European Management Journal 32 (2014) 1323

Contents lists available at SciVerse ScienceDirect

European Management Journal

journal homepage: www.elsevier.com/locate/emj

Deus ex machina? Career progress and the contingent benets of knowledge

management systems
Charles Galunic , Kishore Sengupta, Jennifer Louise Petriglieri
INSEAD, Department of Organisational Behaviour, Boulevard de Constance, 77305 Fontainebleau, France

a r t i c l e

i n f o

Keywords:
Knowledge management
Careers
Talent development
Social identity

a b s t r a c t
Knowledge management (KM) systems are increasingly common in rms which promote self-managed
careers and autonomy, such as professional service rms. Yet, whether or not KM systems provide real
benets is underexplored. Our focus is on the impact of KM use on the career progress of service professionals. We use recorded logs of employee KM system use and career progress data over a two-year period within a strategy consultancy to study the effect of KM use on career advancement speed. We present
a contingency-based model to KM use effectiveness, showing that, although KM use generally boosts
career progress speed, (a) benets vary by seniority (more junior employees benet more), (b) benets
vary by knowledge type (encyclopedic vs. social), with social knowledge use mattering more to career progress, and (c) those service professionals who tap a wider range of knowledge sources progress faster in
their careers. We also nd mediating effects, specically that KM system use operates partly by accelerating the development of task-related skills. We draw the conclusion that KM systems contain neither a
magical Deus ex machine for boosting employee performance and progress but nor do they warrant excessive skepticism, rather their impact on careers is contingent on employee needs.
2013 Elsevier Ltd. All rights reserved.

Introduction
Knowledge management systems are pervasive and spending
on knowledge management services is growing steadily (Musico,
2009). Evidently not a fad, corporate attention and spending have
been paralleled by rising academic interest, which has produced
numerous investigations, ranging from the ontology of organizational knowledge and its transfer (Alvesson & Karreman, 2001;
Nonaka, 1994; Ringberg & Reihlen, 2008) to knowledge management system effectiveness within organizations (Argote & Ingram,
2000; Haas & Hansen, 2005; Olivera, 2000; Thompson &
Walsham, 2004; Zhao & Anand, 2009). Studies of KM (knowledge
management) system effectiveness are important because the
risks of inated expectations are high given the volume of knowledge-based work today and the allure of technology to help eager
managers craft higher performing organizations. However, while
studies of general knowledge transfer consequences/performance
have appeared, studies of KM system benets are underexplored
(Mckinlay, 2004; van Wijk, Jansen, & Lyles, 2008). In particular,
and the focus of our paper, there is a gaping hole in our understanding of the benets of KM systems on individual careers
(including work on the general interplay between knowledge exchange and careers, see Lam, 2007). This lack of understanding is
especially conspicuous because of the converging of modern ca Corresponding author. Tel.: +33 (0)1 60 72 48 39.
E-mail address: charles.galunic@insead.edu (C. Galunic).
0263-2373/$ - see front matter 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.emj.2013.06.002

reers and KM systems. The conversion of labor-intensive jobs into

ones involving some form of information analysis, codication,
and exchange, often involves digital technologies. Also, successfully carrying-out this knowledge-based work means coping with
the fact that relevant knowledge and expertise is often removed
from the immediate workplace. This knowledge and expertise
may require considerable time and effort to acquire, particularly
where knowledge-based work is complex and demands heavy
socialization, as in the case of service professionals (Ramanujam
& Rousseau, 2006; Van Maanen & Barley, 1984). The complexity
of service professional workincluding the physical and temporal
challengesmay make the problems of organizational (Ashforth &
Saks, 1996; Van Mannen & Schein, 1979) and occupational (Pratt,
Rockmann, & Kaufmann, 2006) socialization acute. This requires
that knowledge-workers have tools that can cope with these career challenges.
KM systems should help career development and advancement.
Although operationalized in many different forms, knowledge
management systems are generally IT-supported processes which
collect, categorize, and disseminate organizational knowledge for
the purpose of decision quality, efciency (reuse), and learning or
development. They are tools that help connect workers to knowledge and other people regardless of physical distance, and that
can be used whenever a knowledge worker chooses. Additionally,
they constitute knowledge governance (Foss, Husted, & Michailova,
2010) in that rms attempt to use them as formal structures which
inuence human capital development and progression.

14

C. Galunic et al. / European Management Journal 32 (2014) 1323

Nonetheless, and despite the attention KM systems have received, we know little about the career implications for individuals
who use (or not) knowledge management tools. This gap in understanding is both practical and theoretical. With the tremendous
investments made in KM technology, there is rst the practical
question of whether users benet from them. The research focus
to date, however, is unbalanced, mostly concerned with rm-wide
measures of performance, rather than understanding how knowledge management supports the careers and work of individuals.
In general, empirical studies of knowledge management systems
have focused on the impact of knowledge management on a number of macro measures of rm performance, for example rm
growth rate (Bogner & Bansal, 2007) and patent impact (Miller,
Fern, & Cardinal, 2007). There have also been a number of inuential studies at the team-level (Haas & Hansen, 2005, 2007; Hansen,
1999), displaying important contingencies to when a KM system
adds value and when it does not. Nonetheless, KM systems fundamentally involve individuals and deliver value through individuals; it is mostly the individual who contributes to and usesor
simply ignoresthese systems on a daily basis, including for their
professional development (Foss et al., 2010). This paper, then, focuses on KM system use by individuals (inows) and seeks answers to the following research questions: Does KM system use
impact the performance and, in particular, career progression of the
professionals using them? In their interaction with KM systems, what
behaviors distinguish those who get promoted more often? The link
between KM system use and career progression has received no
attention to our knowledge; assessing this link is important for
the growing number of human capital intensive rms which depend on talent development and advancement and have invested
heavily in KM systems, such as management consultancies (Anand,
Gardner, & Morris, 2007).
Bridging studies of KM systems and careers can also contribute
to theory development. KM systems offer access to two major
knowledge types: documents (codied and explicit knowledge)
and people (experts and their tacit knowledge). The mechanisms
by which these can contribute to career progression, however, differ: the former follows a more classic approach to knowledge
acquisition and competence developmentvia self-study and
familiarity with codied concepts, rules, and procedures, that is
learning the rulebookwhile the latter emphasizes the acceleration of identity formation through access to and interaction with
the right (prototypical) local experts and relevant subgroupthat
is, learning by interacting and socialization. But are these mechanisms equally benecial to career progression and advancement?
Also, do these mechanisms matter equally to the career progress of
entry-level employees and senior staff? Our approach in answering
these questions is to explore the contingent benets of KM systems
for career advancement along these two theoretical dimensions:
knowledge type and seniority. Indeed, the empirical KM literature
has failed to better distinguish knowledge types and explore contingencies (for an exception, see Haas & Hansen, 2005). Well established classications of knowledge exist, for example the common
distinction between explicit and tacit knowledge (Ambrosini &
Bowman, 2001; Nonaka, 1994), or between positivist, social-constructionist, and socio-cognitive modes of knowledge transfer
(Ringberg & Reihlen, 2008). However, the majority of empirical
studies assessing the impact of knowledge management on
performance either focus on one basic type of knowledge (tacit
or explicit), or do not clearly differentiate between the two. Moreover, the question of their relative impact on career progress and
advancement remains unanswered.
To address these issues, we analyze individual use of a knowledge management system, over a two year period, within a management consultancy. We examine the impact of knowledge
management use on individual career progression, measured by

