Framing Wicked Problems For Enterprise-Spanning Innovation

Susan Gasson
2013 Academy of Management Proceedings  
INTRODUCTION Organizations are increasingly focusing on the design of organizational information systems that align business process redesign with enterprise systems (ES) changes. Enterprise-spanning innovation projects cross multiple functional boundaries and business units, requiring many different disciplinary, organizational and political interests to be negotiated and multiple, diverse ways of working to be reconciled. While the critical processes of boundary-spanning design take place
more » ... tream" of the waterfall process of information technology (IT) implementation, the processes of negotiation and reconciliation needed for strategically-aligned business adoption adopts a "liminal space" between the status quo and strategically-desired outcomes. In regular ES change projects, this space must be bridged by human mediators, who negotiate appropriate use and integration procedures (Wagner et al. 2012 ). If boundary-spanners lack sufficient skill or organizational power, this results in fragmented implementation of organizational information systems that are designed around multiple, often conflicting aims (Levina and Vaast 2005; 2008) As a result of prior failures, leading-edge organizations are increasingly adopting approaches based on the co-design of business processes and information systems. These initiatives attempt to integrate the technology and data-driven design interests of IT professionals with the strategic awareness and network influence of business managers. But we do not understand how to manage this type of project, as we have little understanding of how such divergent perspectives can be bridged in practice. This paper follows one such initiative over two years, to understand the processes by which business and IT interests may be bridged. We take a sensemaking approach, to understand how the divergent framing perspectives employed by various OCIS Paper 16707 3 stakeholders are negotiated and merged, to bridge the diversity of domain-related frameworks for action involved in complex innovation projects. CONCEPTUAL UNDERPINNINGS Boundary-Spanning Collaboration As Organizational Sensemaking The creative processes underlying joint sensemaking "emerge from a process of negotiating multiple and potentially competing interests between different communities or groups within the organization" (Drazin et al. 1999, page 286). A stable (accepted and incorporated) design does not need to be based on the establishment of common understandings but rather on the intersection between different positions and perspectives (McLaughlin et al. 1999 ). This involves the elicitation and sharing of organizational knowledge about the meaning of work-practices across multiple organizational and domain boundaries (Carlile 2002; 2004) . The meanings that we attribute to business processes and technology are defined by reference to culturally-situated frames or mental models: the adoption of specific perspectives that derive from our experiential learning, our membership of specific interest groups or communities of professional practice, and our organizational affiliations -and that underpin our belief-systems and expectations, providing a script for action that fits with the situation at hand (Goffman 1974; Tannen 1993; McLoughlin et al. 2000) . To reflect the complexity of real-world innovation and to account for interaction effects in framing, we need to employ multiple levels of analysis (Drazin et al. 1999) . Curtis et al. (1988) , in discussing how understanding varies across individuals, teams, project groups (multiple teams), the organization and its interactions with its external environment, suggest three levels of analysis: individual, group, and organizational. These levels of analysis are also employed by Weick (1995) in his analysis of sensemaking and by Drazin et al. (1999) in their OCIS Paper 16707 4 exploration of organizational sensemaking. We therefore employ these three levels: (i) the individual level of cognitive framing, (ii) the group level, which defines joint frames, and (iii) the organizational level, which reconciles distributed frames across knowledge domain boundaries. Individual Framing and Organizational Learning Specific belief structures, or frames of reference, permit individuals to make sense of phenomena or events in terms of their own, individual interpretation of reality (Goffman 1974). They make sense of events and determine how to act as a consequence of an automatic "ordering" of reality, that we refer to as sensemaking (Weick 1995) . A frame analysis examines how individuals and groups place a communication or interpret an event within a relevant context and interpret its meaning. Frames may be analyzed by means of discourse analysis -by examining language-terms, constructs and metaphors, it is possible to understand how various people take a position on, or interpret the subject of discussion in different ways (Tannen 1993; Ensink and Sauer 2003) . A specific frame embodies a set of expectations that constrain action and the scope of change: this tends to lead to "automatic" or unquestioned action or a resistance to change (Goffman 1974; Tannen 1986; Winograd and Flores 1986) . To remove these unconscious forms of resistance, we need to "break frame," making explicit and questioning the dominant frame that stakeholders in that situation inhabit, causing them to reframe the situation (Goffman 1974; Tannen 1986) . Reframing may be triggered by cognitive