Indicative Case Example: Knowledge Sharing in Compound UK

Compound UK (a pseudonym) sells pharmaceutical products to hospitals and general practices, whilst also undertaking clinical trials of new drugs with participating doctors in Great Britain and Northern Ireland. It is part of a major multinational company and employs around 300 people, half of whom are involved in selling activities while the rest work in areas such as marketing, accounting, human resources, or as medical experts. Lotus Notes (Notes) was introduced as a means to encourage employees to draw on all areas of the organization to work and share information and knowledge across functional and geographic boundaries. Such knowledge sharing was seen as beneficial not least because of the complexity of sales situations which involved many different decision makers (such as different doctors, pharmacists, managers, and accountants), just as clinical trials involved many different physicians.

In addition to an electronic mail (e-mail) facility, there were three main uses of Notes, as summarized in Table 5.2. First, it was used to create a strategic selling database to support the co-operative activities of staff from different areas of expertise (such as sales, marketing, and medicine) involved in seeking to secure a sale to a particular hospital or general practice. The second and most prevalent use of Notes was the contact recording database. Any employees that visited a client were required to record specific details about doctors such as the outcome of the meeting, the doctor’s personal and medical interests, and their involvement in clinical trials. Notes databases were also set up that focused on issues, products, or a particular role with the intention that employees would discuss relevant issues.

Table 5.2 Key uses of Lotus Notes in Compound UK

Sharing knowledge explicitly across boundaries

In relation to the framework represented in Figure 5.1, the introduction of Notes saw employees ‘acting’ in terms of selling products or undertaking clinical trials, and then ‘accounted’ for their practice by making explicit their activities and views on Notes’ various databases which bridged the spatial, temporal, and functional boundaries. In trying to make explicit their sales and medical work, employees were required to make sense of their own activities in ways that were comprehensible to others (sense giving), and also make sense of (sense reading) and utilize the accounts of others. Sense making, giving, and reading are highly reflexive processes and are inextricably interlinked.

Acting and accounting on role or functional specific databases was relatively unproblematic. This was due to the intimate understandings and practice they shared with colleagues working in the same function. This local mastery was sustained not only through the use of Notes, but also through regular formal and informal face-to-face discussions and interactions.

However, when such practices spanned knowledge domains, acting and accounting practices were not as productive. Due to the limited number of company-wide meetings, staff had little experience of working directly with colleagues in different functions. This left the organization-wide Notes databases as the primary mode of collaboration. Such difficulties using Notes to work across functions were explained by one medical expert who said:

Staff from across the different functions thus viewed much of what employees were able to provide on account of the shared databases as only being a simplistic representation of what they sought to convey (sense giving). Reps themselves were also sometimes unsure as to the meanings that others sought to convey in their posts, and they could not understand why some people would request further clarifications about posts they had recorded. This left them feeling ill disposed to record detailed information in the future and thus was further implicated in limiting sense making, giving, and reading processes.

This lack of familiarity with other domains of knowledge resulted in many employees experiencing considerable uncertainty about what they required from each other, and exactly how they could draw on each other’s expertise. They feared not only that what they recorded may be misunderstood but that it might appear stupid or irrelevant to other employees, particularly by those in other professional domains. Further, as their use of Notes was in addition to their existing activities, many claimed that they did not have the time to review and/or contribute to the databases. Indeed, making their activities explicit in this way on an on-going basis was alien to most employees who were only typically required to account for what they had done when problems arose (Giddens, 1984: 6).

Understood from a practice perspective, overall the framework and the case indicate one key challenge pertaining to how knowledge management initiatives typically require people to make sense of the accounts recorded by members of other professions on shared forums and then utilize them in their own activities. The case has indicated that a reliance on accounts recorded on discussion forums, in order to collaborate across professional disciplines, are unlikely to provide a sufficient insight into the ‘path of thinking’ or different conventions that underpin the logic that other professions draw upon in their practice (Schultze, 2000; McDermott, 1999). Sense making, sense giving, and sense reading processes rest upon the shared practices of those posting to and reviewing shared databases. When practice is not shared then a reliance on reflexively monitoring and making sense of the accounts others have recorded on shared forums will always be limited as they are not fully representative of the histories, activities, and assumptions that underpin their formulation. As such, sense giving accounts are only likely to be partial rationalizations of the complexity that underpins their practice (Brown, 1998; Hayes, 2001). This again indicates the importance of recognizing that knowledge cannot be encoded on ITs, and instead is shared through social networks (Hislop et al., 2000). Consequently, for acting and accounting processes to be more productive they will require the ongoing development of social networks and reflexive engagement in joint practice, in addition to solely the use of shared forums (as was the case with the function specific databases). This ongoing sense making process provides a basis for people to develop a deeper understanding of the assumptions that underpin the work practices and formulation of accounts recorded by members of different professional groups (Brown, 1998).

Political and normative dimensions participation

This subsection focuses on the ways in which political and normative issues shape the extent and nature of acting and accounting processes. I draw on Hayes and Walsham’s (2000a, 2000b, 2001) constructs of political and safe enclaves to do this. Political enclaves describe contexts for participation that resemble a public façade, while safe enclaves describe contexts for participation that allow for more genuine views and assumptions to be exposed.

