Professional Documents
Culture Documents
Design Principles For Competence Management Systems - A Synthesis of An Action Research Study
Design Principles For Competence Management Systems - A Synthesis of An Action Research Study
SPECIAL ISSUE
and the action research method, with its iterative quences of our interventions. We conclude with
hypothesis development and testing, is particularly not only a set of revised design principles, but also
appropriate for the development of system design an assessment of our research vis-à-vis the
principles (Walls et al. 1992). Thus, we conducted criteria for evaluating canonical action research.
a 30-month action research study, which consisted
of two cycles with the following phases: diag-
nosing, action planning, action taking, evaluating,
and specifying learning (Susman and Evered Competence in Organizations
1978). The study involved six Swedish organiza-
tions that also partially funded our project. The The literature on competence in organizations
remaining financial support came from VINNOVA.2 appears to be divided along disciplinary lines. The
strategy literature focuses on the macro or
Due to the long duration of the research and the organizational level of analysis and concerns itself
conditions of our funding, we published insights with the notion of core competence as a means of
and intermediate results at various stages of the generating competitive advantage (Prahalad and
project so as to secure ongoing financial support. Hamel 1990). According to Lado and Wilson
(1994, p. 702), core competencies
These publications reported on CMS imple-
mentation failures (Lindgren and Henfridsson
include all firm-specific assets, knowl-
2002), CMS design assumptions (Lindgren et al.
edge, skills, and capabilities embedded
2003), and CMS design principles (Lindgren and
in the organization’s structure, techno-
Stenmark 2002). The research contribution we
logy, processes and interpersonal (and
offer here goes beyond these earlier publications
intergroup) relationships.
in that our analysis considers the 30-month action
research project in its entirety. We develop an
Thus, at the organizational level, structural fea-
integrative model of competence, a competence
tures such as culture (Barney 1986), routines
typology, and consider the unanticipated conse-
(Nelson and Winter 1982), and learning (Hamel
quences of our design principles for the first time.
and Prahalad 1994) are sources of a firm’s core
We thus synthesize all the steps in our study and,
competence, and hence, its competitive advan-
based on the lessons learned, refine our initial
tage.
design principles.
In contrast, the HR literature focuses more on the
The paper proceeds as follows. First, we review
micro or individual level of analysis and views
the literature on competence and develop a model
competence as “an underlying characteristic of a
that integrates macro and micro level definitions of
person, which results in effective and/or superior
competence and incorporates a typology of com-
performance in a job” (Boyatzis 1982). The per-
petence. This is followed by a method section that
sonal characteristics that facilitate high perfor-
describes action research in general, the criteria
mance (and that are therefore part of individual
by which it should be evaluated, and details about
competence) include motivation, disposition, self-
our particular action research project. Then, we
image, values, moral standards, norms of social
present our two action research cycles. In our
behavior, and traits, as well as communication,
discussion of the research findings, we highlight general reasoning, and learning capabilities
both the anticipated and unanticipated conse- (Bergenhenegouwen et al. 1996; Rothwell and
Lindholm 1999).
integrate these two perspectives on organizational all crucial in an organization that seeks competitive
competence through the development of taxono- advantage through core competencies, are gener-
mies and theoretical frameworks (e.g., Muffatto ally neither acknowledged nor developed in the job
1998; Nordhaug 1998; Rothwell and Lindholm paradigm.
1999; Simpson 2002). Indeed, competence-based
theories of the firm have been developed (San- In contrast, the skill-based approach to HR
chez and Heene 1997; von Krogh and Roos 1995). management focuses on the individual and his/her
ability to contribute to the organization’s core
A number of these integration efforts highlight the competence and competitive advantage (Lawler
need for alignment between the organization’s 1994). Instead of relying on job descriptions, a
strategic orientation and the assumptions under- skill-based approach relies on person descriptions,
lying its HR practices (Bergenhenegouwen et al. which identify the skills and behaviors that an
1996; Lado and Wilson 1994). Hagan (1996) sug- individual needs to be effective in a particular work
gests that an organization’s adoption of a core area. With its emphasis on competence, the skill
competence perspective will require shifts in job paradigm focuses more on behavior than on tasks
and reward system design, as well as in staffing and processes. The skill-based approach is
and training practices. For instance, in a core particularly effective in situations requiring knowl-
competence organization, more work is done in edge and/or team work, as both imply a relatively
project teams and individuals move around the high degree of self-management and the individual
organization to complete different assignments. In worker’s ability to add unique value to products
addition to challenging the individual employee by and services. Furthermore, skill-based remunera-
demanding more effort, flexibility, and motivation in tion systems reward employees for learning and
such a competence-based organization (Bergen- flexibility, and for developing skills that allow them
henegowen et al. 1996), these changes in job to complete multiple tasks.
design challenge the value of job descriptions and
HR practices such as hiring and training that are Lawler and Ledford identify a number of chal-
based on assumptions of more stable jobs and lenges that face organizations wishing to manage
individually-assigned tasks (Lawler 1994). competencies. Shifting from a job-based orienta-
tion to a skill-based one requires significant
Lawler and Ledford (1992) distinguish between change in the physical (e.g., systems and prac-
job-based and skill-based approaches to HR tices) and conceptual (e.g., assumptions and
management, and argue that HR departments beliefs) infrastructure of HR departments. For
need to adopt a skill-based approach in order to instance, the selection of individuals for organi-
support their organizations’ development of core zational membership rather than for a particular
competencies. They highlight that the traditional, job is relatively foreign to organizations that have
job-based approach develops job descriptions and traditionally operated in a top-down, planned
then tries to find and shape individuals to fit them. manner rather than an emergent one. Further-
They contend that this paradigm is problematic in more, organizations will need to invest in new
contemporary organizations because job descrip- technology that supports a skill paradigm (Lawler
tions are generally based on how the organization and Ledford 1992).
has operated in the past, with little or no appre-
ciation for its future needs. Furthermore, the job- Having highlighted the need for alignment among
based approach fails to take into account indivi- the structural features of organizational com-
duals’ abilities to contribute to the organization’s petence, especially between the organization’s
success beyond the boundaries of their job. By strategic orientation and its HR infrastructure, we
incentivizing and evaluating employees within their now turn our attention to individual-level com-
job description boundaries, capabilities such as petence and its development. Individual compe-
learning, flexibility, communication, collaboration, tencies are skills that are critical for individuals to
and innovation across organizational boundaries, master if they are to achieve high performance in
the completion of a task (Boyatzis 1982). Even element in an individual developing along a career
though knowledge is central to individual com- track is the individual's desire, interest, and
petence (von Krogh and Roos 1995), the concept learning capability” (p. 386). This suggests that, in
of competence couples practice (Bassellier et al. addition to developing competence merely through
2003) and action (Muffatto 1998) with this knowl- the reproduction of past competencies in a
edge component. Furthermore, Sandberg (2000) situated context, individuals are also purposive in
highlights that workers’ own conception of the work their competence development, motivated either
is central to our understanding of competence. by their own competence interests or organi-
zational competence needs. We label this form of
Emphasizing that competence is the enactment of competence competence-in-the-making.
knowledge, Muffatto (1998) suggests that com-
petence is an ongoing accomplishment (also see Informed by the literature reviewed thus far and by
Orlikowski 2002). It is not an object that either an Giddens’ (1984) structuration theory as a way of
individual or an organization owns, but rather a integrating the mutually dependent realm of
continuous process of production and reproduction organizational structure (macro level) and indivi-
(Scarbrough 1998). In this ongoing process, dual action (micro level), we now develop a model
competence plays a dualistic role, serving both as of competence in organizations (see Figure 1).
input to and output of competent action. Kim’s Given our research objective, namely the
(1993) model of learning is helpful in identifying the development of design principles for competence
various components of the competence develop- management technology, we chose Orlikowski’s
ment process at the level of the individual. Kim’s (1992; also, Orlikowski and Robey 1991) adap-
model is made up of two parts: (1) a dynamic tation of structuration theory for our conceptual
learning cycle consisting of the phases of infrastructure as it highlights the role of IT in the
experiential learning (Kolb 1984), i.e., experience, recursive, organizational structuring process.