breakdowns. Heidegger (1962) argued that objects and their properties are not part of an objective reality, but become apparent only in the event of breaking down, in which they change from "ready-to-hand" (used automatically) to "present-at-hand" (requiring reflection). Breakdowns can be used constructively in design: a breakdown is "a situation of non-obviousness" (Winograd and Flores 1986, p. 165). Cognitive breakdowns uncover aspects of the design or change related to the OCIS Paper 16707 5 context of use or operation and so provide a source of learning. For example, IS analysts appear to define a system according to implicit assumptions that are not questioned or realized until they conflict with explicit user requirements, during user interactions (Malhotra et al. 1980; Urquhart 2001) . So a breakdown may be the mechanism by which an individual breaks out of or adapts an existing design frame to include new evidence (Beynon-Davies and Holmes 2002; Gasson 2007) . To examine what triggers a breakdown, we employ the concept of cognitive dissonance. When there is a discord between incompatible belief structures, for example when an individual's assumptions of how a process works are inconsistent with how the individual is rewarded for performing that process, the individual experiences cognitive dissonance (Festinger 1957) . People avoid exposure to information that is likely to increase dissonance and are more likely to change their attitudes and beliefs to accommodate behaviors to which they have already committed, than to change their behavior (Festinger 1957) . Members of socially-cohesive groups appear to experience dissonance-reduction that brings about a commitment to new methods and ideas, even when these conflict with previously-held beliefs (Nelson and Cooprider 1996) . So one would expect increasing individual commitment to group perspectives as social disparity is reduced. Shared Framing and Group Consensus Members of a local community of practice, such as a functional group within an organization, tend to develop systems of shared values, belief-structures, culture and norms that govern how group members behave, interact and communicate. These shared understandings provide a framework for action, that allows group members to take shortcuts in communication and decision-making -they implicitly know "how we do things here" without the need for debate or agreement (Lave and Wenger 1991). People who work together regularly develop OCIS Paper 16707 6 shared frames that encompass not just how work is done, but why (shared rationale) and by whom (role allocations within the group) (Brown and Duguid 1991). Shared frames represent a negotiated order that permits a group to operate cohesively rather than wasting effort in resolving differences in perspective (Walsh et al. 1988 ). It is important to distinguish between intersubjectively-shared group frames and consensus, which depends on convergent, rather than shared framing. Organizational stakeholders determine how to act collectively based upon the joint meanings that they construct with other knowledgeable actors through argumentation and debate (Rittel 1972) . Sociotechnical studies use the term technological frame to understand how members of competing interest groups understand the role and purpose of technology as it is mutually constituted as part of organizational change. A technological frame represents "the concepts and techniques employed by a community in its problem-solving ... frames are located between actors" (Bijker 1987: 171). This is very different from knowledge that is explicitly shared. Individuals are rarely aware of the influence of the technological frames that they call upon, as these are embedded in unquestioned assumptions and understandings about work that result from shared experience within a specific community of practice. The term "technological frame" is problematic, in that it derives from an explicit focus on the role of technology in sociotechnical studies. But it is also useful in its wider application, as we need to understand that a boundary-spanning design or change group is unlikely to develop shared frames that allow them to define the problems that they face in common. Instead, boundary-spanning projects rely on the negotiation of multiple (technological) frames, each of which has its adherents within the group as these "make sense of" the rationale for design in terms of the framework for action that they are used to applying. We cannot assume cognitively-shared understanding just because group members share a similar OCIS Paper 16707 7 organizational culture -shared understanding depends on a history of developing shared routines and values (Krauss and Fussell 1991; Miranda and Saunders 2003) . Boundary-spanning groups instead rely upon consensus, which accommodates the perspectives of others when these do not entirely conflict with one's own: "Meaning ... encompasses multiple dimensions. Consensus may develop around one dimension of meaning and not around another" (Fiol 1994, pp. 404-405). Consensus-building is emergent, as the knowledge required for complex change is often distributed across many different people and it is difficult to predict what knowledge will be required for specific areas of change (Fiol 1994; Markus et al. 2002) . Joint sensemaking in groups that span multiple knowledge domains relies on the ability to challenge one's taken-forgranted frames and experiential knowledge in order to define new frames that accommodate the views of others. A Punctuated Equilibrium Model of Boundary-Spanning Innovation