relative job promotion speed. Our dataset is unique in that we do

not rely upon perceptions of knowledge use but we track the actual
use (download logs) of the KM system by consultants. We also
have company data on the promotion records of consultants at various stages of their career. Both dependent and independent variables, then, are objective measures of KM system use and career
advancement. Finally, we also test for mediation effects. Consultants are assessed after each engagement on items which reect
their skill development. We include mediation (structural) models
which show to what extent promotion outcomes are driven by the
direct effects of KM use vs. driven indirectly, by the impact of KM
use on skill development.
Background: modern work and talent
In the late 1990s McKinsey & Co. introduced the metaphor of a
war for talent (Chambers, Foulton, Handeld-Jones, Hankin, &
Michaels Ill, 1998). Demographic trends indicated an aging workforce, with particular contraction in the 3545 year-old range, a
management sweetspot, even while the economy would continue to grow (34%). The new metaphor captured the competition
rms would face for an important talent pool of (mostly) knowledge-workers. This importance refers to both the scarcity of such
talent but particularly some form of technical specialization,
involving tasks that require rapidly absorbing diverse information,
creatively synthesizing new knowledge, and coordinating the work
of others in serving customers or clients. Employees capable of this
transformative work must be found and retained (Dellippi,
Arthur, & Lindsay, 2006), and, equally important, they must be
equipped and developed. For the purpose of this paper, talent refers to these difcult-to-replace, high value-adding knowledge
workers, such as partner-track professionals.
Talent management challenges are generally of three types:
recruitment, development, and retention. Research to date has focussed on recruitment and retention, studying for example the
portability of talent between rms (Groysberg & Lee, 2009). Our
interest lies in the development of talenttraining, socialization,
and career advancementbecause of the special nature and challenge of developing contemporary talent. This challenge has its
source in the way organizations tend to be managed today along
with the unique etymology of talent in business rms. In an attempt to increase efciency and exibility, organizations have become atter and spans of control wider. Simultaneously,
individuals have been urged to become more self-reliant (Arthur
& Rousseau, 1996), expected to do more without being told what
to doan economy born out of the need for management
efciencies but based on the simple rule that more autonomous
employees are also more motivated employees. This emphasis on
self-reliance has also spilled-over into careers (Osterman, 1999)
high-potential employees are expected to show self-responsibility
for their career and development, not just their job (Mabey, 2002).
There is irony in that talent, while treasured, must also fend
more for itself. But there is a big difference between self-reliant
development and being on your own. The difference is the
systems and tools that organizations make available for talent to
develop themselves. These tools include various training programmes and opportunities for advanced education, but they also
include critical on-the-job resources, such as KM systems that blur
the boundary between training instruments and in situ problem
solving devicesfor example, industry reports, client engagement
histories, practice guidelines and principles, presentation materials, and peer biographies. Consequently, these tools are critical
for knowledge-based work. The consequences of KM systems,
however, may be broader than the performance of the immediate
project. When companies provide such information, knowledge,
and tools, they are also providing learning resources for career

C. Galunic et al. / European Management Journal 32 (2014) 1323

advancement. The impact of KM systems for career advancement

have, to date, been overlooked, yet we argue that they deserve
attention. Specically, we aim to determine whether KM systems
contribute to the development of professional service employees,
and if so whether there are contingencies to this benet.
Service professional promotion speed and km systems
Career advancement is a primary concern for high-potential
employees who are expected to be self-reliantfaster is better
(Maister, 2003), for employees and their companies alike. This is
because rms benet when people are more rapidly promoted to
higher value-adding work (Engwall & Kipping, 2002). This pressure
for career progression is reected in the now commonplace up-orout systems in the majority of professional service rms. Although
up-or-out systems are stringent, professional talent doesnt move
in lockstep through the rankssome are promoted quickly, others
take longer. This variance in career progression is an opportunity to
assess the value of the learning tools at hand.
Our focus is on KM tools and their effect on career progression.
The early spread of Lotus Notes (Circa, 1989) and the advent of the
Web were the sparks for the rst successful generation of KM systems. While these original systems aimed to improve judgements
or decisions, the KM systems today are more focused on becoming
repositories of accessible knowledge in the form of large searchable databases (Olivera, 2000). In the decade since their release,
KM systems have become pervasive.
Corporate attention has been paralleled by academic interest.
Argote, McEvily, and Reagans (2003) provide a useful orienting
framework for investigations concerning the management of
knowledge, organizing studies along three properties of the context: the units (features of the people, groups, etc.), the relationships
between the units, and the nature of the knowledge (e.g., explicit vs.
tacit). Although we will draw on mechanisms from the rst two domains, our primary concern is with the nature of the knowledge
transmitted. Within this domain our study distinguishes between
two general forms of knowledge: encyclopedic and social (Hansen,
Nohria, & Tierney, 1999). Encyclopedic knowledge concerns facts,
histories, analyses, templates, and examples, all of which are stand
alone information sources; the epistemological mode is mostly positivist, with the knowledge presumed to reside in the object. In contrast, social knowledge is information that directs the user to experts
in the rm; the epistemological mode is a combination of social and
cognitive, based on human, context-rich interaction but also the actors existing private and cultural models (Ringberg & Reihlen, 2008;
Thompson & Walsham, 2004). There is also some overlap between
these knowledge types and the traditional distinction between explicit and tacit knowledge: encyclopedic knowledge tends to be explicit while social knowledge provides access to more tacit
knowledge. This distinction between the two types of knowledge
is important and needs to be measured and compared; the majority
of empirical studies assessing the impact of knowledge management on performance either focus on one type of knowledge (tacit
or explicit), or do not clearly differentiate between them (Majchrzak, Cooper, & Neece, 2004; Tanriverdi & Venkatraman, 2005).
While KM system inuence on organization performance has
provided some positive results (e.g., Bogner & Bansal, 2007; Miller
et al., 2007) studies that investigate relationships between KM and
performance at a more micro level show less straightforward
results. For example, in a study of the success of client bids conducted by management consulting teams, Haas and Hansen
(2005) found that the use of KM systems (self-report, aggregate
assessments of documents downloaded) worked poorly for experienced client teams who faced stiff competition; the KM system
worked only in the case of inexperienced teams who faced little
competition. Personal knowledge gained through networking

15

(self-report assessments of advice taking from peers) worked marginally better as a predictor, but with greater team experience this
effect also diminished. This study suggests the nuances in the relationship between knowledge management and performance at less
aggregate levels, although not at the individual level. The data for
our study is at the individual level and can further our understanding of these mechanisms because we have (a) direct measures of
KM system use (download logs for each user) and (b) data that allow us to parse-out encyclopedic vs. social knowledge downloads.
Our focus is on consultants data usage, not generation.
The measure of performance in our study is career progression,
a novel measure of KM tool effectiveness. As argued, career progression is relevant to talent development, and particularly in the
context of knowledge-based professional service rms. Here, the
career pace of junior professionals is critical as these are the individuals who not only feed the higher positions, but also from the
bulk of the fee charging population of the rm. Amongst professional service rms, the relevance of KM systems is perhaps most
salient for management consultancies. For consultantsoften geographically spreadto maximize their effectiveness they must
have access to the collective experience of the rm (Sarvary,
1999). Yet, to date, the benets and impact of KM systems on the
career progression of key talent pools has not been researched.
Our starting point is the general relationship between the career progression of consultants and their use of the KM system.
KM systems provide important opportunities to enhance learning
and developmentvia access to a breadth of industry knowledge,
presentation styles and techniques, and access to information
regarding people and experts who can be contacted and mined
for further tacit knowledge. Because the rate of career progression
of service professionals rests on their ability to access, assimilate,
and present knowledge, we begin by postulating a positive relationship between the level of KM system use and the career progression of consultants.
H1. Use of knowledge management systems positively impacts
career progression of consultants.
This rst hypothesis identies the expected general trend. But
these effects may differ between individuals in different positions
within a hierarchy. In their study of management consultancies,
Hass and Hansen (2005) found that, while knowledge management
use positively impacted the success rate of junior consultants bids,
this positive effect did not hold for more experienced consultants.
Similarly, we posit that KM systems, by accelerating socialization
and organizational familiarity, will be particularly important for
the progression of junior employees. Socialization into an organization is achieved via a range of tactics, for example formal induction and training, informal discussions with longer serving
employees, and so on. As socialization progresses, employees will
be better able to navigate the organization, with efcacy and
speed, and, in turn, make progress within the organization (Bauer,
Morrison, & Callister, 1998). Whatever socialization tactics an
organization employs, the efforts are moderated by the proactivity
of individuals. Research suggests that the more an employee seeks
to nd his/her place in the organization the more successful will be
the socialization effort (Ashford & Black, 1996; Tae-Yeol, Cable, &
Sang-Pyo, 2005).
KM systems facilitate this process. By containing a large amount
of rm-specic information (best-practices, templates, histories,
etc.) and pointers to the social structure of the rm, KM systems
can be harnessed for employee socialization. For example, junior
employees may prot disproportionately from KM systems in the
construction of their transactive memory systems (or TMS, an individuals knowledge of who knows what in the organization). In
particular, KM systems help clarify knowledge ownership and