reflection, abstraction and testing, and (2) memory,
a stock of conceptual frameworks and operational According to structuration theory (Giddens 1984),
routines. Memory is both the source and the the structural properties of social systems (the
destination of the learning process. structure level in Figure 1) are enacted through
recurrent human action and interaction (the
Viewing competence as ongoing accomplishment agency level in Figure 1). Such enactment is
and applying Kim’s model of learning to com- mediated through a number of elements (i.e.,
petence development, we can distinguish between facilities, norms, and interpretive schemes) that
different types of competence. In order to act both enable and constrain human action. Tech-
competently, individuals rely on their stock of com- nology embeds these mediating elements
petence, which is derived from past actions accu- (Orlikowski and Robey 1991). As individuals use
mulated over time. However, as with the use of technology and thereby draw on these mediating
knowledge (Stehr 1994), the use of extant compe- elements (shown as technology mediating be-
tence is not a mere transfer from stock to a speci- tween the structure and agency level in Figure 1),
fic situation or action context. Instead, applying they recursively produce and reproduce the social
stored competence implies a process of re-crea- structures that shape their action.
tion that transforms the competence taken from
stock. As such, the previously accumulated stock Thus, recurrent actions of organizational members
of competence is distinct from competence-in-use. draw not only on extant competence (competence-
in-stock) to generate new competencies, but also
Taking a life-cycle perspective of competence, the on a variety of assumptions, expectations, and
literature not only suggests a past (competence-in- norms embedded in the structural features of the
stock) and a present (competence-in-use) stage of organization, which include core competencies
competence, but also a future stage. For instance, and HR practices (arrow c), and CMS (arrow b).
Lawler and Ledford emphasize that “a critical By applying these mediating elements, organiza-
Structure
Organizational Core Competence
• Resources, learning, and routines
• HR orientation (skill-based approach)
a
b
Competent Action
Competence-
Agency
Competence- Competence-
in-Stock in-Use in-the-Making
a Technology as a product of human CMS as the result of the competent actions taken by system
action designers and developers
c Institutional conditions of interaction Structural features such as the organization’s core compe-
with technology tence (resources, learning, and routines), as well as con-
ceptual structures (e.g., a skill-based orientation in HR),
define, enable, and constrain individuals’ competent actions
tional participants create and re-create the core action researcher gains both theoretical and
competencies that characterize the organization’s practical knowledge about the phenomenon.
competitive advantage (arrows a and b).
In addition to generating knowledge through
This structurational perspective on organizational experimenting in a real-world setting, action
competence and its development informed our researchers regard change as an important out-
action research study, which was motivated by our come (Susman and Evered 1978). Merely
quest to develop and test CMS design principles. studying a real-world problem without assisting to
As our integrative model demonstrates, CMS form resolve or ameliorate it is perceived as unhelpful.
part of the mediating structure that facilitates the In other words, action researchers see it as their
smooth interaction between competencies at the responsibility to assist practitioners by not only
macro and micro levels of the organization. In developing but also applying knowledge
order to support organizational competence (Mathiassen 2002).
management in day-to-day action, the design of
CMS must appreciate the reciprocal relationship of Although there are a variety of action research
the three competence types (competence-in-stock, approaches available to IS researchers (Basker-
competence-in-use, competence-in-the-making) ville and Wood-Harper 1998), Susman and
and the organization’s core competencies. Evered’s canonical action research method is one
of the most widely adopted in the social sciences
(Davison et al. 2004). As a canon of action
research, the method formalizes the standards of
this iterative, rigorous, and collaborative research
Method process by describing it in terms of the following
five phases:
Action Research
• Diagnosing refers to the joint (researcher and
Given our objective of developing and testing
practitioner) identification of situated problems
design principles that render CMS supportive of
and their underlying causes. During this
knowledge-intensive organizations with a core phase, researchers and practitioners jointly
competence orientation, we selected action formulate a working hypothesis of the
research as our mode of inquiry. Action research research phenomenon to be used in the sub-
has been described as “a post-positivist social sequent phases of the action research cycle.
scientific research method, ideally suited to the
study of technology in its human context” • Action planning is the process of specifying
(Baskerville and Wood-Harper 1996, p. 235). the actions that can improve the problem
Even though there are other methods for studying situation.
technology in its “natural” context of everyday use,
action research distinguishes itself in that it is • Action taking refers to the implementation of
interventionist and dedicated to the development the intervention specified in the action
of knowledge useful to both research and practice planning phase.
(e.g., Baskerville and Wood-Harper 1996; Susman
and Evered 1978). • Evaluating entails the joint assessment of the
intervention by practitioners and researchers.
As an interventionist method, action research
allows the researcher to test a working hypothesis • Specifying learning denotes the ongoing
about the phenomenon of interest by implementing process of documenting and summing up the
and assessing change in a real-world setting. By learning outcomes of the action research
analyzing discrepancies between the hypothesized cycle. These learning outcomes should con-
and actual changes in the real-world setting or the stitute knowledge contributions to both theory
client-system infrastructure (Susman 1983), the and practice, but they are also recognized as
temporary understandings that serve as the identifying implications for other situations and
starting point for a new cycle of inquiry. research contexts.
vendors such as Prohunt, SAP, and Tieto Datema showed that they resulted in both anticipated and
to custom-built applications (Frontec’s Compass unanticipated consequences. A summary of our
and Guide’s Competence Marketplace). research project is presented in Table 1.
Research Process
First Action Research Cycle
In our quest to develop design principles that
would make these systems supportive of knowl- With our intent to identify design principles for
edge-intensive organizations that embraced a core CMS, we started our first action research cycle by
competence approach, we started with an investi- investigating our participating organizations’ earlier
gation into the strengths and weaknesses of the attempts at using IT to support competence
technologies upon which our participating organi- management. Their prior IT solutions included in-
zations had relied for managing competencies in house database applications, spreadsheets, and
the past (e.g., spreadsheets, Word documents, Word documents, and they had been only partially
and databases). Based on the shortcomings of successful. The competence descriptions pro-
these desktop-based solutions, we formulated two vided by these systems were inaccurate and
design principles. Since the objective of the first incomplete and therefore of little use in practice.
action research cycle was to work within the A key problem discussed during our workshops
confines of the CMS that our participating with the participating organizations was that the
organizations had selected, we translated these competence descriptions were not updated
design principles into interventions that guided the regularly and were therefore frequently obsolete.
configuration and implementation of CMS.
Another problem concerned inconsistencies in
We followed Susman and Evered’s cyclical action competence descriptions across organizational
research design, and completed the first of two units. While some managers only recorded com-
action research cycles in 4 months (July 1999 to pleted courses, others also included skills acquired
October 1999). During this time, we implemented from project activities. The most ambitious mana-
and assessed our recommendations in the parti- gers combined different information sources and
cipating organization. To our disappointment, the compiled relatively comprehensive competence
evaluation of our interventions highlighted a set of descriptions. For example, some managers at
barriers that hampered the adoption of the CMS as Guide documented their group members’ career
we had envisaged. ambitions.
With an understanding of these adoption barriers We traced the underlying cause of these problems
and a desire to find ways of overcoming them, we to the fragmented nature of the desktop appli-
embarked on a second action research cycle, a cations that the organizations were using. Our
26-month effort (November 1999 to December working hypothesis of the first action research
2001). Instead of working within the confines of cycle was that the problem of inaccurate and
the organizations’ CMS solutions, as we had in our incomplete competence data could be handled by
first action research cycle, we sought to develop using CMS (i.e., information systems specifically
and empirically test a set of design principles by designed to manage competencies in organiza-
implementing them in prototypes. Given the con- tions).
siderable effort required to develop, implement,
and evaluate CMS prototypes, we invited only two In view of the competence data problems related
of our six organizations, namely Volvo IT and to prior IT solutions, the research team and the
Guide, to participate in our second action research practitioners agreed that it was important to use
cycle. Our assessment of the four design prin- these experiences as guidance for configuring and
ciples on which we had built our two prototypes implementing CMS. Since we did not consider
CMS as turn-key solutions but rather technologies their commitment to and understanding of it (see
that needed to be integrated into a social system Lawler and Ledford 1992). In our role as action
of everyday work, we relied on a series of col- researchers, we therefore initiated and evaluated
laborative workshops (involving practitioners from competence data entry routines for such employee
all six organizations) to outline design principles involvement in all six participating organizations.
