance of this perspective is the focus on goal-oriented actions and not simply the implementation of technology (Karimi, Somers, and Bhattacherjee 2007b). From this position, technology is perceived to be a contributing cause but not the sole cause of outcomes proceeding from technology initiatives (Markus and Silver 2008). We therefore expand the ERP-PIM-CIMO theory to include organizational transformation as the context in which organizational actors undertake further actions to utilize the capabilities offered by ERP post-implementation modifications. In other words, situational mechanisms (automational/informational/transformational ERP capability) by themselves do not suffice as an explanation of BPO. Rather, the actualization of ERP capability which, in many cases, requires a separate effort from the implementation of the ERP-PIM initiative must also be considered. The data indicated that although the implementation of ERP-PIMs provided situational mechanisms in the form of automational, informational, and transformational capabilities, action-formation mechanisms in the form of separate actions were necessary to actualize these capabilities and in turn produce business process efficiency, effectiveness, and flexibility. A typical example of this is the installation of EHP6 initiative (I5) at Bexplo. This initiative (I5) did not automatically result in BPO. After the installation, Bexplo needed to actualize its informational ERP capability (i.e., to further utilize the ERP system’s ability to collect, store, process, and disseminate information that had been provided by the enhancement pack) by feeding information into the ERP system. This is well described in I6, where it was observed that only after such actualization actions did it become possible for Bexplo to transform data into context-rich knowledge to support business analysis and decision-making. I6 improved automation, in addition to enhancing Bexplo’s reporting and response to customers, thus enhancing business process efficiency, effectiveness, and flexibility. In summary, installing EHP 6 (I5) provided the ERP capabilities that were later actualized in I6.

5.2 Factors influencing actualization actions

Other than the identification of necessary actions for actualizing capability, the data also identified some factors that determined whether actualization actions were undertaken. These factors enabled or constrained the actualization of ERP capability following the implementation of ERP-PIMs. They include knowledge, effort/cost, and training.

1. Knowledge

A recognition of the capability inherent in post-implementation modifications is important as organizational actors are required to undertake actions to actualize capability (Strong et al. 2014). Reflected in the following comments, if organizational actors have no knowledge of the capabilities, actions to actualize the capabilities cannot be undertaken:

‘ … if the business has more understanding of what you're giving to them, they can use more of the functionality … rather than you give everything to them and they don’t even know what you give them anyways’ (SAP Basis Administrator)

‘First we had to team up with SAP development and that gave us the ability to review what the functionality was in the enhancement pack to determine … the usability within the system’ (Business Executive)

• (ii) Effort and cost

The extent of effort and cost that is required to actualize a capability was also highlighted to be a factor that encourages or discourages actualization actions. This factor has not been identified to date in the post-implementation literature. The following comments indicate that modifications requiring considerable time and effort were usually avoided:

We find that sometimes simple changes can sometimes be quite expensive and complicated … (SAP Development Manager).

‘Upgrading is generally accompanied by the introduction of new functionality … and the thing is that implementing any new functionalities is not a cheap exercise … If an installation has a large number of custom programs, the syntax checking can be very expensive’ (IT Executive)

• (iii) Training

In agreement with several scholars (Nicolaou 2004; Pan, Baptista Nunes, and Chao Peng 2011; Ram, Corkindale, and Wu 2013), this study found training to be an important factor for gaining optimal outcomes from ERP-PIM initiatives. This is hardly surprising, given that training as been identified by several authors as a critical success factor in ERP implementations. As observed in the following comments, training provided or not provided would influence the actualization of capabilities.

‘So although the plant maintenance projects are not particularly complex, they do require in the order of thousands, pieces of master data to be collected and organised and put into the system. We also need to get people trained to use the system’ (SAP Global Support Lead)

‘In collaboration with a consulting firm, we created a solution which is enabled by our ERP system and completely aligned with existing design so that only minimal amount of training and changes are necessary … ’ (IT Manager)

Until users have been trained and have learned to use the capabilities provided by an ERP-PIM initiative, capabilities may remain dormant and under-utilized.