16

C. Galunic et al. / European Management Journal 32 (2014) 1323

establish a protocol for knowledge dissemination, features which

are less well known to junior employees but important to their
progress in the rm; and these features have been linked with positive TMS development (Jarvenpaa & Majchrzak, 2008). KM systems can also make sector or industry knowledge more readily
accessiblea form of socialization that is critical to consultants
to develop into credible experts (Fincham, Clark, Handley, & Sturdy, 2008)from which more junior employees should disproportionally benet.
While employees at all levels accrue the instrumental benets
from KM systems when struggling with pending problems and
assignments, newer employees should additionally accrue the
socialization benets inherent in these systems. Thus, we expect
that junior employees will benet disproportionately from KM
systems.
H2. The positive effects of knowledge management use on career
progression will be stronger for more junior consultants.
We analyze two basic forms of knowledge in this study: encyclopedic, which concerns codied, explicit knowledge such as reports, presentations, and analyses, and social, which concerns
contact information, directing the user to experts within the rm.
There are distinct advantages to each knowledge type, roughly distinguished along the criteria of speed and quality of the information
exchange. Encyclopedic knowledge has the advantage of speed. Service professional time is costly and actively managed. When confronted by problems or ambiguity, service professionals need
rapid answers. For example, consider a consultant who is completing a complex assignment, one part of which involves a process
(say, alterations to the supply chain) with which s/he is unfamiliar;
having access to best-practice procedures and (sanitized for condentiality reasons) actual presentation slides should speed up the
work. Those professionals who use KM systems in this way should
be more efcient and learn faster, which should accelerate their career progression.
However, there are downsides to encyclopedic knowledge and
that may impede project success and, in turn, career progression.
Galunic and Weeks (1999) document the danger of non-expert
consultants getting hold of presentation materials and, under
deadline pressures, grafting the content to their presentation but
without properly understanding the process or technology at hand.
One Principal in their study related a close encounter (p.12):
. . .I had a couple of associates that were brand new, walk into my ofce. They
said, Weve heard you know a lot about how to organize xyz. And so weve
pulled this stuff of yours off [KM system], and weve got a proposal written
thats due tomorrow, can you look at it? I ipped through this document which
was on the face of it beautiful. Beautiful graphics, beautiful headlines, but it
made no sense because they had cobbled together a bunch of stuff that they
didnt understand. It was terrible. And the lesson from that is that the content
thats out there is only as good as the peoples ability to use it well.

This story is compelling because it not only highlights the dangers of purely explicit knowledge but simultaneously points to the
value of having live contact with colleagues and experts who can
help you make proper sense of the codied information (Fincham,
1999). In other words, social knowledge should have quality and
learning advantages. The core ingredient in this advantage is the
access to rich, tacit information exchange, where misunderstandings can be cleared-up, priorities highlighted, non-codied information distributed, and, at the extreme, follow-up and active
support offered. This should add considerable quality to the learning and development of service professionals.
There is further support for this comparative advantage of social
vs. encyclopedic knowledge at the team-level. Hass and Hansen
(2007) investigated a similar split of knowledge type, focusing on
their impact on the performance of sales teams in a management
consultancy. They found that seeking information via advice from

other individuals both improved the quality of the work and signaled competence to clients, while seeking codied information increased the speed of work but did not positively impact work
quality or perceived competence. Similarly, we posit that, at the
individual level, there should be signicant advantages to social
knowledge use, as it should enhance the quality of the learning
and development experience and, in turn, accelerate career progression. Relatedly, social knowledge through a KM system can
help users develop their social network, through which they can
better locate key resources and opportunities. In fact, because
many resources and opportunities will not be obvious through a
KM system but can only be accessed through familiarity with specic individuals, assembling the appropriate network is a vital precursor to accelerated career progression (Seibert, Kraimer, & Liden,
2001).
Social knowledge has at least one further advantage. KM systems can also facilitate socio-cognitive processes (Ringberg & Reihlen, 2008) such as professional identity formation, an aspect of
development which is also important to career advancement. The
view here is different from the view expressed above; networks
above provide instrumental value to the extent that they connect
users to real experts. But service professionals seek more than just
instrumental value; they seek a sense of belonging, clarity in
expectations, and condence in their conduct, all of which contribute to their professional identity. The need for accelerated identity
formation is particularly critical for service professionals. While
work roles in all professions are accompanied by a set of expectations for behaviour and general presentationor display rules
(Rafaeli & Sutton, 1989)and the adoption of display rules is
important for individual success (Hochschild, 1983), service professionals have a heightened need to follow display rules because
they are, in some ways, the product. Moreover, the construction
of a professional identity and adoption of display rules that t
with the image of ones rm and/or expertise subgroup is essential
to client-facing work, where professionals represent and must embody the rms identity (Ashforth & Humphrey, 1993). Alongside a
general image of competence, each rm and specialized subgroups,
have specic identity prototypes of the essential type of person
(Hogg & Terry, 2000). The career progression of service professionals, such as management consultants, is therefore likely to be
linked to the construction and display of a professional identity.
KM systems can facilitate identity development. Because professional identity construction necessitates exposure to the appropriate peopleexperts and prominent or prototypical colleagues in
the knowledge arena within which an actor wants to build competencehaving a tool that identies and locates such people should
accelerate development and advancement. Hence, overall we argue:
H3. Use of Social information has a stronger positive impact on
consultants career progression than use of encyclopedia
information.
One further contingency to knowledge use that should moderate career progression is the diversity of knowledge sources. Most
service professions are neither linear nor simple in their transformation of inputs into outputs. That is, there is some temptation
to naively approach knowledge management as a mechanical tool,
one that centralizes the sacrosanct methods of an institutionand
provides denitive mappings of the expertsand then expects the
rank-and-le service professionals to repetitively execute those
procedures. Certainly, the strategic management consultancy studied here is far from this scenario; consultants have an important
transformative role in their various projects (Fincham, 1999)their
career progression depends to some degree on the creativity of this
transformative process. Moreover, they are judged by the depth of
their eventual expertise and understanding.

17

C. Galunic et al. / European Management Journal 32 (2014) 1323

If creative transformation of inputs and knowledge depth is

essential to consultant success (and resulting career advancement)
then one likely driver of that success will be the diversity of the
knowledge sources or inputs. Consultants who use the KM system
to access many different sources in their development should gain
two advantages. Because different sources and types of KM inputs
(from broader issue papers and industry reports to more specic
project presentations) will offer different nuances, angles or perspectives on a topic, consultants should nd their knowledge deepening as they read across knowledge sources and build links across
the knowledge sources. Second, these alternative nuances or angles
should provide at least some material for creative recombinations.
There is a good deal of research that points to the importance of
variance in knowledge inputs to innovation (Burt, 2004; Rodan &
Galunic, 2004). For example, Obstfeld (2005) argues that individuals best placed to generate ideas and follow them through are
those who can both bring together heterogeneous information
and people and coordinate action (although the diversity of knowledge is not necessarily measured). In general, to the extent that diverse knowledge sources are likely to contribute to knowledge
depth and creativity, we posit:
H4. Accessing diverse sources of knowledge is positively associated with consultants career progression.

