Role-related misfits

Soh et al. (2003) study the use of an ERP package in a hospital and identify ‘opposing forces’ between the ERP package and the organization as sources for misalignments.

Among others, the authors identify the embedded process-oriented structure of the ERP package as being in opposition to the function-oriented structure of the organization, which creates ‘job scope’ misalignment between the system and the users in the organization. The traditional functional structure of the hospital with clear demarcation between administrative and medical staff was thus in opposition to the process-oriented structure of the system where the handling of transactions was not limited by functional boundaries. An example of a resulting misalignment in their case study is that “nurses’ job scope would be expanded to include capturing information about patient location, attending physician, and treatment department because they were the “person on the spot” at certain important points of the transaction process (in this case, patient movement).” (p. 92). The increased job scope of the nurses may thus be interpreted as entailing lower role specialization, as a result of the introduction of the ERP system. This change is categorized as a misalignment by Soh et al. (2003), due to the extra effort in checking and correcting data that occurred from medical staff being unfamiliar with administrative tasks, although the change also resulted in the more current and detailed information on patients and more efficient workflows.

Volkoff and Strong (2010) report similar role related misfits. Using Grounded Theory (Strauss and Corbin 1990) as the methodological frame, they conduct a longitudinal case study of a manufacturing company during the implementation of SAP for the purpose of identifying ES misfits. Strong and Volkoff propose that “role misfits occur when the roles in the ES are inconsistent with the skills available, create imbalances in the workload leading to bottlenecks and idle time, or generate mismatches between responsibility and authority” (p. 742). Similarly to the findings by Soh et al. (2003), Strong and Volkoff report that each role needed more understanding and knowledge of the network of tasks to be performed and had to spend more time performing coordination activities between functional areas. The case study thus reports an example where the ‘material planner’ role was enlarged to include acquisition of the

required material and the organization thus had to create a new ‘buyer material planner’ role to accommodate the embedded processes of the system. The authors report that the users perceived the system as the foundation for this new role and that the responsibilities concurrently changed from a tactical to strategic level that some users could evolve into while others could not. On the other hand, Strong and Volkoff point out that the fixed role authorizations in the SAP system increased the number of people involved in any action and made it difficult to overlap the tasks carried out by each role, entailing a role narrowing. In turn, this caused limited managerial flexibility for dynamically reassigning workforce as needed. Finally, (Strong and Volkoff 2010) reports the location of tasks related to the ‘shipping role’ being spread across many of the predefined roles in the system.

Another example from the literature of potential role misfit between ESs and organizations is Carlsson and Hedman’s (2004) evaluation of 329 EP “role templates”

in SAP’s mySAP Portals. Using Quinn and Rohrbaugh’s (1983) Competing Values Model as theory for the evaluation, they point out that the predefined roles primarily supported operative and middle management roles but lacked support for top manager roles.

In order to arrive at a more theoretical level of abstraction, as opposed to a descriptive level, we may introduce Strong and Volkoff’s (2010) misfit types of deficiencies and impositions (see section 3.4.4) to the domain of role-oriented ESs. We may thus apply the deficiency label to misfits that are either attributable to the particular role-oriented ES and the imposition label to misfits that are inherent to the role-oriented approach in general. By using these two types of misfits as overall categories we are able to distinguish between the misfits that can be addressed through adaption of the ES at the design or implementation and misfits that are inherent to representing organizational roles in ESs.

3.4.5.1 Role aggregation

The research in paper V and VI illustrates misfits related to role aggregation of user in organizations. This misfit occurs when a user occupies multiple roles and these roles are represented in different user interfaces. Although predefined system roles may exist in the role-oriented ES that maps perfectly with the respective roles of the user, the tasks and information are located in different user interfaces and the user thus has to switch between the interfaces when switching roles, as illustrated in Figure 9. The partner companies reported that since the NAV 2009 RTC was primarily targeted at SMEs, many users in the customer organizations had multiple roles. The level of role aggregation of the users did thus not always fit with the roles aggregated in the predefined user interfaces. The role aggregation misfit was also found in the study of client organizations (Customer 4) where the interviewed user had multiple roles that required switching between the predefined user interfaces.

Figure 9. Role aggregation misfit.

Depending on the difficulty of switching between the role user interfaces, an amount of time is spent switching between the interfaces, decreasing the performance of the user, and undermining the ease of use associated with the role-oriented approach (see section 3.4.3.1). In the case of NAV 2009 RTC, users can only be associated with one role-oriented user interface at a time, and significant time consumption and inconvenience is thus associated with switching between two interfaces. Consequently,

some partner companies reported, that when users experienced the role aggregation misfit, they were prone to abandoning the role-centers in NAV 2009 RTC altogether and use the old front-end client with the conventional user interface instead. As the appearance of the role aggregation misfit depends on the role aggregation of the predefined system roles, the misfit may be categorized as a deficiency misfit, as flexible role aggregation or a match between the predefined role aggregation and the actual roles may remedy the misfit.