that incorporated the lessons learned from the
organizations’ earlier competence management Second, addressing the problem of inconsistencies
solutions. in competence descriptions, we agreed that CMS
should balance formal and informal indicators of
First, tackling the problem of irregularly updated competence in competence descriptions. We refer
competence descriptions, we agreed that user to this as the principle of balanced competence
control is an important prerequisite for keeping descriptions. While formal descriptions of compe-
competence data in CMS up-to-date. Thus, we tence refer to documented knowledge and skills
formulated the design principle of user control, acquired and rated through official channels (e.g.,
which specifies that knowledge workers whose attending training courses or earning a degree),
competencies were captured and stored in CMS informal descriptions of competence refer to on-
would have control over their information. Rather the-job experiences, interests, personal charac-
than having the responsible manager enter the teristics, and behaviors. This principle specifies
competence data, individual knowledge workers that the formal and informal aspects of compe-
should be able to update their own competence tence complement each other and together
information. We reasoned that employees’ active provide a more holistic representation of individual
involvement in managing CMS would increase and organizational competence. For instance, the
formal competence indicator that an individual unable to address a more fundamental problem.
speaks German at an expert level is enriched and We therefore embarked on a second action
contextualized by the additional description that research cycle in order to identify and test CMS
the person lived in Germany for many years.3 design principles outside the confines of existing
CMS.
Even though all of the participating organizations
recognized the value of representing both formal
and informal indicators of competence in their
CMS, they implemented our suggested interven- Second Action Research
tion to varying degrees. This was primarily due to
Cycle
the limitations imposed by the design of their CMS
packages. For instance, Prohunt and SAP R/3 did
In our second action research cycle, we sought to
not include free-text features, thus limiting the
develop and empirically test design principles by
organization’s ability to incorporate informal com-
implementing them in prototypes separate from,
petence descriptions. Therefore, only two organi-
but complementary to, the participating organi-
zations, Frontec and Guide, whose CMS could zations’ CMS. Only two of our research partner
accommodate additional competence descriptions, organizations, Guide and Volvo IT, participated in
implemented our first intervention. this action research cycle. We targeted these two
organizations because they had well-established
At each site, we evaluated the CMS in use. Our competence management practices covering both
evaluation highlighted a number of barriers to the strategies and technologies. In fact, this strength
adoption of the CMS (Lindgren and Henfridsson could be traced to the fact that Guide and Volvo IT
2002). For instance, users indicated that they did invested a larger portion of their revenues in
not want to disclose competencies that they were competence management than the other four
no longer interested in applying or developing. In organizations and as a consequence they had
other words, the CMS stored knowledge workers’ most HR staff dedicated to develop and maintain
extant competencies, which then served as the these practices on an everyday basis. They were
basis for project assignments. Thus, the CMS thus better prepared for engaging in a new action
tended to reproduce competencies (e.g., once a C research cycle requiring intense researcher-
programmer, always a C programmer), causing practitioner collaboration.
users to misrepresent their knowledge and skill.
In addition, we had good access to these organi-
Thus, despite our attempts to improve data quality zations. The general manager of the Viktoria Insti-
by implementing CMS guided by the two design tute had previously served as the chief knowledge
principles, the problems remained. Our working officer at Guide and our contact at Volvo IT was a
hypothesis was not supported. The CMS imple- senior information architect at Volvo IT’s Web
mented and used by the six organizations were Program Center, who was enrolled in the Viktoria
unable to sufficiently solve the problems of Institute’s industrial Ph.D. program. These rela-
inaccurate and incomplete competence data. We tionships increased the likelihood that a second
concluded that we needed deeper and richer and rather lengthy action research cycle could be
insight into what caused these CMS adoption completed successfully.
problems. We suspected that the assumptions
about competence embedded in the CMS had
hampered the success of our interventions. Thus
Diagnosing
our two design principles were isolated solutions
CMS at Guide
Guide had approximately 800 employees at 10 agement and organizational planning, (2) organi-
offices located in 3 countries. Since 1988, Guide zational development, (3) technology, tools, and
had invested 15 percent of its annual revenues systems, and (4) methods and models. The
into competence development. Of its many system stored an individual’s competencies in
competence-enhancing activities, Guide con- terms of four levels: beginner, some knowledge,
sidered the development of individuals’ compe- experienced, and expert (see Figure 2).
tencies through project assignments the most
important. Guide’s business and group managers
thus endeavored to staff projects in ways that both CMS at Volvo IT
satisfied a customer’s need and the individual
knowledge worker’s competence development With offices in Belgium, Brazil, Great Britain,
objectives. Malaysia, Sweden, and the United States, Volvo IT
is the Volvo Group’s resource and expertise center
In 1999, Guide implemented Competence Market- for IT. At the time of our study, Volvo IT had
place, a system that was intended to facilitate staff approximately 2,500 employees. Some 1,400 of
allocation and competence management. At the these worked in Sweden, with roughly 900 in
heart of the system was a database storing Göteborg, Volvo IT’s global headquarters. As do
descriptions of staff competence levels in areas many large organizations, Volvo IT recognized the
such as client-server technology and project advantage of knowing who within the organization
management. The system organized the different held what expertise. To help manage its network
competencies into four major categories: (1) man- of globally dispersed knowledge workers, Volvo IT
decided to implement a CMS in June 1999. They taking that involved the individual knowledge
chose TP/HR, an off-the-shelf, modular application worker and his/her manager.
developed by Tieto Datema (Sweden). The focus
of our research was the education/competence
module. Initial Assessment of CMS
knowledge and skills to fulfill a job without knowl- about the competence-reproducing mechanism of
edge sharing. In the CMS we studied, individual CMS. These concerns threatened to undermine
users could only see their own competence the accuracy with which users represented their
descriptions, not those of others. The restricted competence:
access to competence information was intended to
limit internal recruiting, but it hampered knowledge In order to avoid assignments in certain
sharing and the ability to leverage internal knowl- areas, I hide competencies that I don’t
edge resources. A project manager at Guide in want to apply.… I’m capable of working in
Göteborg put it as follows: a COBOL project, but I don’t want to do
it. If I express my competence in
Maybe it should be possible to connect COBOL there is an obvious risk that I
this group of people with similar interest have to take on assignments where this
profiles in some way. For instance, mark type of programming skill is needed. So
here [in the system] that I’m a member of my group leader advised me to leave out
this network. Then I have more search COBOL when I filled in my competence
paths and this would make it easier to description.
find knowledgeable colleagues. At pre-
sent, there is no interactive forum for This concern was underscored by an account
exchanging opinions and competence manager at Guide, who highlighted that the
sharing. It is important to make it easier omission of knowledge workers’ interests and
to initiate a dialogue. career objectives in the CMS limited its value:
One of the HR managers at Volvo IT identified The major disadvantage of the system is
another challenge with regard to the CMS’s focus that it is not possible to make compe-
on past competence. He pointed out that in a tence analyses of teams and groups dif-
dynamic and fast-paced environment like IT, main- ferent sizes. The system supports ana-
taining accurate representations of competence lyses on the individual level in an excel-
required a CMS capable of dealing with skills and lent way. But we must be able to use the
knowledge as they evolve: system in order to form a project team
and analyze the total competence level.
Earlier it was easier since there were few So the system needs features facilitating
programming languages. Now the devel- evaluation of groups in different sizes.
opment is so fast. Yes, there are the Moreover, the system should also be
fourth, fifth, and sixth generation. And more flexible with regard to analyses of
individuals change as well…their com- competence status at a certain point in
petencies change over time. time and competence status changes
over time.
Third, the CMS were rigid in their reporting of
Based on our diagnosis of the CMS used in Guide
organizational competencies. The parameters for
and Volvo IT respectively, we concluded that the
competence analyses were limited (e.g., for
problems associated with their extant systems
predefined group sizes and at system-stipulated
could be traced back to a system design that
points of time). Since ad hoc reporting of com-
embraced a job-based approach to competence
petencies for gap analyses, for instance, is
management. Applying Lawler’s (1994) theorizing
increasingly important in volatile business environ-
that a skill-based approach is more conducive to
ments, the rigidity of the reporting functions limited
management of organizational competencies than
the usefulness of the CMS. One project leader at
a job-based approach, we formulated the following
Guide pointed out that the Competence Market-
working hypothesis: CMS that embody the skills-
place was deficient in its support of strategic based paradigm are more effective in knowledge-
planning because reporting focused on micro level intensive organizations embracing a core com-
competencies only: petence orientation than are CMS reflective of the
job-based paradigm.