Finally, we consider mediators which can help us to better

understand the mechanisms by which KM system use impacts career progression. The above arguments imply, in general, that KM
systems will impact core consulting skills, which are in turn necessary to advance a consultants career. Of course, consulting skills
will include a multitude of competencies: from basic numeracy
to analytical skills to teamwork skills to presentation skills. Given
the focus of this paper, however, our concern will be on skill development that KM system use is most likely to impact and which is
particularly central or core to being a consultant, and so we
concentrate on task-related skills, such as specialist knowledge
acquisition and problem solving. We distinguish these from
team-or-social skills (getting along with teammates and/or clients)
or related personal work habits (being organized, communication
skills, coping with stress, etc.). There is insufcient and tenuous
theoretical rationale to suppose that KM system use will make people better team players, better communicators, or better able to
cope with stress. But there is better reason to believe that KM system use will have a comparative advantage with more task-related
and particularly technical knowledge. For example, Thompson and
Walsham (2004) nd that while KM systems can help codify technical knowledge they face severe challenges in highly contextual
knowledge (or as one senior sales manager noted in their study
of a software rm, The technical side of things is easier to document than the organizational side of things p. 731). Haas and Hansen (2007) also nd that task-related benets can be garnered by
users (e.g., efciency and competence signaling) although the effects are contingent (efciency gains came from document exchange while competence and quality gains came from sharing
personal advice, something the KM system in our study can help
facilitate). The acquisition of these task-related skills will also be
something that consultancies are likely to monitor and want to
see developed in their people. This was the case in the consultancy
that we studied, which measured (after each engagement) the
development of these task-related skills. Certainly, KM system
use may have effects which are captured by much more than the
development of task-related skills (e.g., it may inuence network
building and identity formation) but task-related skills should at
least partially mediate KM system use (i.e., our models should
show direct effects on career progression but also indirect effects,

through task-related skill development). To sum, we hypothesize

that the positive impact of KM system use on career progression
will be partially mediated by the development of task-related skills.
H5. The impact of KM system use on consultants career progression will be partially mediated by the acquisition of task-related
skills.

Materials and methods

We tested the hypotheses on a data set that tracked the use of
knowledge documents by consultants within the home market of
Morpheus (a pseudonym), a well established strategy consulting
company, with dispersed ofces, multiple disciplines, and a coordinated federation structure (Malhotra & Morris, 2009). Morpheus is
organized around 20 competency centers that focus on specic
industries such as automotive and chemicals. The larger competency centers have as many as 35 consultants; the smaller ones
have fewer than 10 members. Career progression for consultants
follows a pattern typical of many such companies: junior consultant (JC) consultant (CONS) senior consultant (SC) project
manager (PM) and eventually associate partner (AP). As is commonly the practice in consulting rms, Morpheus has invested signicantly in KM systems.
The KM system at Morpheus contains the core elements that
have been associated with KM systems more generally (knowledge
capture, codication, and electronic access) (Reihlen & Ringberg,
2006). In this regard, they are similar to KM systems that would
be found in other consulting companies in that the core focus is
on the storage and dissemination of knowledge objects. In this
KM system, knowledge objects typically consist of artifacts such
as power point presentations, project proposals, summaries of projects, issue papers, tables of contents, analyses of market trends
and logs of team room interactions. The documents are accessed
by users through keyword searches on document title, authors, topic, etc. Each access is recorded in a log that captures the identity
number of the employee, date, name and type of document accessed. Our data analysis focuses on the usage, not generation, of
data within the system. The system was launched in late 1999.
The recording of logs started in January 2000.
Dataset
Morpheus provided us with access to two datasets. The rst consists of a log of KM system activity between January 2000 and June
2002. This dataset is exceptional in that we can observe actual
usage rates of various KM materials. The second contains data on
employees careers for the corresponding period (employee identity numbers are retained, but names are stripped out to better ensure condentiality). The data on careers include demographic
information, evaluations in projects, and afliations in competency
centers during 20002002. Since both data sets contain employee
identity numbers, we were able to match career data with logs of
KM system use by consultants.
Our analysis focused on consultants, as opposed to administrative and support staff. We considered employees who had so called
partner-track positions, meaning JC, CONS, SC, PM or AP at any
Table 1
Career paths for consultants in sample.

No promotion
One promotion
Two promotions
Total

JC

CONS

SC

Total

0
32
56
88

6
68
16
90

21
51
0
72

27
151
72
250

18

C. Galunic et al. / European Management Journal 32 (2014) 1323

Table 2
Classication of knowledge objects by consultants.
Type of knowledge
object

Classied as
encyclopedic

Project presentation
Table of contents
Issue paper
Market trend
Newsletter
Project reference
Transcripts of team
room chats
Competency center
service folder

10
10
12
12
12
1
0

11
12

Encyclopedic
Encyclopedic
Encyclopedic
Encyclopedic
Encyclopedic
Social
Social

12

Social

Classied
as social
2
2

Classication used
in analysis

Table 3
Factor properties of mediator.
Items retained in the scale

Mean

S.D.

Task-related skills applied in projects factor

Application of specialist knowledge
Problem solving
Knowledge management
GFI

4.21
4.10
3.88

0.7
1.8
1.1

v2
P-value
Reliability

0.96
22.4
0.5
0.84

point during the period of the log les. Only two employees who
were PMs in early 2000 were promoted to Associate Partner during
20002002, therefore, promotion from PM to AP was not considered. Thus, our analysis focuses on 250 professionals. Table 1 lays
out the distribution of ranks and promotions.
Variables and model
The dependent variable is the number of promotions awarded to
a consultant during the period. At Morpheus, promotion decisions
are announced once a year. Thus, for the period under consideration, a consultant could receive 0, 1, or 2 promotions. Accordingly,
for each consultant, we categorized values in the dependent variable as 0, 1, or 2.
The independent variables for our model consisted of the frequency and diversity of encyclopedic and social information accessed. In order to categorize the different types of knowledge
objects as encyclopedic or social information, we conducted interviews with six Morpheus consultants to ascertain how knowledge
objects are used in the company. Each consultant was given all
types of knowledge objects with two examples of each. S/he was
asked to categorize them as encyclopedic or social. As Table 2
shows, there was a high degree of agreement among the consultants on how knowledge objects are used at Morpheus. Disagreements were resolved through further discussion. For each
consultant, we classied the documents s/he used in these two categories and aggregated them into frequency counts for each year.
The diversity of knowledge sources or objects accessed by each
consultant was calculated as the sum of different types of knowledge objects accessed (across both encyclopedic or social information categories). For example, if a consultant chose to access
project presentations, issue papers, newsletters and project reference documents in a particular period, then the diversity of sources
counts for that consultant during that period would be four.
Because of the nature of our dependent variable (0, 1, 2 as the
only outcome conditions), we used a multinomial logistic regression to test for the main effects. Logit analysis is more appropriate
with so few outcome conditions (even if they are not binary or
ordinal) because it is more robust than OLS regression, where