5.3 Consolidated findings – the ERP-PIM CIMO middle-range theory

Through the development of the ERP-PIM-CIMO model (initially presented as Figure 2b, now revised as Figure 4), this paper presents a middle-range theory to explain how causal mechanisms trigger business process optimization when organizations conduct ERP post-implementation modifications.

Figure 4. Revised ERP-PIM-CIMO Model – A middle-range theory of ERP-PIM and BPO.

This practitioner-focused theory explains business process efficiency, effectiveness, and flexibility (the outcome), and its association with ERP post-implementation modifications (the IS initiative), identifying ERP capability as the mechanism within the context of BPO. Referring to Hedström and Swedberg (1998)’s classifications of mechanisms – situational, action-formation and transformational mechanisms, the data reveals situational mechanisms (earlier identified) and action-formation mechanisms to incorporate the actions to actualize ERP capability. Automational, Informational, and Transformational ERP capabilities as situational mechanism describe how specific modifications to ERP systems influence the abilities and opportunities of individual actors with respect to the execution of business processes, and inform action-formation mechanisms describing how individual choices and actions are influenced by specific combinations of opportunities. Hence, mechanisms are now expanded to include action-formation mechanisms, incorporating actions necessary for actualizing ERP capability. The ERP-PIM-CIMO congiguration buttresses the idea that key users, through their actions, may impact the organization (Jan-Bert, Paul and Joseph 2016). Particularly, ERP systems can only continue to be a tool for business transformation through individual users who transform ERP systems’ capabilities into organizational performance (Liu et al. 2011). The place of factors that influence the actualization actions is also demonstrated as findings highlight three factors (knowledge, effort and cost, and training) that influence the actualization of capabilities provided by ERP-PIM initiatives. This revised model represents our middle-range ERP-PIM-CIMO theory and includes, based on the data, (i) Business process optimization as the context of the investigation of (ii) ERP-PIM as structure for attaining (iii) Business Process efficiency, effectiveness, and flexibility as observable events, facilitated by (iv) ERP capability as causal mechanism actualized through (v) actions initiated by actors and (iv) influenced by (vi) factors such as effort and cost, knowledge, and training.

5.4 Validity concerns for the ERP-PIM-CIMO middle range theory

As a CR study, there are valid concerns about the findings presented in this paper. To the CR researcher, internal validity establishes that a mechanism is a possible cause of the events observed in the study. In our study, it was important to distinguish between events (BPO in the form of efficiency, effectiveness, and flexibility), and the mechanisms (ERP capability) that produce the events. In evaluating how well identified ERP capabilities represent the causal mechanisms for BPO resulting from ERP-PIM initiatives, two questions were addressed:

First, have all ERP capabilities been identified? Unfortunately, one simply cannot know if all capabilities have been identified. For instance, there may be some capabilities that have been missed because they do not directly link to outcomes that can be experienced in the domain of the empirical setting. However, as the data identified a large number of ERP-PIM initiatives across two cases, a wide variety of scenarios are well represented in the study and theoretical saturation has been reached. From a critical realist perspective, the understanding of mechanisms for BPO from ERP-PIM initiatives provides at the very least a good explanation.

Second, are the identified capabilities suitable for explaining business process optimization? In other words, do identified mechanism provide a ‘good enough’ explanation that is useful and better than the alternatives? One may ask if identifying capabilities is enough, since the case data reveal that there are capabilities which have not been actualized or may have been ineffectively actualized. The response to this question is that the selection of several ERP-PIM instances across several organizations facilitated the identification of relevant mechanisms. In summary, automational, informational, and transformational ERP capabilities represent situational mechanisms for business process efficiency, effectiveness, and flexibility as outcomes of ERP post-implementation modification initiatives.

For external validity, the operation of the causal mechanism was assessed by utilizing multiple data sources; in-depth investigations of ERP-PIM initiatives in more than one organization revealed the working of causal mechanisms. As each instance of ERP-PIM represents a context for initiating ERP-PIM, several instances investigated were a means of conducting replicated studies under various contexts and serves the ultimate goal of identifying the causal mechanisms that explain the phenomenon of interest. To ensure the reliability of findings, the paper adopted measures such as time and cost savings that had been tried and tested in previous research to detect business process efficiency, effectiveness, and flexibility as outcomes of post-implementation modifications.

