See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/223765159

Managing Dispersed Workers: The New Challenge in Knowledge Management

Article in Technovation · May 2006

DOI: 10.1016/j.technovation.2005.02.003

CITATIONS READS
80 1,552

5 authors, including:

Mariano Corso Antonella Martini

Politecnico di Milano Università di Pisa
116 PUBLICATIONS 2,385 CITATIONS 117 PUBLICATIONS 2,254 CITATIONS

SEE PROFILE SEE PROFILE

Luisa Pellegrini Silvia Massa

Università di Pisa Università degli Studi di Genova
76 PUBLICATIONS 1,689 CITATIONS 40 PUBLICATIONS 1,191 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

CONTINUOUS INNOVATION INTERNATIONAL SURVEY View project

MIT-UNIPI 2015 View project

All content following this page was uploaded by Silvia Massa on 24 October 2017.

The user has requested enhancement of the downloaded file.

 Technovation 26 (2006) 583–594
www.elsevier.com/locate/technovation

Managing dispersed workers: the new challenge

in Knowledge Management
Mariano Corsoa, Antonella Martinib, Luisa Pellegrinib, Silvia Massac, Stefania Testac,*
a
Department of Management, Economics and Industrial Engineering, Polithecnic of Milano, P.zza L. da Vinci, 32-20133 Milano (I), Italy
b
Faculty of Engineering, University of Pisa, Via Diotisalvi, 2-56126 Pisa (I), Italy
c
Department of Communication, Computer and System Science, University of Genova, V.le Causa, 13-16145 Genova (I), Italy

Abstract
Knowledge Management (KM) has become a central theme in today’s business environment and a commonly cited source of
competitive advantage. In today’s global economy, many economic activities in and among firms are based on the contribution of
workers that are geographically dispersed and have loose contractual links with the company. Building on evidence from three case
studies, this paper explores the relations between dispersed worker activities, KM tools and worker satisfaction.
q 2005 Elsevier Ltd. All rights reserved.

Keywords: Knowledge Management; Dispersed worker; Worker activity; Worker satisfaction; Case study