It’s not sufficient to know the employees’
competencies. You must be able to
manage those competencies in a stra-
tegic way, but [the system] does not Action Planning
contain a complete package. Our inten-
tion is to categorize the activity with Guided by our working hypothesis, we set out to
regard to strategic goals and critical com- develop design principles that would improve the
petencies. At present, however, we can usefulness of CMS in our two participating organi-
merely visualize competencies on an zations. In formulating these design principles, a
individual level by using the system. But process that was completed in collaboration with
we want to manage competence on a, for practitioners from Guide and Volvo IT, we relied
instance, departmental level. When ana- not only on the three problems that we had
lyzing different groups of employees, we identified in the diagnosis phase, but also on the
need [the system] as a management tool three types of competence that we had identified
for the activity. in our integrated model of organizational compe-
tence: competence-in-stock, competence-in-use,
This shortcoming was echoed by a HR manager in and competence-in-the-making. Thus, we devel-
Volvo IT: oped the following four design principles:
• The principle of flexible reporting: CMS As its name suggests, Competence Visualizer
should support ad hoc analyses of the relied on graphical modes of presentation to allow
organization’s competencies, both with regard users to visualize Guide’s competencies-in-stock
to different units of analysis (e.g., individual or and competence interests. Users could view com-
group) and time frames (e.g., from-to dates).
petencies at the individual, group, or organizational
This principle addresses the rigidity in CMS’
level. Furthermore, competencies could be shown
reporting functionality. By providing mana-
as a snapshot view (Figure 4), i.e., single point in
gers with flexible reporting, CMS support
time, as well as across a user-defined time period
strategic competence development by influ-
(Figure 5). These graphical representations were
encing competence-in-the-making.
expected to help users identify patterns and trends
in organizational competence needs.
These four design principles were then applied in
the design of two CMS prototypes. These were
In order to present competence-in-stock and
Guide’s Competence Visualizer and Volvo IT’s
competence interests at the individual and aggre-
Volvo Information Portal (VIP).
gated group or organizational level, Competence
Visualizer assigned numeric values to the descrip-
tive competence ratings in Competence Market-
Action Taking place. A beginner rating was assigned a value of
0.25; some knowledge was 0.5; experienced was
The CMS Prototype at Guide a 1.0; and expert was 1.25. To satisfy the design
principle of flexible reporting, Competence
In collaboration with Guide practitioners and three Visualizer displayed the competencies not only in
M.Sc. students, the first author developed the graphical, but also in tabular form.
In April 2000, the Competence Visualizer was Based on his knowledge on search engines and
demonstrated to Guide users. Our original intent recommender systems, the Volvo Information
was to implement the prototype in Guide so that Portal (VIP) system was developed as a comple-
users could evaluate our system after using it in ment to TP/HR. VIP captured knowledge workers’
their everyday work, but we were unable to do so competence interests in real time and made
due to unforeseeable organizational changes. In competence-in-the-making more transparent. Re-
early 2000, Guide merged with the Internet consul- source constraints did not allow us to implement all
tancy organization Framfab. This meant that four design principles. The VIP prototype em-
Guide’s management was preoccupied with the re- bodied the principles of transparency, real-time
organization following the merger, which made it capture, and interest integration. In this way, each
difficult to move ahead on our implementation of our design principles was covered by at least
plans, particularly as all internal development was one prototype.
put on hold. We thus opted for an alternative
evaluation strategy, which involved workshops and VIP was built on Autonomy’s AgentWare platform,
focus groups with intended Guide users. We a commercially available tool that uses neural
describe our evaluation in more detail later. networks and advanced pattern-matching tech-
niques to find similarities in textual data. VIP
allowed the users to define intelligent agents that
The CMS Prototype at Volvo IT searched an index database for intranet docu-
ments matching the user’s interests. By defining
At the time of this research, one of the members of one or more agents, VIP users were thus able to
our research team was a senior information monitor the corporate intranet for items that
architect at Volvo IT’s Web Program Center. matched their interests. The users defined their
interests in a free-text natural language format abling knowledge sharing and collaboration across
from which the system then created an internal Volvo IT’s many offices. When invoking this fea-
digital representation. ture, the interest profiles embedded in the user’s
agents were matched with the profiles of all other
The search results from each agent were dis- agents. Matches were displayed, listing the name
played in a simple list similar to those generated and contact information of the identified users.
by search engines. By clicking on the hyperlinks Finally, the VIP prototype supported searches for
generated by the agent, users could retrieve the users with specific interests. VIP users could enter
documents matching their interests. When users a search for an interest in natural-language format
had read and identified one or more of the and VIP would return all users whose agents
returned documents as relevant, they could contained the specified interest. Figure 6 presents
provide the agent with explicit feedback through a VIP’s response to such a search.
retrain feature, which adjusted the agent’s search
criteria based on the characteristics of the In April 2000, VIP was implemented on the Volvo
retrieved document. IT intranet and made available to all intranet users
in Göteborg as part of their everyday work. Fifty
The VIP prototype also provided a community fea- people tried out the prototype when it was first
ture, which allowed users to find other users with introduced and there were twenty users with active
an interest profile similar to their own, thus en- agents at the end of the evaluation period.
Moreover, it is easier to understand what organizations have to explicate and publish their
we mean when we discuss employees’ competence needs in order to stimulate their
opportunities in relation to the organi- development.
zation’s activities.
It’s quite easy to appreciate the value of
Additionally, the interest-integration design prin- this system. It’s really interesting to
ciple was seen as addressing the data accuracy check out the competencies within the
problem that plagued the original CMS. A group organization on a regular basis; be able
manager at Guide’s office in Göteborg used the to find out our strengths and weak-
following example to explain the benefits of nesses. And the interest module indi-
incorporating interest competencies into a CMS: cates what work tasks the employees are
interested in working with. But, then it is
As an example: I have been working our duty to create areas of interest. That
with FORTRAN for 10 years; I’m a really is basically the best approach. In case
good FORTRAN programmer, but I don’t we need 15 experienced project leaders
want to work with it. Then I have re- within the next 12 months, we have to
moved that competence in the system introduce that specific interest to our em-
[Competence Marketplace]. For this rea- ployees. In this way you are able to
son, this interest dimension is good: I’m affect both of the curves [of the Compe-
good at it and I want to work with it, I’m tence Visualizer system]; it’s simply a
good at it, but I’m not interested in matter of marketing.
working with it, and my knowledge is
limited at the moment, but I’m eager to This interview quote highlights that organizational
learn more. competence needs and individual competence
interests cannot be treated independently in a
As this group manager highlighted, Competence CMS that embodies the skill-based paradigm.
Visualizer allowed employees to describe their Instead, managers have to work through indivi-
competence in more comprehensive ways. They duals’ competence interests in order to meet the
were able to express both existing competencies competence development needs of the organiza-
(competence-in-stock) and individual competence tion. This approach to competence development
interests, which drive competence-in-the-making. might prove to be particularly challenging in
Thus, competence interests were not represented situations where employees’ interests are at odds
at the expense of existing competencies, allowing with the organization’s (and its customers’) needs.
Guide to generate a more accurate picture of its
existing and emerging competencies, and to
compare them with its strategic core competence Evaluating at Volvo IT
goals.