3.4.5.2 Role scope

Role scope differs conceptually from role aggregation in that role scope is concerned with the relationship between a role and its tasks while role aggregation is concerned with the relationship between a user and her roles (see section 3.1.1). By combining the findings in paper V and VI, we may argue for two different types of role scope misfits.

First, a user may occupy a role that entails carrying out a number of tasks, but the predefined role assigned to the user includes more tasks than required, as illustrated in Figure 10. We may term this type of misfit as the role overload misfit, similarly to the term used in organizational role theory to describe the situation in which a role has too many tasks (see section 3.1.3). The user does thus not gain full benefits of the role-oriented approach, as tasks and information that are superfluous to the role of the user do not serve any purpose and limits the ease of use gained from the role-oriented approach.

Figure 10. Role overload misfit.

Second, a user may occupy a role that entails carrying out a number of tasks, but the tasks are located in different predefined role user interfaces, as illustrated in Figure 11.

We may term this type of misfit as the role segregation misfit. The user thus needs to switch between the user interfaces while working in the same role to gain access to the required tasks. The consequences of this misfit are thus similar to those of the role aggregation misfit. Both types of role scope misfit (role overload and role segregation)

may be categorized as deficiency misfits, as they not inherent to the role-oriented approach itself, but to the fit between particular system roles and actual roles.

Figure 11. Role segregation misfit.

Additionally, a user may occupy a role that entails carrying out a task that is not supported by the system and hence not present in any of the predefined role user interfaces. However, this situation simply represents a situation where the system does not support a required task and should not be mistaken as a misfit that is attributable to role-oriented ESs, as the issue applies to ESs in general – not just role-oriented ones.

3.4.5.3Role specialization

The client companies in the research used a different number of the role-centers in NAV 2009 RTC, ranging from two to eight (see paper VI). While this difference may be partially explained by differences in organizational coverage of the ES, it points to differences in role specialization of organizations as a potential misfit for role-oriented ESs. An example of the role specialization misfit is the industry-specific role misfit discussed in paper V. In the case of industry-specific role misfits, the desired tasks or information may be available in a “generic” variation but not in the specialized industry-specific variation required by the role. In the case of NAV 2009 RTC, the ISVs that had upgraded their add-ons (both industry-specific and cross-industry) to be compatible with NAV 2009 RTC had made the functionality of the add-ons available in all the predefined role-oriented user interfaces. Although certain industry specific functions may be relevant to all roles in a company, most companies have generic cross industry roles that are not specialized for a certain industry (Pugh et al. 1968).

Making all the industry-specific functions available to all predefined roles thus undermines the role-oriented approach and limits the role specialization. As it is possible to address this misfit by only including the industry-specific functions in the predefined roles where they are needed, the role specialization misfit may be categorized as a deficiency misfit.

The concept of cross-industry misfit is also addressed in paper V. However in the case of cross-industry misfit, no user in any industry is able to access the desired task or information, as it is simply not available. While the absence of cross-industry functions

thus constitutes a misfit for organizations needing the functions, it is thus not a misfit that is attributable to the role-oriented approach per se but rather a case of missing functionality all together.

3.4.5.4Role narrowing

Previous studies have identified examples of both role narrowing and role enlargement in implementations of conventional ESs (see section 3.4.5). A central condition for achieving the benefits of role-oriented ESs (see section 3.4.3) is to syndicate tasks and information into different user interfaces. Role-oriented ESs thus inherently cause additional role narrowing, compared to conventional ESs with a single unified user interface where all tasks and information are available to every user. For the same reason, it is unlikely that role-oriented ESs will entail role enlargement when compared to conventional ESs. Strong and Volkoff (2010) classify role narrowing as an imposition misfit and identifies limited flexibility for dynamically reassigning workforce as a consequence caused by role narrowing. Although client organizations may even perceive the role narrowing and demarcation of responsibilities as a distinct advantage of oriented ESs (see section 3.4.3.3) this pivotal attribute of role-oriented ESs is also the cause of role-related imposition misfits.

3.4.5.5 Role transition

Role-oriented ESs may entail a potential misfit related to transition (see section 3.1.3 for elaboration of role transition). Although radical changes in organizational structure and the role set of an organization may cause misfits at the role level in conventional ESs as well, role-oriented ESs are more susceptible to such changes as they are adapted to a specific role set at the individual and the organizational level. Changes to the role set at either the individual or organizational level may thus cause role transition misfits. This type of misfit is arguably magnified by the role-oriented approach regardless of the approach to representing the roles at the interaction level and may

thus be perceived as an imposition misfit of role-oriented ESs when compared to conventional ESs.

However, the misfits related to role transition may be increased depending on representation of roles at the interaction level of role-oriented ESs. In the unified approach to representing roles at the interaction level, the predefined system roles are tighter coupled (see section 3.2.1.2), and changing the role(s) of a user thus requires more adaption than in the componentized approach, where mappings to predefined roles are more readily addible or removable. This circumstance suggests a deficiency dimension to the role change misfit of role-oriented ESs, as the degree of the misfit can be lowered depending on the approach to representing the roles in the system.