assumptions are likely to be violated (e.g., homoscedacity). Our

analysis also employed several control variables: gender, starting
position, tenure, age, size of the competency center, and project
length, each of which may be a factor in why some people are promoted faster than others. For example, larger competency centers
may offer more scope of work, and so more opportunities for
advancement. Also, longer projects (average length of projects)
may mark assignments which are more complex or difcult to
accomplish. A consultants starting position was operationalized
as JC, CONS, or SC. The size of the competency center was operationalized as small (fewer than 10 employees), medium (1020
employees) and large (more than 20 employees). Gender was 1
for females, 0 for males. Age was age in years in 2002. Tenure
was years worked in the rm in 2002. And project length was average length in months of projects.1
For our mediation analysis, we used a structural equation model
that incorporated data on evaluations in consulting engagements.
For each engagement, a consultant is evaluated by the projects leader on eight dimensions: specialist knowledge development, problem solving ability, knowledge management capability (leveraging
existing KM data and adding to the KM database), teamwork skills,
organized work habits, ability to handle stress, client communication skills, and client acceptance. As noted, our mediation analysis
is concerned with task-related skills, and so the rst two items
above were the main targets for the analysis (specialist knowledge
and problem solving ability). Items 47 concern mostly social or
personal skills (teamwork, communication, etc.) and with less connection to KM system usage, while item 8 (client acceptance) is too
broad and incorporates a host of drivers (including social skills).
Item 3 (knowledge management capability), however, is more
clearly task-related. It captures how the project leader assesses
the use of and contributions to the wider KM system in the company by the consultant; in a sense, it captures whether consultants
are able to display their KM skills and knowledge appropriation to
the leader (clearly a valuable skill in the promotion struggle). Its
also tting that actual KM use should translate into KM skills being
seen to develop through these assessments. Moreover, in an
exploratory factor analysis of the 8 items, the rst three items
loaded onto one factor (and the rest on a second factor); for this
and noted theoretical reasons, we included these three items together in one scale called task-related skills. A conrmatory factor analysis of the properties of this scale is presented in Table 3.
Also, these evaluation data are for each client engagement,
whereas the promotions are yearly events. Accordingly, we aggregated the evaluation data on an annual basis. We note that
whether we built our task-related skills scale from all three items
or just items 1 and 2 (specialist knowledge and problem solving
ability, and so excluding the KM capability measure) the results
were the same.
For the mediation analysis, we followed Iacobucci, Saldanha,
and Dengs (2007) recommendation and estimated a structural
model that accounts simultaneously for the causal relationships
among the independent, mediating and dependent variables
(Fig. 1).2 Because the dependent variable is not continuous, we estimated the parameters following the approach suggested by Arbuckle
(2007). That is, we used the Bayesian estimation procedure of AMOS
16.0, which incorporates a Markov Chain Monte Carlo (MCMC)
1
For each independent variable, we performed a ShapiroWilks test to examine
the normality of errors, which was conrmed. We then checked for heteroscedasticity
by inspecting the plot of residuals and by conducting a White test of rst and second
moment specication. No heteroscedasticity was found.
2
We also collapsed the variables for KM usage for rst and second years into one
variable that covered both of the years, one for KM social and one for KM
encyclopedic. There were no signicant differences in our results, with KM social
remaining signicant while KM encyclopedic was not. We retain the variables for
each year as it offers greater granularity in the analysis.

19

C. Galunic et al. / European Management Journal 32 (2014) 1323

Fig. 1. Mediation model.

Table 4
Means, SD and correlations total sample (N = 250).
Variable

Social KM use rst year

Social KM use second year
Encyclopedia rst year
Encyclopedia second year
Diversity in KM inputs
Employee age
Tenure

*
**

Mean

38.44
81.73
38.84
47.67
2.55
31.09
2.99

S.D.

18.09
32.48
26.74
30.17
1.94
2.67
3.47

Social information

Encyclopedia

First year

First year

0.43**
0.21*
0.23*
0.31**
0.13
0.02

Second year

0.18
0.22
0.33**
0.08
0.06

0.44**
0.31*
0.005
0.09

Diversity in KM inputs

Employee age

Second year

0.23*
0.04
0.11

0.11
0.06

0.07

p < 0.1.
p < 0.05.
p < 0.01.

approach. The procedure enables us to study ordered-categorical data

and obtain 95% credible intervals of the posterior distribution of direct, indirect and total effects. We generated 15,000 samples using
the MCMC algorithm. Regression parameters were estimated for each
single arrow, and covariances were estimated for the double arrows.
Results
Descriptive statistics
The descriptive statistics are presented in Table 4. The use of
both social and encyclopedic information increased from the rst
to the second year. However, the increase in the use of social information was much higher.3
Table 5 shows the results of the logistic regression models taking into account all 250 employees. Model 1 shows only the control
variables; model 2 shows the independent variables, and model 3
includes all variables, and which offers a signicant improvement
in t than either model 1 or 2. In models 2 and 3, the use of social
information is signicant for both years under consideration. Actors who use knowledge management objects that support social
exchange are more likely to be promoted. In contrast, the use of
encyclopedic objects is only weakly associated with the number
3
In the structural model, encyclopedic and social information are aggregated over
two years in order to construct a more parsimonious model. Control variables are not
included in the structural model, although adding them did not change the overall
results.

of promotions, and only in the second year, thus indicating that

knowledge transfer may be associated not only with benets but
also with costs (Haas & Hansen, 2007). Models 2 and 3 also show
that diversity in the use of KM sources or inputs is signicantly
associated with the number of promotions. That is, the greater
the range of knowledge sources accessed by an employee, the
greater the likelihood of promotion.
Our data also show that those actors who are relatively junior
(JC position) tend to receive more promotions within Morpheus,
not surprising for entry level positions where the early competence
testing may occur faster than for more senior roles. However, and
in line with Hypothesis 2, our results show that advantages to
KM-use also benet disproportionately JCs and CONS vs. senior
consultants (SCs). Examining our results (Table 6) on these three
seniority levels, we nd, rst, that JCs who use social information
tend to get promoted moreboth in the rst and second year. Also,
those JCs who use encyclopedic information see marginally significant (p < 0.08) promotional benets in year 1, and stronger benets in year 2 (p < 0.05). For CONS, the association between social
information and promotions is signicant and benecial in both
years, and the relationship between encyclopedic information and
promotions is much more tenuous in both years (p < 0.09 and
p < 0.08, respectively). For SCs, however, the association between
knowledge use of any sort is not signicant in either year.
For the mediation analysis, we estimated structural equation
coefcients for the entire sample, and for the respective samples
of JCs and CONS (see Table 7: a separate analysis for the senior
consultants did not yield results and is omitted here). The direct

20

C. Galunic et al. / European Management Journal 32 (2014) 1323

Table 5
Multinomial logistic regression analysis for entire sample (N = 250).
Model 1 control variables
Estimate
Intercept
Control variables
Gender (F)
Starting Position (CO)
Starting Position (JC)
Age (yrs)
Tenure (yrs)
CC size (L)
CC Size (M)
Avg. Project length (in months)
Social KM use in rst year
Social KM use in second year
Encyclopedia use in rst year
Encyclopedia use in second year
Diversity in KM inputs
Likelihood ratio Chi-square testa/F
Pseudo R-square/R-square

*
**

Model 2 independent variables

Chi-square

3.21

67.31

0.01
0.03
0.09
0.001
0.001
0.008
0.008
0.05

1.41
2.85
8.91
0.24
0.25
0.69
0.68
4.98

Estimate
**

3.32

Model 3 all variables

Chi-square

Estimate
**

70.01

0.18
0.15
0.05
0.07
0.22
33.11
0.22

0.12

**

17.39
14.71
5.01
7.23
30.43

**

**
**

Chi-square

3.46

77.01

0.005
0.01
0.09
0.002
0.001
0.006
0.002
0.03
0.14
0.13
0.05
0.09
0.21
13.21
0.31

0.85
1.52
9.01
0.38
0.21
0.56
0.26
2.68
13.91
13.01
5.02
9.01
30.15

**

**
**

*
**
*

p < 0.1.
p < 0.05.
p < 0.01.
a
Compared with previous model.

Table 6
Multinomial logistic regression analysis conducted separately for junior consultants, consultants, and senior consultants.

Intercept
Control variables
Gender (F)
Age
Tenure
CC size (L)
CC size (M)
Average project length (months)
Social KM use in rst year
Social KM use in second year
Encyclopedia use in rst year
Encyclopedia use in second year
Diversity in KM inputs
Pseudo R-square/R-square

*
**

Junior consultants (N = 88)

Consultants (N = 90)

Senior consultants (N = 72)

Estimate

Estimate

Estimate

Chi-square

0.04
0.03
0.01
0.02
0.03
0.05
0.20
0.18
0.08
0.15
0.30
0.38

3.05
2.88
2.38
2.56
2.92
3.81
20.34
17.89
7.02
14.00
30.46

**
**

*
**

0.01
0.01
0.005
0.002
0.01
0.10
0.17
0.16
0.07
0.08
0.28
0.26

Chi-square

2.05
1.98
0.52
0.61
1.82
4.92
16.45
15.49
6.31
7.38
30.07

**
**

**

0.001
0.001
0.0005
0.0001
0.0003
0.002
0.02
0.01
0.01
0.04
0.03
0.18

Chi-square

0.51
0.21
0.10
0.05
0.07
0.72
1.75
0.93
1.08
3.91
3.19

p < 0.1.
p < 0.05.
p < 0.01.