1. Introduction positions and work in multi-disciplinary teams. As a

The intensification of competition results in companies
relying more and more on the competencies and flexibility
of their personnel. Knowledge Management (KM) is
becoming a critical issue. New emerging organizational
models have characteristics that challenge traditional ways
of sharing and managing knowledge. These characteristics
can be summarized in two ways:
(i) Space dispersion. Today the concept of the workplace
is changing. People spend a growing part of their
working time outside the physical boundaries of their
companies, often mobile and interacting with
customers or colleagues of partner organizations. Since
compe- tition occurs not so much among companies as
among networks, the network, more than the single
company, becomes the context in which individuals
operate and have to assimilate and transmit knowledge.
Also when working inside the company, people
often change
* Corresponding author. Tel.: C39 10 3532135; fax C39 10 3532948.
E-mail addresses: mariano.corso@polimi.it (M. Corso), antonella.
mar t i ni@ dsea.u ni pi.i t (A. Martini), luisa. pel legr in i@dsea.uni pi.it
(L. Pellegrini), silvia.massa@unige.it (S. Massa), stefania.testa@unige.it
(S. Testa).
0166-4972/$ - see front matter q 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.technovation.2005.02.003
consequence, individuals are less and less exposed to
their peer colleagues with the same ‘functional’
competencies and can hardly rely on their experience.
(ii) Contractual dispersion. The provisional nature of
employment, loose forms of contractual links to the
company and the high level of turnover render people
to have a temporary and partial involvement in
organiz- ations, thus creating barriers to the sharing of
knowl- edge and expertise among individuals.
These two factors converge in what we call ‘dispersion
of work’. Being a ‘dispersed worker’ (DW) is in many
cases a need rather than a choice. It requires individual
qualities such as autonomy and proactiveness that are not
always fundamental for the success of traditional
workers. Dispersed workers cannot simply rely on
formal career development paths but have to care
individually for their own professional growth.
While dispersed workers nowadays represent a growing
part of the workforce, traditional management systems are
not designed to answer their needs.
When dealing with DWs, Knowledge Management
becomes a fundamental issue for twofold reason:
(i) Dispersed worker knowledge and experience are
becoming fundamental assets, that have to be properly
managed.
584 M. Corso et al. / Technovation 26 (2006) 583–594
(ii) A proper Knowledge Management System (KMS) can
create a connection between dispersed workers and the
professional and social network of the company.
It provides DW with opportunities for interaction and
learning and thus contributes to their professional
growth, long term employability and ultimately
improves job satisfaction and attractiveness.
The biggest opportunities are now represented by the
availability of new Information and Communication
Technologies (ICT), and in particular web and mobile
communication services. At a rapidly decreasing cost, these
services make it possible to overcome geographical, time
and organizational barriers to communication and knowl-
edge transfer in dispersed networks.
Nowadays, the theoretical debate on KMS for DWs is
still in its infancy and limited to mainly anecdotic
evidences about best practices. The challenge for
management theory is therefore clear: to provide
empirically grounded and actionable knowledge for
companies in order to design and implement effective
KMS for DWs.
This paper presents the results from three case studies
where organizational, managerial and technological actions
have been taken to support KM in networks of dispersed
workers. Based on case comparison, implications are drawn
on how to design KMS in such environments.
The paper is organized in four sections. The state of the
art management literature on knowledge dispersion and
mobile/dispersed workers is discussed in Section 2. It also
details the empirical background of the research. Results
from the case studies are presented and analyzed in Section
3. This analysis allows us to derive the implications for
KMS design in a dispersed work environment. Implications
for management actions and plans for further research are
finally discussed in Section 4.
2. Theoretical and empirical background
The issue of workforce dispersion has been treated in
economic and management literature mainly in connection
with the problems of preserving knowledge and
competence from loss and spill over (Minkler, 1993;
Tsoukas, 1996; Becker, 2001; Cramton, 2001; Maskell,
2001).
For example, Minkler (1993) focuses on firms that tried
to assess the importance of their dispersed knowledge; he
emphasizes that some organizational innovations—
employee participation, organization in cooperative teams,
just-in-time stock management systems, forms of guarantee
towards employees—are solutions that stem from firms’
awareness of the dispersion of their knowledge assets.
The dispersion of knowledge exists in all organizations
(Cramton, 2001). However, in traditional companies
workers interact within the organizations’ physical bound-
aries, thus facilitating interactions. The level of knowledge
dispersion has progressively increased in the past few
years,
exceeding previous boundaries. Various factors contribute
to this trend, including internationalization of markets,
diffusion of companies across wide areas, and product and
service customization. In many cases, these factors induce
firms to decrease the physical distance that separates them
from their customers/suppliers, having some employees
temporarily located in the customers/suppliers offices.
In this context, the main problem is how to effectively
create a technological and social infrastructure allowing
knowledge and information management.
Several authors (Becker, 2001; Cramton, 1997, 2001,
2002) have pointed out that there are many relevant
organizational and managerial consequences resulting
from the level of knowledge dispersion. Becker (2001)
indicates three factors as the cause of organizational
problems: large numbers, asymmetries and uncertainty.
Becker also identifies some strategies that enable a better
use of dispersed knowledge:
(i) substitute knowledge with providing access to
knowledge;
(ii) provide users with the capability of completing
incomplete knowledge;
(iii) design institutions with appropriate coordination
mechanisms;
(iv) decompose organizational units into smaller ones;