6. Conclusion and limitations

Combining theoretical and methodological IS perspectives, this paper explains the development of a middle-range theory on the initiation of ERP post-implementation modifications for business process efficiency, effectiveness, and flexibility in the context of business process optimization. The paper discusses and exemplifies CR-informed IS research. The proposed theory incorporates a critical realist’s Context-Initiative-Mechanism-Outcome (CIMO) configuration of the adaptation of existing ERP systems for continued optimization of business processes and draws on the notion of ERP capability as a causal link between ERP-PIM and BPO. Specifically, the CIMO perspective was adopted as a unifying view, and the resource-based view as a theoretical perspective. The theory includes the development and evaluation of propositions suggesting circumstances under which ERP-PIM generate business process optimization through the mechanism – ERP capability.

The paper makes contributions in three ways: First to the literature, the paper presents a methodological addition through successful application of the critical realist CIMO framework as a useful perspective that richly explains variations in business process outcomes resulting from ERP post-implementation modifications. Second, the paper represents the first known work that conceptualizes and operationalizes the notion of ‘ERP capability’ as ‘mechanism (M)’, drawing on the existing IS literature in which IS capability has been discussed. Third, the paper exemplifies the notion of two specific dimensions of ‘mechanism’: (a) situational – generating the existence of ‘ERP capability’, and (b) action-formation – accounting for actions necessary for exploiting ERP capability. Presently, in the CIMO literature, no explicit differentiation exists on the generation of appropriate mechanism for ideally triggering outcomes and actual application (i.e. utilization) of the generated mechanisms. The ERP-PIM-CIMO suggests that mere presence of appropriate mechanisms does not automatically trigger generation of any outcomes, unless those mechanisms have been actioned (i.e. actualized) within a specific context. Through actions, key ERP users offer organizations an opportunity to better utilize information technology (IT)-induced innovations. Expanding on the concept of actions to actualize capabilities embedded in post-implementation modifications, the paper identifies that actions are moderated by factors such as knowledge, effort and cost, and training.

A practical application of realist evaluation of ERP-PIM using a Context-Initiative-Mechanism-Outcome (CIMO) configuration pattern, the paper includes transferable and cumulative lessons to further facilitate internally and externally valid findings from IS evaluations following a CR perspective. The paper is however not without limitations. First, not all post-implementation modification initiatives undertaken by the participating case organizations were reported. Interviewees only reported modifications that they had knowledge of, or had been involved in. While the selection of participants included the consideration of participants who had good knowledge of modification initiatives, it is possible that some initiatives may have been forgotten by the participants. Additionally, in some cases, the ERP-PIM initiatives were on-going projects, and relevant information such as outcomes gained from the initiatives could not be reported. In such cases, propositions could not be evaluated, and were therefore inconclusive. Second, the adoption of CR-CIMO perspective required the identification of observable outcomes. While this allows us to focus on BPO as outcomes of ERP-PIM initiatives, it also indicates that there are possibly other outcomes worth studying in this context of digital business transformation. Despite these limitations, this paper highlights the need for further research into contexts, mechanisms, and outcomes of technology initiatives. The middle-range theory (the ERP-PIM-CIMO theory) may be applied in different contexts, with a focus on different outcomes. In addition, the operations of the identified causal mechanism in other settings and contexts beyond the present study can also be further explored.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Notes on contributors

Taiwo Oseni

Taiwo Oseni is an Information Systems Lecturer at Federation University Australia and is responsible for coordinating the Master of Enterprise Systems Research program. A current priority is to investigate Information Systems initiatives and healthcare interventions using the Critical Realism lens, particularly the Context-Initiative-Mechanism-Outcome perspective. Her goal is to facilitate the identification of affordances inherent in Information Systems and investigate causal mechanisms explaining experienced outcomes.

Mahbubur Rahim

Md Mahbubur Rahim is an Associate Professor of Information Systems at Sultan Qaboos University.

Susan Foster

Susan Foster is a Special Interest Group Lead in Organisational Change and Training at Australian SAP User Group.