At the organizational level (where most change projects operate), boundary-spanning initiatives involve distributed cognition, where understanding is "stretched across" rather than shared between stakeholders (Lave 1988). Negotiated outcomes cannot rely on leveraged intersections, or interdependencies between stakeholders, who align their joint interests around a slim degree of shared understanding: " Distributed cognition is the process whereby individuals who act autonomously within a decision domain make interpretations of their situation and exchange them with others with whom they have interdependencies so that each may act with an understanding of their own situation and that of others." (Boland et al. 1994, p. 457) OCIS Paper 16707 8 The process by which groups develop a solution in organizational change projects has therefore been described as "punctuated equilibrium," where long periods of equilibrium around a consensus model of the organization are punctuated by rapid disruptions during which the project group or organizational managers reframe joint understandings to form a consensus that fits with changes to the firm's environment. This theory has been applied to explain why firms periodically restructure or reorganize their technology infrastructure (Sabherwal et al. 2001; Silva and Hirschheim 2007) . If we apply the concept to boundary-spanning design and problem solving initiatives, we might argue that relatively stable organizational frames are punctuated by disruptions that cause the group to completely redefine their problem and as a consequence, rethink their solution. There is some evidence for this. Gersick studied a number of project groups, from MBA student groups involved in 3-month projects, to hypothesizes that this is driven by group perceptions of time-constraints, observing that there is always a major redefinition of the problem structure midway through the project duration (Gersick 1988; 1989; 1991) . A similar type of punctuated or interrupted process has been observed in user-analyst system requirements definition (Newman and Robey, 1992) and in larger scale studies of heterogeneous IS requirements analysis (Bergman et al., 2002a; Bergman et al., 2002b). But these studies focus on relatively heterogeneous groups and do not delve into the detail of how or why disruptions occur. To investigate this issue in enterprise-spanning groups, we need an analytical lens that explores not only the evolution of problem and solution spaces over time, but also the catalysts that drive an evolving design consensus across design participants and stakeholders from multiple knowledge domains. The study presented here suggests a framework for boundary-spanning problemsolving and design that incorporates the three levels of analysis. OCIS Paper 16707 9 RESEARCH SITE AND METHOD This study explored the design of strategic management systems in a US University (not the author's own institution) by participant observation in a strategic taskforce for enterprise information systems design. The author engaged in a longitudinal, ethnographic study of monthly or bi-weekly taskforce meetings to define changes to the business processes and enterprise systems used for University management, over two years. While two members of the taskforce left and were replaced over the period of the study, representation of specific groups Periodic interviews were performed with taskforce participants and other stakeholders, to understand how they --and their organizational groups --framed the requirements for change. Graphical and textual representations of business processes, information flows and requirements, organizational problems, and proposed solutions were collected and analyzed, to explore their role in the processes of group design. Discussions from 27 meetings over the two-year period OCIS Paper 16707 10 were recorded and interpreted by means of a qualitative discourse analysis (Tannen 1993; Klein and Truex III 1996; Davidson 2006) . The research interpretation of stakeholder frames was validated in ad hoc discussions and in formal interviews as the project proceeded. Stakeholder frames were identified and traced across the duration of the project, using a grounded theory approach (Glaser and Strauss 1967; Glaser 1998) . Themes were identified, to trace commonalities and differences across individuals, the group (consensus) perspective, and points of divergence in project debate, both within the change management group and between the group and external organizational stakeholders, such as influential managers. During this process, it became clear that there were discontinuities in consensus during which the levels of disagreement and conflict were significantly higher than at other times during the project. Explanations were sought in the literature to explain this, which allowed the development of concepts provided in the discussion of findings, below. FINDINGS Precursors To Enterprise System Framing The taskforce was assembled by the Director of Information Services and the University Registrar, in response to a perception that the University's Enterprise systems (ES) were inadequate for financial management and reporting. The University employed an ES that was used by many similar institutions, but which had not been fully implemented due to political considerations. Several functional groups, in particular Human Resources, were suspicious of the introduction of an overarching administration system, viewing this as an attempt to impose
doi:10.5465/ambpp.2013.17607abstract fatcat:hiph73aamrehffajsnxtds4pji