(v) increase the information available to the decision-
maker.
Focusing on knowledge workers, Cramton (1997, 2001
and 2002) identifies five problems that result in failures
connected to establishing and maintaining mutual knowl-
edge among DWs: (1) failures to communicate and retain
contextual information regarding different members; (2)
interference between unevenly distributed information and
team-level collaboration; (3) differences in evaluating the
relevance of information, (4) differences in speed of access
to information (5) difficulties in interpreting the meaning of
silence/lack of communication. She also suggests that these
failures can have serious consequences for the viability of
dispersed collaboration.
Many authors have analysed and proposed possible
solutions to facilitate cooperation between geographically
dispersed workers or virtual teams (Cramton, 1997, 2001;
Furst et al., 1999; Duarte and Snyder, 1999; Lipnack and
Stamps, 1997; Pe´ rez et al., 2002).
Two aspects contribute to an effective management of
distributed knowledge: an efficient communication/interac-
tion structure (Cohen and Levinthal, 1990) and a focused
human resources management system (Tsoukas, 1996).
Communication can be enabled by creating information
channels, both formal and informal (Nahapiet and Ghoshal,
1998; Gupta and Govindarajan, 2000). The introduction of
job rotation is an example of an organizational solution,
while intranet and corporate portals are technological tools
facilitating the interaction among experts.
 A fundamental assumption that is common to recent
literature is that coordination and decision making do not
require knowledge centralization, but rather the access to
knowledge not necessarily centralized (Nonaka, 1990).
However, access to knowledge requires users to have
both competencies and capacities for understanding,
assimilating and using retrieved knowledge. In addition,
users should also be able to correct possible mistakes
(Collins and Kusch, 1998), adapt knowledge to the
specific problem and complete possible gaps (Nonaka
and Takeuchi, 1995).
With regard to the impact of KMS on performances, a
big stream is emerging in literature (Haanes and
Lowendhal,
1997; Petrash, 1996; Roos et al., 1997; Schiuma and Marr,
2001; Sveiby, 1997). The impact on performances is
strongly related to the approach adopted in the KMS design
and implementation (Davenport and Prusak, 1998; Wiig,
1997) that directly influence organizational behaviors in
terms of knowledge creation, transfer and capitalization.
Overall, current literature seems to focus on the problem
of coping with knowledge dispersion, rather than on the
phenomenon of managing networks of dispersed
knowledge workers. In particular, there is a lack of
empirical research regarding its diffusion, characteristics
and effects on Knowledge Management Systems. Up to
now there have been few studies that can help
companies in designing effective Knowledge
Management Systems to support dispersed workers.
In order to answer to the need of empirically grounded
research, in 2003 we started a project whose background
benefits from two past studies exploring KMS in Italian
firms (Corso et al., 2004a).
On the basis of a literature review and of a field research
consisting of six explorative case studies and a survey on
410 Italian firms (with a response rate of 20.2%), the
research investigation framework has been developed
(Fig. 1). It analyses three groups of variables and their
relationships: DW Profiles, KM Systems and Performances.
Fig. 1. The research framework.
In the framework, the choice of the KM System, made
accordingly to DW Profiles, produces effects in terms of
Performances.
By dispersed worker we refer to the worker who:
(i) is contractually linked to the firm and can be subjected
to firm control and/or coordination, such as employees
(both permanent or temporary) and coordinated and
continuative assistants;
(ii) satisfies at least one of the following conditions:
(a) (s)he mainly works outside the firm. Employees
always dislocated outside the firm (such as, in
some cases, the sales agents) and mobile workers
who, though having their own site in the firm,
mainly work outside the firm (for example
consultants and maintenance people) are included;
(b) (s)he is an employee who works in a firm with less
than 10 employees.
The DWs profiles are described in terms of: (a) the work
they do (that is the tasks the DWs perform); (b) the
relationship which links the DWs to the firm.
For our purpose, we adopt the following definition of
KM
and KMS (Corso et al., 2004b):
Knowledge Management is about creating an environ-
ment that encourages people to learn and share knowl-
edge by aligning goals, integrating bits and pieces of
information within and across organisational boundaries,
and producing new knowledge that is usable and useful
to the organisation. Systems supporting knowledge
processes (and their management, for that matter)
therefore exist in, and must be designed to fit, the
internal and external context of the organisation. The
key aspects of KM are:
(i) Knowledge Management is a configuration of
technical, organisational and managerial choices
(KMS).
 (ii) Knowledge Management directly influences human
behaviour and through that company performance.
(iii) Knowledge Management can improve effectiveness
in all phases of the knowledge lifecycle, including
the acquisition, capitalisation and reuse, transfer
and sharing of knowledge.
The lack of research and especially the lack of
empirically grounded analyses on DWs led us to carry out
the investigation into three different phases:
(i) selection of the relevant variables for the research
protocol. We conducted six explorative case studies
which highlighted how the different types of DWs are
the result of a specific combination of variables that
describe the relationships which link dispersed
workers to the firm.
(ii) development of a first-step telephone survey aimed at
measuring their diffusion (Corso et al., 2004b).
(iii) distribution of a second paper/electronic survey to the
firms identified in the previous point in order to
analyze the DW activities and the adopted KM tools.
The questionnaire was addressed to the same person
we previously contacted.
As regards the telephone survey, 1504 firms were
contacted and 899 answered the questionnaire (59.8%
response rate). As the survey addresses the large Italian
firms, which Mediobanca calculate are 2500, we had a 36%
response rate. Within the 899 firms, which answered the
survey, 410 have DWs (45.6%). With regards to the