Table 7
Mediation analysis.
Entire sample
Encyclopedia
Direct effect
Indirect effect

0.045
0.015

Social
Direct effect
Indirect effect

0.085
0.046

Source diversity
Direct effect
Indirect effect

0.164
0.042

**

JR consultants
0.04
0.05

0.115
0.087

0.105
0.158

**

Consultants
0.054
0.009

0.086
0.034

0.150
0.041

**

p < 0.1.
p < 0.05.
**
p < 0.01.
*

effect coefcients measure the effects of independent variables

after controlling for the effects of the mediating variable (task-related skills). In all cases, there are direct effects of KM usage for

all of Encyclopedic and Social knowledge, and a direct effect of

KM source diversity. The indirect effect coefcients measure the
additional effects of the independent variable mediated by task-related skills development. For the entire sample, the task-related
skills affect the relationship between the use of KM systems and
promotions for social information and diversity, but not for encyclopedic information. This is similarly the case for CONS. However,
for JCs, the picture is differentindirect effects are signicant for
all three independent variables, indicating that the mediator plays
a particularly prominent role for more junior employees. Moreover, the percentage of total effects captured by the indirect effect
is signicantly higher for JCs (56% for encyclopedic, 43% for social,
and 60%, for KM input diversity) than for the entire sample (25%,
35%, and 22%, respectively) or CONS (13%, 25%, and 18%, respectively). Thus, the mediation effect is more pronounced for junior
employees, that is KM system usage amongst junior consultants
boosts promotion chances particularly through the development
of task-related skills. This is less true for more experienced

C. Galunic et al. / European Management Journal 32 (2014) 1323

consultants. Mid-level consultants do well by using the KM systemtheir promotion chances are signicantly boostedbut
task-related skill development is not a major force. Rather, network development through social knowledge becomes more
important, as we will discuss. Finally, for the most senior consultants, on the road to partnership, there are no effects whatsoever
of KM usage on promotion pace, either directly or mediated by
task-related skills.
To summarize our ndings, Hypothesis 1 is partially supported
KM use in general positively impacts career advancement pace, but
mostly through social knowledge. Hypothesis 2 is supportedKM
use benets accrue mostly to more junior employees. Hypothesis
3 is supportedsocial knowledge matters more than encyclopedic
knowledge to promotion chances. Hypothesis 4 is supporteda
diversity of knowledge sources positively impacts career advancement. Finally, Hypothesis 5 is partially supportedtask-related
skills mediate the relationship between KM usage and promotion
likelihood but particularly for junior employees.

Discussion
Our ndings reveal that KM system value depreciates with seniority, a phenomenon which also has implications for our understanding of how knowledge comes to be embedded within organizations
(Argote et al., 2003, p. 579). First, KM system benets are conditional: encyclopedic knowledge provides at best a marginal boost
to the career progression of low-ranking consultants. Encyclopedic
knowledge, to the extent that it has impact, works by boosting the
task-related skill development of the most junior employees, helping them get up to speed at very early stages of their careers. This
is when junior consultants are most in need of skill development
yet access to senior mentors may be limited: encyclopedic KM resources can provide a useful alternative. On the other hand, social
knowledge resources have more robust inuencethey help mid-level consultants as well as the most junior ones. Many consultants,
therefore, improve substantially their promotion chances through
the ongoing use of social knowledge objects, helping them to access
people/experts, and so richer insights and development opportunities. It reminds us that KM systems are, as Alvesson and Karreman
suggest (2001), more about managing people and facilitating their
interactions than forcing rm standards from the top to the bottom.
Of course, the strong positive effects of accessing social knowledge become less strong as consultants become more senior. KM
system value depreciates over time. In their early years, consultants
have to rst prove themselves in the details of the work, and they
need to do this as quickly as possible, forcing them to become rapidly
familiar with technical foundations, industry background and context, case histories, and so on. They also have to establish themselves
socially and access the more tacit insights (i.e., opinions and viewpoints) held by experts, and so this necessitates locating and building a network, something the KM system facilitates. KM system use
assists junior consultant development most of all. But, with time and
exposure, as their working habits and heuristics are constructed
layer upon layer with each new experiencethey rely less on the
KM system and more upon their own understandings and relations.
This depreciation in value is consistent with the nding of Haas and
Hansen (2005), who also found that while knowledge management
systems benetted inexperienced consulting teams, they worked
poorly for more experienced client teams. It may be that as consultants mature and gain experience and competence, relatively higher
levels of encyclopedic KM use is a signal of weakness and a net cost
(see Haas & Hansen, 2005).
We need to be careful not to mistake KM system value depreciation for a lack of value to the KM system overall. In the early years,
KM systems can accelerate learning and career progress, outcomes

21

knowledge-based rms will highly value. Moreover, these systems

provide relatively unbiased, affect-free, instrumental bridges to
knowledge at an inuential time in the lives of young employees.
Recent research has shown that affect plays a crucial role in the
decision to seek out task-related ties: people will tap for advice
those they like and actively avoid those they dislike, regardless of
competence (Casciaro & Sousa Lobo, 2008). Think of a young consultant who only relies on the network and opinion of a few local colleagues to seek out wider expertise in the rm. That consultants
search will be tainted by the likes and dislikes of his/her colleagues
of other colleagues, even though the focal actors relationship with
those same experts may produce different affective experiences and
outcomes. The knowledge search that results may be less than optimal. On the other hand, KM systems, during this critical developmental period, can expose young employees to expertise with
fewer emotional and political lters and the bias they may impose.
This also allows affect to lag and develop as a result of task-related
experiences, and not strictly precede task-related search. Of course,
over time, an employees affective judgments will grow, and, as a
senior member of the organization, they will come to prefer their
emotionally close and liked colleagues. And with potentially good
result: they can better leverage these positive affect ties, because
such ties tend to do more for the focal actor. But young employees
do not have the same affective ties to experts, and so are less likely
to benet in this way; rather, they would do better to search
through a relatively affect-neutral tool during their early careers.
Overall, our nding points to the uidity and temporal dynamics
of transferring knowledge within organizations: different methods
and tools will be more or less appropriate at different times.
The predominant inuence of social knowledge suggests that
KM systems can also contribute to professional identity formation.
Much has been written on the need for service professionals to
identify and imitate role models (Ibarra, 1999)and experts within
a consultancy are presumably one important type of role model
but less is understood about exactly how junior professionals connect with these role models in the rst place. KM systems may provide one valuable pathway. Relatedly, our ndings improve our
understanding of how capabilities are translated into task-level
experiences. As Haas and Hansen (2007) argue, there is too little
transparency and a lack of micro-foundations to our understanding
of how broader rm capabilities are deployed and impact task-level performance. Our study contributes to our understanding of
these mechanismsfor example through the mediation effects
and points to the important yet contingent value of KM tools,
which add value disproportionately to younger consultants. As
socialization and capability deployment tools, we suspect their impact will grow, as IS/IT technologies continue to improve and the
web savvy of new hires grows.
The nding that accessing diverse knowledge sources positively
relates to career progression suggests the importance of variety
to work that requires innovative thought and problem solving.
We must be careful to point out that diverse sources of knowledge
do not guarantee true diversity in the content of knowledge, but
because the genre of sources are quite different within Morpheus
(from issue papers to market trends to specic and numerous projects) and they are generated by a variety of users of the system,
variety in perspectives and ideas is not unreasonable. Consultants
at Morpheus pride themselves on insightful and creative strategic
consulting and this often requires the triangulation of numerous
pieces of data and knowledge picked-up from various sources.
Our ndings point to the advantage that consultants gain by
accessing diverse knowledge objects, exposing their minds to at
least some of the variety useful in creative recombinant thinking.
This may matter less in consultancies whose business model is focused on the roll-out and implementation of tightly pre-packaged
technologies, where the main goal is awless execution of dened