After implementing the VIP prototype on Volvo IT’s
In addition to these anticipated consequences of intranet, we initiated a 10-week evaluation with a
our prototype, there were some unanticipated workshop at Volvo IT’s headquarters in Göteborg.
ones. For instance, the objective of our flexible At this workshop, which was attended by depart-
reporting design principle was to highlight compe- ment managers, HR managers, project managers,
tence gaps between existing competencies, system developers, and technicians, we intro-
individuals’ competence interests, and the organi- duced VIP as a complement to TP/HR and des-
zation’s competence needs. As the following inter- cribed its basic features. The workshop partici-
view quote from a Guide group manager high- pants (approximately 30) were then encouraged to
lights, the implication of adopting a skill-based use VIP in their everyday work. After the 10-week
approach to HR management and taking indivi- evaluation period, during which we collected
duals’ competence interests seriously, is that observational data, we interviewed 16 VIP users to
assess their perceptions of VIP, e.g., its useful- I see [the find users with similar interest
ness in their work and its value with regard to feature] as a very useful feature; as an
competence management. enabler for building networks. It is
interesting to be able to find colleagues
Our assessment revealed that VIP generated both who are interested in the same things.
anticipated and unanticipated consequences. With Because our main problem here is that
regard to anticipated consequences, a number of there are people working with similar
the practitioners remarked on the advantages of things everywhere and you don’t really
VIP’s ability to capture competence and compe- find them.
tence interests in real-time. They noted that the
original CMS contained information on neither In addition to these anticipated consequences of
competence-in-use nor competence-in-the- our design principles as implemented in VIP, our
making, whereas the prototype did. One HR man- evaluation also identified a number of unanti-
ager felt that the VIP prototype represented a “next cipated consequences. For instance, the principle
step” in competence management as it captured of transparency was expected to increase
the tasks for which knowledge workers used their knowledge sharing in the organization. However,
competencies: the feedback from one Volvo IT project manager
highlights that this is only true if the competence a
TP/HR is a lot about order and being in user seeks exists within the CMS:
control of the situation, i.e., to know what
we have and the level of education of our Then you suddenly realize that the topics
employees.…Then this prototype is
searched for are of no interest to others.
something else. It is what people do on
Naturally it is valuable to know that there
an everyday basis. It is what they use
are no organizational members but me
their skills for. It is sort of the next step.
who are interested in these kinds of
issues. Resulting from this, you under-
Several practitioners indicated that the interest-
stand clearly that there is no point to walk
integration design principle implied a future orien-
all over the office trying to find relevant
tation for CMS and they believed that this would
people to talk to.
stimulate competence development. A technology
analyst expressed this in the following way:
Thus, in the event that a competence cannot be
located in the CMS, the system potentially created
Interest is tremendously important for the
disincentives for organizational members to
development of competence. When you
appreciate something and find it chal- contact each other. Furthermore, the quote above
lenging, the fundamental conditions for illustrates how the system gives users a perspec-
learning are in place.…So, to cultivate tive of themselves in relation to the rest of the
learning, it is important to provide positive organization. It is unclear what emotional reac-
tools like VIP; tools that cultivate a tions people have when they cannot find anybody
positive spirit increase the competence, in the organization that either shares their interest
the interest, the speed, and the quality. or possesses the interest or competence that they
are seeking. While some may view this as an
A number of VIP users indicated that the principle opportunity to develop a niche in which they can
of transparency, which allowed users to identify differentiate themselves, others may view it as an
colleagues based on mutual or specific interests, indication of their own isolation.
rendered VIP a valuable tool for the creation of
knowledge sharing networks or communities. As Another unanticipated consequence of the proto-
the CMS project manager highlighted, the trans- type was the increased vulnerability that some
parency principle created the conditions for effec- knowledge workers perceived. Discussing the VIP
tive competence management in Volvo IT: system, one process developer suggested that the
system’s manipulation of personal information, you with their problems and questions,
such as individuals’ interests, made some users resulting in a heavier workload for you as
anxious: an individual.
Typically individuals tend to be a little bit Our prototype evaluation also revealed that there
frightened when they are not in control. were few incentives for managers to promote the
They are uncertain about how the system use of the CMS. Due to a lack of metrics and per-
works, how detailed they should be, and formance criteria related to the tracking of compe-
how much personal information they are tence-in-use and competence-in-the-making, a
supposed to provide. To me it seems member of the HR staff at Volvo IT highlighted the
that some people feel anxious about the ambivalence managers felt toward our prototype:
role of the technology.
TP/HR is much more about structure,
Since interests are very personal and frequently order, and control, and basically the
related to the tentative exploration of a topic (as system is intended to support the HR
opposed to a firm commitment to a skill or function with statistical analyses of the
knowledge, as might be inferred from earning a competence status of the employees.
degree or attending a course), it is not surprising However, the VIP system deals with
that VIP users felt protective of this private knowledge too, but in a different way.
information and concerned about what inferences Fundamentally the system represents
were made from it. Furthermore, interests are an knowledge quite differently, and I’m
expression of the self. Thus coupling the principle positive about this system’s ability to
of interest-integration with the principle of contribute to a lot of things….Although
transparency and real-time capture was rather I’m pretty confident that you can’t
problematic. Indeed, the combination of these measure everything, the problem with
three design principles created a system that had this system is really how to measure it.
the potential of infringing on a user’s privacy. Because the result generated by the
system is not measurable, however, the
In our first action research cycle, we noted that problem is that nobody can take the
competence data was frequently inaccurate credit. If you can’t gain credit, you are
because individuals did not want to be assigned to not willing to deal with it.
projects based on their existing competencies
(competence-in-stock) rather than their compe- In sum, our evaluation of the two prototypes
tence interests. The principle of interest integra- highlighted that our design principles generated
tion was expected to address this issue by allowing not only anticipated but also unanticipated conse-
knowledge workers to specify their competence quences. In light of these findings, we revisited
interests separately from their competence-in- and refined our initial design principles. We
stock. However, as the following quote from an discuss these in the next section.
interview with a system developer at Volvo IT
indicates, our interest integration principle was
unable to provide sufficient incentives for indivi-
duals to report their competence truthfully: Discussion
Sometimes people choose to hide their Prior research on organizational competence and
interests. The main reason for this is its management highlights the importance of
basically that people want to remain aligning an organization’s core competence orien-
anonymous. In case you express your tation with HR practices and systems that support
personal interests openly there is an the development of organizational competencies
obvious risk that people will approach (e.g., Muffatto 1998; Nordhaug 1998; Rothwell and
Lindholm 1999; Simpson 2002). However, there is and unanticipated consequences. Reflecting on
no prior research on the role of IT in supporting the these results and the entire 30-month action
management of organizational competence. Thus, research project, we realized that we had to
our research objective was to develop and test address the unanticipated consequences of our
design principles for CMS so that these systems design principles. In particular, the design prin-
support knowledge-intensive organizations ciples did not embed the user control idea pursued
embracing a core competence approach. Given in the first action research cycle. Thus, we now
that action research is a strategy particularly revise and refine them based on our assessments
appropriate for the development of design of the prototypes. As our research did not reveal
principles (Walls et al. 1992), we conducted an unanticipated consequences stemming from the
action research study consisting of two action principle of flexible reporting, which essentially
research cycles (Susman and Evered 1978). incorporated the principle of user control, we
revised only the first three design principles.
In our first action research cycle, we identified two
design principles intended to improve the quality of
the organization’s competence information.
Working within the confines of the CMS that the Revised Principle of Transparency
participating organizations had selected, our
interventions—to enhance formal competence The principle of transparency specifies that com-
descriptions with informal ones and to grant users petence should be made visible and accessible
control over their competence descriptions—were throughout the organization. This design principle
implemented to varying degrees. Our evaluation was motivated by the desire to leverage infor-
of the two interventions revealed a number of mation about individuals’ competencies in the
barriers that hampered CMS use in the parti- building of knowledge sharing networks and
cipating organizations. For example, we learned communities. By blurring the distinctions between
that CMS reproduced extant competencies. This business units in this way, the conditions for core
meant that knowledge workers interested in competence development and more flexible,
developing new skills were inclined to use the emergent organizational structures are created.