Political enclaves

A first important theme pertaining to acting and accounting processes concerns the way senior managers may harness the surveillance capacity that political enclaves provide to reinforce their dominant positions (Newell et al., 2000; McKinley, 2000). Senior managers harnessed the surveillance capacity Notes provided to try to ensure that sales reps worked in the ways that they promoted. For example, not long after Notes was introduced, the Commercial Director instructed the Notes developer to devise a league facility which could indicate centrally how many contact records and strategic selling sheets had been completed by each rep. A senior manager would then send out electronic messages to field force managers commenting on their reps’ position in the league tables. Further, some senior managers would regularly review detailed comments and observations that reps had recorded, and contact their area manager if they thought they could be approaching sales or recording contacts better. This confirmed many of the reps’ suspicions about the intention behind the electronic contact recording and strategic selling databases, namely to increase central surveillance and control. This process of surveillance had the effect of simultaneously increasing the dominance of senior management while reducing the autonomy of middle managers dealing directly with the sales force, which strengthened hierarchical control and reduced trust and co-operation (Pan and Scarbrough, 1999; Zack, 1999a; Storck and Hill, 2000).

Such surveillance and control activities influenced the extent and nature of acting and accounting processes in Compound UK. When staff made sense of their own activities and then decided how they would convey (sense giving) this to others, they did this by identifying what would be politically and normatively acceptable to senior managers and others that may review their accounts. For example, in relation to the contact recording and strategic selling databases, even though non-careerist reps saw some of the use they made of them as being futile, they still felt normatively bound to record a sufficient amount and in enough detail so as to not be identified by senior managers. They felt frustration with their peers for having to undertake what they saw as being a futile and time-consuming task.

Similarly on the discussion databases careerist employees harnessed the visibility to indicate to senior managers that they were contributing to the development of new insights, while also concurring with their views. However, if non careerist reps disagreed with comments posted by senior managers, or those supporting the views of senior managers, rather than stating this on a database they considered it better not to post their own views. Indeed this led to a homogenizing of views that reflected those of senior management. Such examples highlight how sense making, giving, and reading processes are inextricably interlinked with the political and normative context (McDermott, 1999). Examining the latitude which employees have to distance themselves from received wisdom is a fundamental challenge for acting and accounting processes (Brown and Duguid, 1991).

In relation to sense making, giving, and reading, such career orientated politicking further reinforced the homogenizing of acting and accounting processes and had certain implications. For example, careerist reps formulated their activities and accounted for them in ways that they thought would meet senior management approval, and hence enhance their career prospects. Furthermore, when non-careerist reps made sense of the representations others recorded on the cross-functional discussion databases, many were aware of the dominating and careerist undertones that lay behind them. Further still, a result of this politicizing was that many non-careerist employees did not see any benefit to participating in discussions that were dominated by what one rep described as: ‘ambitious people competing with each other about who could shout the loudest.’ Though several studies have made a cursory reference to the importance of the reward structure in knowledge management initiatives, very few have explicitly considered how the career reward system can influence the shape and nature of participation. Of those that have, Orlikowski’s (1993) study found that consultants did not make any use of Lotus Notes as it was not associated with either the career or the financial reward structure, while Quinn et al. (1996) discovered that modifying the reward structure to emphasize reciprocity was vital to the success of Merrill Lynch’s and NovaCare’s knowledge management initiatives. However, neither study highlighted how the visibility that political enclaves provide may exacerbate careerist motivations which arise from an individualistic reward structure, and how this may assist or hinder knowledge management initiatives. Such an understanding thus suggests that accounts of knowledge management need to be more sensitive to the hierarchical, coercive, and competitive nature of knowledge work (Knights et al., 1993).

Safe enclaves

In contrast, safe enclaves did provide opportunities for productive acting and accounting processes (Pan and Scarbrough, 1999; Zack, 1999a; Storck and Hill, 2000). Some discussion databases were considered to be safe(r) from heavy politicizing and careerism. Some non-careerist employees did include themselves in the functional or regional specific shared databases. On these databases, many non-ambitious reps would discuss the ways they had approached sales situations and other areas of interest that they shared. In addition, those employees that had experience of working with members of other functions would discuss and provide advice to less experienced members of their own community on how they could best interact with members of other functions. However, ambitious reps did not make much use of these local forums, as they viewed their time being better spent on career enhancing national discussion forums. In this sense, while safe enclaves did provide opportunities for staff to express their thoughts, and take into account the posts of others (sense reading) in their activities, there were still many other employees who did not participate, as they saw their time as being better spent using the community wide political enclaves to further their own career aspirations.

The concept of safe enclaves suggests that the relative success of acting and accounting processes rested upon several important features. First, they were all domain specific, which meant that employees felt confident that their representations would be understood by members of their own professional domain and that they could easily be made sense of as being genuine. Second, they were all optional, which meant they were not associated with the normative control structure. Third, senior managers were not privy to the discussions that occurred on safe enclaves.