22

C. Galunic et al. / European Management Journal 32 (2014) 1323

procedures; here, we would expect encyclopedic knowledge to have

disproportionate inuence on career progression, and future studies could explore this contingency. In the case of Morpheus, however, a rm that prides itself on customized strategic consulting
services, it is reasonable that social and diverse knowledge sources
has greater importance, as we nd.
Finally, our mediation analyses point to the robust nature of KM
benets. Indirect effects exist, and particularly for JCs, such that
KM usage is facilitating task-related skill development, something
partners are looking for as they assess engagement performance
and make decisions on promotions. However, direct effects remain.
Consultants must be picking-up more than just task-related skills
in their KM system use as they struggle for promotion; gains to
creativity, social capital, and a deepening sense of what it means
to be a professional consultant are likely, an exchange that is context-rich (Thompson & Walsham, 2004).
Implications for strategic management practice
A senior HR or IT executive reading the literature on KM systems will likely conclude that they have a positive impact on rm
performance and are particularly advantageous in knowledgeintensive industries (Bogner & Bansal, 2007). Our study tempers
the enthusiasm around KM systems and tools by making us aware
of the contingent benets of KM systems. The implications for
leaders are at least threefold. First, it is important to realize that
KM systems are costly to acquire and maintain with current information. Deluxe systems which attempt to cater to all levels in the
company may be misguided. If the value created is mostly at the
junior ranks, systems could be designed for this population more
squarely in mind, focusing scarce resources on KM structures and
attributes which benet primarily junior employees. This approach
would emphasize the task-related skill development, socialization,
networking, and education advantages of KM systems.
Second, and from the opposing viewpoint to the previous point, our
study suggests that KM systems are not adding enough value to the
career progress of more senior employees. Rather than focus just on junior employees, the implication may be that executives and designers
need to be more creative with KM systems and ask how they can provide more value to senior employees. This is not an easy problem to
solve. Senior service professionals and partners will have a tendency
to think they know it all; they also tend to engage mostly in relational
(rather than technical) work, both with clients (negotiating, selling)
and with staff (managing, leading). Even where this work can be codied, there is the problem of motivating senior partners to divulge their
best practices, although longer tenure has been associated with greater
inputs to KM systems (Watson & Hewett, 2006); attempting to better
capture this knowledge through encyclopedic means may be nave,
but providing better social tools may be possible and worth exploring.
This approach would emphasize the collaboration-building and search
capacities of KM systems.
Finally, KM systems may be expected to establish greater connectivity and collaborativeness in the rm, and executives may desire
these broader cultural benets. But it is not given that the connectedness of the organization will be more than instrumental and taskoriented. Just because (junior) employees tend to use and benet
from connections to experts (who will tend to be more senior
employees) does not mean bridges will be built horizontally or that
a culture of sharing and support will be established. The difference is
between a KM system that is, more or less, selsh (use it because it
helps your work and the specic problems that you are facing, getting you up-to-speed faster; contribute because it gives you greater
visibility and status) and one that is more collaborative (create
connections that will solve not only immediate issues but create
new platforms for ongoing innovation and organic growth, generating broader-based trust and cooperativeness in the rm). The rapid

development of public social networking tools may create new technologies, and demands, for these tools within companies. The question of value-added, however, will remain. Executives should be
cautious with the promises and hubris that may surround these systems, and future studies should examine whether such more sophisticated KM approaches offer such extended benets.

Limitations and future research

Our study benets from a rare database. Having actual KM
usage logs is unique and much more objective than self-reported
survey data on KM usage. However, our study and method has several limitations. First, without survey data it is more difcult to
capture additional baseline controls, such as better differentiating
project differences (difculty, learning opportunities, etc.) faced by
consultants and personal differences which may explain promotion rate (e.g., personal absorptive capacity). Future research
should explore ways to expand the baseline model in assessing
promotion speed and KM use. Second, it is important to acknowledge that KM system use could work through political skills and
behaviour in accelerating promotion pace. Consultants (particularly junior ones) could possibly use KM information to manage
better impressions but without necessarily developing true
skills. This is a real possibility. We can only counter by suggesting
that over time, such behaviour is discoverable and not really sustainable as a substitute for competence development. Third, the
period of study was during an economic downturn, making it much
less likely for service professionals to voluntarily leave their jobs,
and much more likely that professional service hiring new workers
could capture the cream of the talent pool. Although this should
not impact our main results, because we take tenure into account,
conducing a similar study in a period of rapid growth would be a
helpful repetition. Fourth, while Morpheus uses a KM system
which incorporates generic KM features (experience input by
users, codication, dissemination, all through electronic and
searchable means) it may be the case that ner-grained differences
in the KM system makes a difference to how, for example, junior
and senior people make use of the tool. We must be cautious not
to generalize without more and comparative research.

Conclusions
Although knowledge management systems are pervasive, there
are too few studies of their benet in the development of a companys talent pool. The danger is that we presume a broader benet
than is the case, a Deus ex machina approach to this technology and
its role in shaping strategic processes. Deus ex machina comes to us
from Greek tragedies, where gods were lowered by cranes onto
the stage to offer decisive solutions to plots (literally, gods from
the machine), in other words magical solutions to the complexity
of unfolding human drama. In a way, this is perhaps what we expect
of KM systems, magical solutions to complex problems. Our study
tries to provide some demystication of KM benets- there are certainly no Deus ex machine through KM systems. But there is also
danger at the other extreme: if we dont know the true value of such
systems, it is easier to grow cynical and totally dismiss their impact.
This paper tackles the issue of value-added by focusing on employee
career promotion rate, and offers a contingent-benets model of KM
system effectiveness. While promotion rate benets are evident,
they exist mostly for junior employees and largely through social
KM tools. By parsing-out the contingent value of such systems,
mindful of the career and organizational dynamics which support,
or constrain, their use, organizations and designers will be better
able to design intelligent and effective knowledge transfer systems.