CMS in inappropriate ways by, for instance, failing
to disclose all of their competencies. These Even though our evaluation of this principle
insights provided the impetus for the second cycle
(implemented in VIP) suggested that transparency
of action research.
supports knowledge sharing and organizational
learning, it also highlighted a number of unantici-
In the second cycle, we relied on further analyses
pated consequences. These arose out of com-
of two CMS, Guide’s Competence Marketplace
bining the principle of transparency with the prin-
and Volvo IT’s TP/HR, as well as Lawler and
ciple of interest-integration. Users expressed a
Ledford’s (1992) argument that the job paradigm
sense of vulnerability as information about their
is at odds with the core competence orientation of
contemporary, knowledge-intensive organizations, interests, which are frequently more tentative
to develop four design principles for CMS: the expressions of the self than are competencies
principles of transparency, real-time capture, acquired via formally recognized means (e.g.,
interest integration, and flexible reporting. These courses), became widely available. Related to this
design principles embody a skill-based approach, is users’ concerns that the public display of their
which Lawler and Ledford considered more appro- competence (both competence-in-use and com-
priate for competence-oriented organizations. We petence-in-the-making) would lead to increased
subsequently implemented the four principles in workloads. Both of these conditions are likely to
two prototypes: Competence Visualizer and VIP. compromise the quality of competence data as
knowledge workers either fail to disclose compe-
Our evaluation of the prototypes in Guide and tence information in the CMS or misrepresent their
Volvo IT respectively revealed both anticipated competencies.
Another unanticipated consequence arises when stood in terms of the users’ uncertainty about how
there are gaps in the organization’s competence. the system worked and how the information
As the transparency principle makes these gaps contained in it would be used.
visible to individual knowledge workers, CMS may
impede the development of knowledge sharing To address this problem, we suggest that the real-
networks as competence gaps suggest that time capture principle be refined to read “real-time
contacting others is a waste of time. capture with feedback loop.” In short, users must
be provided with a clear understanding of how the
To address individual users’ concerns about too system handles the competence-in-use and
much visibility, we suggest that the transparency competence-in-the-making data it gathers. Users
principle be renamed “user-controlled trans- should also not only have access to the system’s
parency.” In other words, the control over which representation of their competence data, but also
competence data is publicly displayed should rest the ability to amend it. For instance, the search
with the individual knowledge worker. In this way, criteria contained in an agent (some of which are
individual users can control the presentation of based on the user’s search criteria, and some of
their competencies and the amount of personal which are derived through retraining) should be
data that is disclosed about them. For instance, in made accessible to users. Finally, users should
the case of VIP, users should be able to make be able to edit their implicitly derived competence
agents public or private. A public agent would be profiles. By combining implicit and explicit profiles
searchable by other users’ agents, while a private of an individual’s competencies, competence data
agent would not. is generated in an emergent, in-use fashion with-
out users relinquishing control over the system’s
Additionally, organizations could address the representation of their interests and competencies.
unanticipated consequences of the transparency
principle with management policies. By making By providing such a feedback loop, the ambiguity
knowledge sharing an activity on which the associated with real-time capture and, as a
performance of individual knowledge workers is consequence, users’ perceived anxiety should be
measured, users might see transparency as an reduced. Furthermore, users’ periodic review of
opportunity to market their competence rather than the implicit, use-derived profile in the CMS might
as a threat. Indeed, as its name implies, Guide’s also serve as a basis for individuals’ reflection on
Competence Marketplace embraced this notion of and learning about their competence and compe-
competence marketing. tence development.
formally recognized competence. This design incorporates both the speed with which the
principle was motivated by the desire to enrich the knowledge workers reporting to them reach the
definitions of competence with interests as an competence goals set by the organization and the
indication of the skills and knowledge that indivi- degree to which this competence development is
duals are motivated to develop. Thus, this consistent with the competence interests of
principle supports competence-in-the-making. individuals. Alternatively, a metric representing an
overview of the diversity of interests that an
Even though our evaluation of the interest organizational unit encompasses can work as an
integration principle indicated that this principle led indicator of the unit’s flexibility and responsiveness
to system features that users valued, especially in to change. In this way, individual competence
the Competence Visualizer system, it was also interests become highlighted and can be utilized
associated with a number of unanticipated conse- for leveraging the core competencies needed for
quences. Specifically, our evaluation of the developing customer value.
Competence Visualizer at Guide highlighted that
the combination of interest integration and trans-
parency created conditions in which management
will have to market the organization’s competence
needs to its employees so as to generate the Evaluating Our Action
requisite interest in the development of the neces- Research
sary organizational competencies. In situations
where employees’ interests are at odds with the In this section we assess how our action research
organization’s (and its customers’) needs, this compares to the principles of canonical action
approach to competence management promises research (Davison et al. 2004). Our evaluation
to be particularly challenging. Furthermore, our shows that our research satisfies all of the criteria
evaluation of VIP at Volvo IT highlighted that there put forth by Davison et al. (see Table 2 for a
were few incentives for managers to promote the summary). In the ensuing section, we describe
use of the CMS because performance metrics our assessment of our study.
related to knowledge workers’ interests were non-
existent. The principle of researcher-client agreement
highlights the importance of facilitating trust-
We propose a revision of the interest integration building and knowledge between researchers and
principle, which we label “multi-perspective interest practitioners (clients) (Davison et al. 2004). In
integration.” This principle suggests that it is not addition, it stresses the need for developing a
only the individual’s competence interests that mutual understanding of the scope, focus, and
should be taken into consideration in CMS, but mode of inquiry of the research project. In our
also the organization’s, which are assumed to be project, a third party, the research funding agency
reflective of its customers’ competence needs. In (VINNOVA) played an important role in outlining
this way, the potential tension between individuals’ the terms for each organization’s participation.
interests and the organization’s needs can be VINNOVA required that each participating organi-
addressed, as the organization’s strategic direction zation sign an agreement specifying resource
(i.e., the core competencies that it seeks to allocations and regulating the potential commercial
develop) is made explicit to the system’s users. utilization of the research results. We believe that
this agreement helped foster a working environ-
Our multi-perspective interest integration design ment where opportunism, which is counter-
principle also provides the foundation for devel- productive with respect to learning, was minimized.
oping metrics and evaluation criteria related to For instance, organizations such as Frontec,
competence-in-the-making. For instance, mana- Guide, and Volvo IT were willing to participate in
gers can be evaluated on the basis of a metric that the project and contribute resources to it even
though they competed with each other.4 Thus able research contributions. Indeed, it triggered
VINNOVA provided a control structure through the development of an integrative model of com-
which some of the tensions associated with the petence including a competence typology, and this
divergent interests involved in action research provided the basis for further learning and action.
could be handled (Avison et al. 2001; Mathiassen The synthesis conducted in this paper suggests
2002). the need of additional action research cycles to
test the revised design principles. At the end of
The principle of the cyclical process model asserts our research, there were discussions about a new
that progressing through all five action research project that would build on the lessons learned in
phases in a sequential manner is important to the first action research cycle. However, no follow-
ensure that the action research is conducted on project was initiated once the research funding
systematically and rigorously (Davison et al. 2004). ended. Nevertheless, the first author of this paper
While a single research cycle is sufficient in some continued his relationship with Volvo IT as a
cases, additional iterations through the five phases consultant.
can often yield additional learning. We applied
Susman and Evered’s (1978) action research The principle of theory highlights the importance of
cycle model composed of diagnosing, action using theory to guide the research activity and
planning, action taking, evaluation, and reflective relate the findings to extant theory (Davison et al.
learning. We relied on the first action research 2004). The development of the integrative model
cycle to help us define the research problem and of competence provided the central theoretical
our interventions in an emergent way. With the scaffolding for our research, enabling us to
lack of empirical support for our initial working synthesize findings from macro and micro level
hypothesis, we embarked a second action competence literatures. The model and its compe-
research cycle, which was more focused on the tence typology guided the analysis of CMS in use,
key problems of CMS design and use. the development of design principles, the inter-
ventions, and the evaluation of the prototypes.