C. Galunic et al. / European Management Journal 32 (2014) 1323

References
Alvesson, M., & Karreman, D. (2001). Odd couple: Making sense of the curious
concept of knowledge management. Journal of Management Studies, 38(7), 995.
Ambrosini, V., & Bowman, C. (2001). Tacit knowledge: Some suggestions for
operationalization. Journal of Management Studies, 38(6), 811829.
Anand, N., Gardner, H., & Morris, T. (2007). Knowledge-based innovation:
Emergence and embedding of new practice areas in management consulting
rms. Academy of Management Journal, 50(2), 406428.
Arbuckle, J. L. (2007). AMOS 160 users guide. Chicago: SPSS.
Argote, L., & Ingram, P. (2000). Knowledge transfer: A basis for competitive
advantage in rms. Organizational Behavior & Human Decision Processes, 82(1),
150169.
Argote, L., McEvily, B., & Reagans, R. (2003). Managing knowledge in organizations:
An integrative framework and review of emerging themes. Management Science,
49(4), 571582.
Arthur, M. B., & Rousseau, D. M. (1996). The boundaryless career as a new
employment principle. In M. B. Arthur & D. M. Rousseau (Eds.), The boundaryless
career: a new employment principle for a new organizational era (pp. 120). New
York, Oxford: Oxford University Press.
Ashford, S. J., & Black, J. S. (1996). Proactivity during organizational entry: The role
of desire for control. Journal of Applied Psychology, 81(2), 199214.
Ashforth, B., & Humphrey, R. (1993). Emotional labor in service roles: The inuence
of identity. Academy of Management Review, 18, 88115.
Ashforth, B. K., & Saks, A. M. (1996). Socialization tactics: Longitudinal effects on
newcomer adjustment. Academy of Management Journal, 39(1), 30p.
Bauer, T. N., Morrison, E. W., & Callister, R. R. (1998). Organizational socialization: A
review and directions for future research. Research in Personnel and Human
Resources Management, 16, 149214.
Bogner, W., & Bansal, P. (2007). Knowledge management as the basis of sustained
high performance. Journal of Management Studies, 44(1), 165188.
Burt, R. S. (2004). Structural holes and good ideas. American Journal of Sociology,
110(2), 349399.
Casciaro, T., & Sousa Lobo, M. (2008). When competence is irrelevant: The role of
interpersonal affect in task-related ties. Administrative Science Quarterly, 53(4),
655684.
Chambers, E. G., Foulton, M., Handeld-Jones, H., Hankin, S. M., & Michaels Ill, E. G.
(1998). The war for talent. McKinsey Quarterly (3), 44.
Dellippi, R. J., Arthur, M. B., & Lindsay, V. J. (2006). Knowledge at work. Blackwell.
Engwall, L., & Kipping, M. (2002). Introduction: Management consulting as a
knowledge industry. In M. Kipping & L. Engwall (Eds.), Management consulting:
Emerging dynamics of a knowledge industry. Oxford: Oxford University Press.
Fincham, R. (1999). The consultant-client relationship: Critical perspectives on the
management of organizational change. Journal of Management Studies, 36(3),
335351.
Fincham, R., Clark, T., Handley, K., & Sturdy, A. (2008). Conguring expert
knowledge: The consultant as sector specialist. Journal of Organizational
Behavior, 29(8), 11451160.
Foss, N. J., Husted, K., & Michailova, S. (2010). Governing knowledge sharing in
organizations: Levels of analysis, governance mechanisms, and research
directions. Journal of Management Studies, 47(3), 455482. http://dx.doi.org/
10.1111/j.1467-6486.2009.00870.x.
Galunic, D. C., & Weeks, J. (1999). Managing knowledge at Booz-Allen & Hamilton.
INSEAD Case Study.
Groysberg, B., & Lee, L.-E. (2009). Hiring stars and their colleagues: Exploration and
exploitation in professional service rms. Organization Science, 20(4), 740.
Haas, M., & Hansen, M. (2005). When using knowledge can hurt performance: The
value of organizational capabilities in a management consulting company.
Strategic Management Journal, 26(1), 124.
Haas, M. R., & Hansen, M. T. (2007). Different knowledge, different benets: Toward
a productivity perspective on knowledge sharing in organizations. Strategic
Management Journal, 28(11), 11331153.
Hansen, M. T. (1999). The search-transfer problem: The role of weak ties in sharing
knowledge across organizational subunits. Administrative Science Quarterly,
44(1), 82111.
Hansen, M. T., Nohria, N., & Tierney, T. (1999). Whats your strategy for managing
knowledge? Harvard Business Review, 77(2), 106116.
Hochschild, A. (1983). The managed heart. Los Angeles: University of California
Press.
Hogg, M., & Terry, D. (2000). Social identity and self-categorization processes in
organizational contexts. Academy of Management Review, 25(1), 121140.
Iacobucci, D., Saldanha, N., & Deng, X. (2007). A meditation on mediation: Evidence
that structural equations models perform better than regressions. Journal of
Consumer Psychology (Lawrence Erlbaum Associates), 17(2), 15p.
Ibarra, H. (1999). Provisional selves: Experimenting with image and identity in
professional adaptation. Administrative Science Quarterly, 44(4), 764791.

23

Jarvenpaa, S. L., & Majchrzak, A. (2008). Knowledge collaboration among

professionals protecting national security: Role of transactive memories in
ego-centered knowledge networks. Organization Science, 19(2), 260276. http://
dx.doi.org/10.1287/Orsc.1070.0315.
Lam, A. (2007). Knowledge networks and careers: Academic scientists in industry
university links. Journal of Management Studies, 44(6), 9931016. http://
dx.doi.org/10.1111/j.1467-6486.2007.00696.x.
Mabey, C. (2002). Mapping management development practice. Journal of
Management Studies, 39(8), 11391160.
Maister, D. (2003). Managing the professional service rm. Free Press.
Majchrzak, A., Cooper, L., & Neece, O. (2004). Knowledge reuse for innovation.
Management Science, 50(2), 174188.
Malhotra, N., & Morris, T. (2009). Heterogeneity in professional service rms.
Journal of Management Studies, 46(6), 895922. http://dx.doi.org/10.1111/
j.1467-6486.2009.00826.x.
Mckinlay, A. (2004). Knowledge management. In S. Ackroyd, R. Batt, &
P. S. Tolbert (Eds.), Oxford handbook of work and organization (pp. 242262).
Oxford: OUP.
Miller, D., Fern, M., & Cardinal, L. (2007). The use of knowledge for technological
innovation within diversied rms. Academy of Management Journal, 50(2),
307326.
Musico, C. (2009). Knowledge (management) is Power: the need for knowledge now
permeates the entreprise. CRM Magazine, 13, 49.
Nonaka, I. (1994). A dynamic theory of organizational knowledge creation.
Organization Science, 5, 1437.
Obstfeld, D. (2005). Social networks, the tertius iungens orientation, and
involvement in innovation. Administrative Science Quarterly, 50(1), 100130.
Olivera, F. (2000). Memory systems in organizations: An empirical investigation of
mechanisms for knowledge collection, storage and access. Journal of
Management Studies, 37(6), 811.
Pratt, M. G., Rockmann, K. W., & Kaufmann, J. B. (2006). Constructing
professional identity: The role of work and identity learning cycles in the
customization of identity among medical residents. Academy of Management
Journal, 49(2), 28p.
Rafaeli, A., & Sutton, R. I. (1989). The expression of emotion in organizational life.
Research in Organizational Behavior, 11, 142.
Ramanujam, R., & Rousseau, D. M. (2006). The challenges are organizational not just
clinical. Journal of Organizational Behavior, 27(7), 811827.
Reihlen, M., & Ringberg, T. (2006). Computer-mediated knowledge systems in
consultancy rms: Do they work? Research in the Sociology of Organizations, 24,
317347.
Ringberg, T., & Reihlen, M. (2008). Towards a socio-cognitive approach to
knowledge transfer. Journal of Management Studies, 45(5), 912935. http://
dx.doi.org/10.1111/j.1467-6486.2007.00757.x.
Rodan, S., & Galunic, D. C. (2004). More than network structure: How knowledge
heterogeneity inuences managerial performance and innovativeness. Strategic
Management Journal, 25(6), 541562.
Sarvary, M. (1999). Knowledge management and competition in the consulting
industry. California Management Review, 41(2), 95107.
Seibert, S. E., Kraimer, M. L., & Liden, R. C. (2001). A social capital theory of career
success. Academy of Management Journal, 44(2), 219237.
Tae-Yeol, K., Cable, D. M., & Sang-Pyo, K. (2005). Socialization tactics, employee
proactivity, and personorganization Fit. Journal of Applied Psychology, 90(2),
232241.
Tanriverdi, H. S., & Venkatraman, N. (2005). Knowledge relatedness and the
performance of multibusiness rms. Strategic Management Journal, 26(2),
97119.
Thompson, M. P. A., & Walsham, G. (2004). Placing knowledge management in
context. Journal of Management Studies, 41(5), 725747.
Van Maanen, J., & Barley, S. R. (1984). Occupational communities. Research in
Organizational Behavior, 6, 287365.
Van Mannen, J., & Schein, E. H. (1979). Toward a theory of organizational
socialization. In B. M. Staw (Ed.). Research in Organizational Behavior (Vol. 1,
pp. 209264). Greenwich, CT: JAI Press.
van Wijk, R., Jansen, J. J. P., & Lyles, M. A. (2008). Inter- and intra-organizational
knowledge transfer: A meta-analytic review and assessment of its antecedents
and consequences. Journal of Management Studies, 45(4), 830853. http://
dx.doi.org/10.1111/j.1467-6486.2008.00771.x.
Watson, S., & Hewett, K. (2006). A multi-theoretical model of knowledge transfer in
organizations: Determinants of knowledge contribution and knowledge reuse.
Journal of Management Studies, 43(2), 141173. http://dx.doi.org/10.1111/
j.1467-6486.2006.00586.x.
Zhao, J. Z., & Anand, J. (2009). A multilevel perspective on knowledge transfer:
Evidence from the Chinese automotive industry. Strategic Management Journal,
30(9), 25p.