Using this model, it became possible to draw
Our application of the canonical action research
relevant insights from our action research project
method, including a sequential use of the action
for HR researchers working with a core compe-
research phases over two full iterative cycles,
tence perspective. The model also guided our
helped us specify and re-specify the research
ongoing collaboration with practitioners in that it
problem based on our own active involvement with
pinpointed the direction for designing CMS that
it. For instance, the collaborative workshops, the
cater to the three different competence types and
jointly specified design principles, and the evalua-
that are reflective of skill-based assumptions.
tions of our interventions in the first action
research cycle deepened our understanding of the
The principle of change through action stipulates
competence data problem over the different cycle
that interventions appropriate to the problem and
stages. The successive learning eventually led to
the client organizations should be designed and
the recognition that we needed to explore our
implemented (Davison et al. 2004). This principle
design principles for CMS outside of the confines also concerns itself with the degree of practitioner
of the existing systems. involvement throughout the research process. In
our case, the practitioner involvement was high
Stimulated by the cyclical process model, this re- throughout the project. This translated into both
specification was a key element in our efforts to opportunities and challenges for our research. For
develop theoretically sound and practically valu- example, the second action research cycle
involved intense researcher collaboration with
Guide and Volvo IT, thus situating prototype
4
Even though Volvo IT was a division of the Volvo development and evaluation in a real-life, everyday
Group, it competed with IT consulting firms such as context. As such, the solution we implemented
Guide and Frontec because Volvo Group’s business
units were free to source services from external vendors. promised to be appropriate to the problem.
At the same time, high practitioner involvement sible to practitioners and practice has to be made
also requires a high degree of sensitivity toward accessible to academic theory. In this regard, we
the demands that everyday work places on the found that our prototypes played an invaluable role
participants. For example, Guide’s merger with as boundary objects; they made theoretical
Framfab meant that the resources for and the concepts more tangible and comprehensible. For
interest in our Competence Visualizer prototype instance, the VIP prototype at Volvo IT afforded
declined. Thus, to complete the project, we relied our practitioner collaborators sensemaking oppor-
on evaluation strategies involving workshops and tunities, especially with regard to their current
focus groups to simulate as realistic an environ- CMS’ support of the different types of competence.
ment as possible for the intended Guide users. The prototypes also facilitated discussion about
Nevertheless, we do not consider these evaluation the distinction between skill-based and job-based
methods as authentic as the ones we were able to assumptions. We thus found the prototypes
use in Volvo IT. effective in communicating, collaboratively devel-
oping and evaluating our theoretical ideas.
The principle of learning through reflection high-
lights the importance of drawing insights from the As our project also illustrates, however, the use of
research and of identifying implications for other prototypes can divert the attention from the original
situations and contexts. Apart from providing research agenda. In our case this was due to the
progress reports and research papers throughout considerable effort—both technical and social—
the stages of an action research project, this needed to develop and implement prototypes
principle highlights that it is central to investigate within the realm of practice. Our interventions in
the magnitude and sustainability of the research the second action research cycle were based on
efforts over time. Even though it is difficult to the disappointing outcome of the first cycle and
measure the extent to which our research informed by skill-based assumptions. However, in
outcomes resulted in further action in practice, it our quest to develop CMS that support compe-
should be noted that Volvo IT initiated a worldwide tence-in-use and competence-in-the-making, we
project to outline its competence management failed to consider the increasing exposure of the
strategy. In this project, personal interest profiles user following the transparency and the real-time
were incorporated into individuals’ competence capture principles. In this regard, we did not pay
descriptions. We contend that this feature can be enough attention to our initial user control principle
traced back to the interest dimension of our in the development of the design principles in our
integrative model, which was also incorporated into second action research cycle.
the VIP prototype. Guide practitioners also
recognized the learning implications of our project. The strengths and weaknesses of our use of
For instance, Guide developed a new version of prototypes might provide input to improve the
Competence Marketplace that incorporated canonical action research method for design-
features of Competence Visualizer. This system oriented IS research. In our case, it would have
became part of Guide’s offering to their customers. been useful to have guidelines to help us ensure
Moreover, we believe that our results were that the prototypes contributed to both knowledge
important in terms of its implications for other types generation and the implementation of change.
of HR systems. In our collaboration with Volvo IT, The development of design principles is not simply
for example, the implications of our research were about operationalizing theory into neat principles
discussed as part of career management, for normative action, but it involves also an
recruitment and selection, and training sessions. assessment of available tools and situated condi-
tions such that these principles render to tech-
As action researchers we need to balance the nically and organizationally feasible solutions. As
theoretical demands of academia with the suggested above, the complexity and magnitude of
problem-solving demands of practitioners. In prac- the technical challenges made us lose sight of
tice, this means that theory has to be made acces- users’ needs and concerns. This was not the case
in the first cycle in which we could focus on the we identify a competence typology including three
user and organizational requirements with respect types of competence: competence-in-stock, com-
to competence management. petence-in-use, and competence-in-the-making.
Our findings suggest that for CMS to be suc-
In design-oriented action research, the practical cessful, they will have to support all three of these
challenges of handling the socio-technical chal- competence types.
lenges of prototype deployment must be balanced
with the scientific process of operationalizing Our four CMS design principles extend earlier
theory into design principles (action planning), work on HR management by empirically demon-
converting design principles into IT artifacts (action strating that an infrastructure reflective of the job-
taking), and inferring use data back to theory based paradigm present problems for competence
(evaluating/specifying learning). Neither canonical management in contemporary, knowledge-inten-
action research (Davison et al. 2004; Susman and sive organizations (Lawler 1994; Lawler and
Evered 1978) nor the newly developed framework Ledford 1992). As our assessment of CMS in
for the dual imperatives of action research (McKay organizations during our first action research cycle
and Marshall 2001) addresses this dimension of demonstrated, systems embodying the job-based
action research in IS. paradigm with its predefined job descriptions and
taxonomies of formal competence create barriers
to the adoption and appropriate use of CMS.
Furthermore, the findings of our second action
Conclusion research cycle suggest that HR infrastructures that
embrace the skill-based approach must be
Synthesizing the insights gained from a 30-month accompanied by sufficient user control over the
action research study involving numerous data information that represents the user’s competence.
collection strategies and interventions such as Our findings highlight the interdependence be-
prototypes, our study has generated an integrative tween organizational needs for competence and
model of competence in organizations, a compe- individuals’ competence interests. Organizations
tence typology, and a set of design principles for adopting a skill-based approach will find that they
CMS. Our integrative model contributes to the have to market their competence needs to their
body of literature seeking to align the macro and workers in order to stimulate individuals’ interests
micro levels of competence (e.g., Muffatto 1998; in a particular competence.
Nordhaug 1998; Rothwell and Lindholm 1999;
Simpson 2002). It provides a conceptual frame- These insights elaborate on Lawler and Ledford’s
work for incorporating IT into our understanding of assertions about the skill-based approach by
competence management in organizations. providing an empirically grounded picture of it.
Informed by Orlikowski’s (1992) structurational Furthermore, through our revised design prin-
model of technology, we conceptualize IT as a ciples, we identify ways of addressing the prob-
mediating element in the integration and alignment lems that are likely to arise in the application of the
of competent action at the micro level and core skill-based approach. For instance, our principle
competencies at the macro level. Given the of multi-perspective interest-integration seeks to
increasingly important role of IT in competence resolve the potential conflict between organiza-
management (Alavi and Leidner 2001; Andreu and tions’ competence needs and individual knowledge
Ciborra 1996; Davenport and Prusak 1998), the workers’ competence interests. Our revised
literature on aligning macro and micro level design principles, therefore, serve not only as the
competence is incomplete if it fails to make IT an basis for future academic research on design
integral part of its theorizing. Our integrative principles, but also as guidelines for practitioners
model further extends the conceptualization of seeking to design and implement CMS in knowl-
competence at the level of the individual worker. edge-intensive organizations that are pursuing a
Relying on learning theory (Kim, 1993; Kolb 1984), core competence orientation.
First, our research has highlighted that the degree functionality and its development methodology
of competence transparency in a CMS is a (Markus et al. 2002; Walls et al. 1992). Our re-
challenging design choice. While transparency is search has focused only on the former. We thus
desirable because it facilitates the transfer of see the need for future research on not only the
expertise and the sharing of knowledge, our efficacy of our revised, functionality focused design
research highlights the trade-off between compe- principles, but also on design principles that guide
tence transparency and users’ privacy. Our design the CMS development methodology.
principle of user-selected transparency captures
the need to make transparency subordinate to
user control. Another design challenge is to sup-
Acknowledgements
port data accuracy and system flexibility, especially
with regard to the emergence of competence.
VINNOVA (formerly NUTEK) and the participating
While real-time capture promises data accuracy as
organizations funded this work. We are also
it tracks what competence knowledge workers
deeply indebted to the participants of this study
apply in their everyday work and what interests
and our research colleagues Dick Stenmark, Jan
they are pursuing, our research revealed that
Ljungberg, Kalevi Pessi, and Magnus Bergquist.
implicit, system-generated descriptions of users’
Special thanks to Jonny Holmström for construc-
competence and interests creates anxiety among
tive and insightful comments on earlier versions of
users. Our revised design principle of real-time
this manuscript. Thanks are also due to Ole
capture with feedback loop highlights the impor-
Hanseth, Christian Hardless, Eric Monteiro,
tance of providing users access to system-
Jeremy Rose, and the special issue editor, Michael
generated, implicit profiles. The implicit profiles
Myers, the associate editor, and four anonymous
should be made adjustable through users’ explicit
reviewers for their useful comments.
rankings of their competence and interest. Thus,
again, user control should override the designer’s
desire for the accuracy and efficiency of real-time
data capture. References
In addition, CMS need to support not only Alavi, M., and Leidner, D. E. “Knowledge Manage-
individual knowledge workers’ needs, but also the ment and Knowledge Management Systems:
needs of the organization. Our research identified Conceptual Foundations and Research Issues,”
the potential for tension in competence manage- MIS Quarterly (25:1), 2001, pp. 107-136.
ment when the organization’s (and its customers’) Alvesson, M. “Organizations as Rhetoric: Knowl-
competence needs are subordinate to individual edge-Intensive Firms and the Struggle with
knowledge workers’ interests. Our design principle Ambiguity,” Journal of Management Studies
of multi-perspective interest integration captures (30:6), 1993, pp. 997-1015.
the notion that the organization’s competence Andreu, R., and Ciborra, C. “Organizational
needs and strategic competence direction must be Learning and Core Capabilities Development:
balanced with individual knowledge workers’ The Role of IT,” Journal of Strategic Information
competence development interests. Finally, given Systems (5), 1996, pp. 111-127.
the strategic importance of competence in Avison, D., Baskerville, R., and Myers, M. “Con-
knowledge-intensive organizations, the summary, trolling Action Research Projects,” Information
visualization, and reporting of competence infor- Technology & People (14:1), 2001, pp. 28-45.
mation is a key design challenge. Our research Baladi, P. “Knowledge and Competence Manage-
has highlighted the need for flexible, user- ment: Ericsson Business Consulting,” Business
controlled reporting in CMS. Strategy Review, (10:4), 1999, pp. 20-28.
Barney, J. B. “Firm Resources and Sustained
IS research on design theories distinguishes Competitive Advantage,” Journal of Manage-
between design principles that address a system’s ment (17:1), 1991, pp. 99-120.
Barney, J. B. “Organizational Culture: Can it be a Hevner, A. R., March, S. T., Park, J., and Ram, S.
Source of Sustainable Competitive Advan- “Design Science in Information System Re-
tage?,” Academy of Management Review search,” MIS Quarterly (28:1), 2004, pp. 75-
(11:3), 1986, pp. 656-665. 105.
Baskerville, R. L., and Wood-Harper, A. T. “A Kim, D. “The Link between Individual and Organi-
Critical Perspective on Action Research as a zational Learning,” Sloan Management Review
Method for Information Systems Research,” (35:1), 1993, pp. 37-50.
Journal of Information Technology (11), 1996, Kolb, D. A. Experiential Learning: Experience as
pp. 235-246. the Source of Learning and Development,
Baskerville, R. L., and Wood-Harper, A. T. Prentice Hall, Englewood Cliffs, NJ, 1984.
“Diversity in Information Systems Research Lado, A. A., and Wilson, M. C. “Human Resource
Methods”, European Journal of Information Systems and Sustained Competitive Advan-
Systems (7:2), 1998, pp. 90-107. tage: A Competency-Based Perspective,” Aca-
Bassellier, G., Benbasat, I., and Reich, B. H. “The demy of Management Review (19:4), 1994, pp.
Influence of Business Managers' IT Compe- 699-727.
tence on Championing IT,” Information Systems Lawler, E. E. “From Job-Based to Competency-
Research (14:4), 2003, pp. 317-336. Based Organizations,” Journal of Organizational
Bergenhenegouwen, G. J., ten Horn, H. F. K., and Behavior (15), 1994, pp. 3-15.
Lawler, E. E., and Ledford, G. “A Skill-Based
Mooijman, E. A. M. “Competence Develop-
Approach to Human Resource Management,”
ment: A Challenge for HRM Professionals:
European Management Journal (10:4), 1992,
Core Competencies of Organiziations as Guide-
pp. 383-391.
lines for the Development of Employees,”
Lee, A. S., and Baskerville, R. L. “Generalizing
Journal of European Industrial Training (20:9),
Generalizability in Information Systems Re-
1996, pp. 29-35.
search,” Information Systems Research (14:3),
Boyatzis, R. E. The Competent Manager: A
2003, pp. 221-243.
Model for Effective Performance, John Wiley &
Leonard-Barton, D. “Core Capabilities and Core
Sons, New York, 1982.
Rigidities: A Paradox in Managing New Product
Conner, K. R., and Prahalad, C. K. “A Resource-
Development,” Strategic Management Journal
Based Theory of the Firm: Knowledge versus
(13), 1992, pp. 111-125.
Opportunism,” Organization Science (7:5),
Lindgren, R. “Competence Visualizer: Generating
1996, pp. 477-501. Competence Patterns of Project Groups,” Jour-
Davenport, T. H., and Prusak, L. Working Knowl- nal of Information and Knowledge Management
edge: How Organizations Manage What They (2:1), 2003, pp. 1-14.
Know, Harvard Business School Press, Boston, Lindgren, R., and Henfridsson, O. “Using Compe-
1998. tence Systems: Adoption Barriers and Design
Davison, R. M., Martinsons, M. G., and Kock, N. Suggestions,” Journal of Information and
“Principles of Canonical Action Research,” Knowledge Management (1:1), 2002, pp. 65-77.
Information Systems Journal (14), 2004, pp. 65- Lindgren, R., and Stenmark, D. “Designing
86. Competence Systems: Towards Interest-
Giddens, A. The Constitution of Society: Outline Activated Technology,” Scandinavian Journal of
of the Theory of Structuration, Polity Press, Information Systems (14:1), 2002, pp. 19-35.
Cambridge, MA, 1984. Lindgren, R., Stenmark, D., and Ljungberg, J.
Hagan, C. M. “The Core Competence Organi- “Rethinking Competence Systems for Knowl-
zation: Implications for Human Resource Prac- edge-Based Organizations,” European Journal
tices,” Human Resource Management Review of Information Systems (12:1), 2003, pp. 18-29.
(6:2), 1996, pp. 147-164. Markus, M. L., Majchrzak, A., Gasser, L. “A
Hamel, G., and Prahalad, C. K. Competing for the Design Theory for Systems that Support
Future, Harvard Business School Press, Emergent Knowledge Processes,” MIS Quar-
Boston, 1994. terly (26:3), 2002, pp. 179-212.
Dr. Lindgren has published in European Journal of mation Systems Journal, Scandinavian Journal of
Information Systems, Scandinavian Journal of Information Systems, and Journal of Information &
Information Systems, Journal of Information & Knowledge Management.
Knowledge Management, and Journal of Knowl-
edge Management Practice. Ulrike Schultze is an assistant professor in Infor-
mation Technology and Operations Management
Ola Henfridsson is currently managing an action at Southern Methodist University. Her research
research project on in-car nomadic computing focuses on the impact of information technology on
support at the Viktoria Institute, Göteborg, work practices and she has been particularly
Sweden. He is also an assistant professor in interested in knowledge work, i.e., the social
Informatics at the School of Information Science, processes of creating and using information in
Computer and Electrical Engineering, Halmstad organizations. Dr. Schultze’s research has been
University. Dr. Henfridsson holds a Ph.D. in published in MIS Quarterly, Information Systems
Informatics from Umeå University, Sweden, and is Research, Information & Organization, and other
a member of the editorial board of the Scandi- journals. Dr. Schultze currently serves on the
navian Journal of Information Systems. Dr. Hen- editorial boards of MIS Quarterly, Information &
fridsson has published in Accounting, Manage- Organization, and Journal of Information Tech-
ment & Information Technologies, Database, Infor- nology.