paper/electronic survey, we obtained 83 answers (response
1
rate 20.2%). Table 1 summarizes samples and returns data.
DW activities were grouped in the following three
typologies (Corso et al., 2004b):
(i) Sales force. Forty-seven percent of the firms. Usually
commercial competencies are supported by technical
competencies regarding products.
(ii) Technicians. In 35% of the firms DWs are technicians
such as maintenance people or installers. In some
cases, commercial competencies are required.
(iii) Consultants. In 18% of the firms DWs are consultants,
researchers, plant directors, inspectors or responsible
of branch offices.
As activity seems to play a key role in differentiating
DWs, we developed case studies in order to analyze how
these typologies impact on KM tool selection. Referring to
the research framework in Fig. 1, in the paper we intend to
1
Sampling was based on 2002 AIDA database (Bureau van Dijk),
excluding firms with less than 250 employees and industries as
Agriculture, Fishing, Mining, Financial intermediation, Public
Administration and Defence. The person in charge of answering to the
questions had the possibility to select the modality of transmission of the
paper questionnaire between snail mail, e-mail or fax.
Table 1
Samples and response rates for the two surveys
Survey
a
Sample Response Response
rates (%)
Telephone survey 1504 899 59.8
Paper/electronic 410 83 20.2
survey
a
Population: 2500 firms.
explore the relations between dispersed worker activities,
KM tools and worker satisfaction (Fig. 2).
3. Investigation framework and empirical settings
In order to explore the influence of DW activities on
KMS effectiveness, the paper draws on Perrow’s model of
technology and structure (1967). Perrow’s model highlights
two key dimensions in order to describe a task (Fig. 3): the
existence (or otherwise) of well-established techniques for
performing tasks (task analyzability), and the degree of
variety in the tasks encountered (number of exceptions).
The two dimensions are commonly, and collectively,
referred to as task uncertainty. Where established tech-
niques for handling tasks do not exist (low analyzability),
or where substantial variety or novelty in the tasks
encountered exists (high number of exceptions), Perrow
describes the task setting as ‘non-routine’. Conversely,
when tasks are analyzable with few exceptions, the task
setting is ‘routine’. Perrow proceeds to describe the
structural arrangements necessary to achieve effective
control and coordination for the different task
environments faced by an organization. Fig. 3 presents a
diagrammatic representation of Perrow’s framework.
Tasks in Cell 1 are repetitive and familiar (few
exceptions), with known routines for handling them (high
analyzability). Perrow also argues that organizations facing
such tasks would be able to rely on procedure guides,
operating manuals, job codification and rigid lines of
reporting and accountability for controlling employee
behaviour.
In contrast, Cell 3 represents the ‘non-routine’ situation
and it is here that Perrow expects that formal, bureaucratic
controls will not be effective for controlling performance.
Tasks cannot be ‘programmed’ and thus behaviour cannot
be controlled by implementing procedures which pre-
specify desired actions or by monitoring individual actions
through the use of supervisors. The remaining two
situations depicted in Fig. 3 (Cells 2 and 4) involve task
settings where it is more difficult to predict which forms
of control will prove most effective. Tasks may have
little variety, yet exhibit uncertainty in the transformation
process (Cell 2); others may have little uncertainty but a
great deal of variety (Cell 4). They are in the middle of the
routine/non-routine continuum, but they are not the same.
Considering Cell 2, tasks here are low in variety
suggesting the potential
 DW PROFILES
- Activities - Organisational mechanism
Sales force
- Managerial
Technicians - ICT tools
Consultants
Fig. 2. The paper focus.
suitability of behaviour controls. However, notwithstanding
the lack of variety of tasks, little has been learned about the
process by which inputs are transformed into outputs;
programs for performing these tasks have not been
established (i.e., analyzability is low). It is, therefore, not
possible to develop a set of manuals, guides or procedures
which can be drawn upon during task execution.
In Cell 4, much variety is encountered in tasks, but
notwithstanding this variation, input–output relations can
be readily specified. This enables the development and
implementation of programmed controls, such as procedure
guides and operating manuals, and/or the use of superior
authority to ensure that behaviour is in accordance with
what is known to achieve desired results. It is worth noting
that even if this is feasible, it is often too time consuming.
Based on this classification and on existing literature on
Knowledge Management, we can formulate hypotheses
about the characteristics of the Knowledge Management
System that can better suit each specific situation. The
general hypothesis, therefore, is as follows:
Hyp. 1. there is a relation between the characteristics of the
task performed by dispersed workers, classified according
to Perrow’s model, and the characteristics that a
Knowledge Management System should have in order to
be effective.
KMS effectiveness is measured through user satisfaction
(Choe, 1996; Chang et al., 2003) that is a commonly used
proxy in literature.
In order to test and refine the research hypothesis, three
case studies have been developed:
(i) Case A is an Italian System Integrator that recently
developed a KMS to support its dispersed consultants.
(ii) Case B is a Telecom Operator that developed a KMS
to support its indirect sales personnel.
(iii) Case C is a provider of facility management services
that developed a KMS to support its field force
technicians.
The consulting process in A is characterized by a low
repetitiveness of activities, an absence of standardized
procedures and practices and an unavailability of a clear
body of knowledge which can guide the work. According
to Perrow’s model this task can be classified as the
‘non- routine’ situation (Cell 3).
The sales task in B is characterized by a great deal of
variety because of the high number of customers and a little
uncertainty as tasks can be performed through a predefined
KMS
PERFORMANCES
- Worker satisfaction
and limited set of procedures. According to Perrow’s model
this task can be classified as having a high analyzability and
a high number of exceptions (Cell 4).
Finally, the activities carried out by field force
technicians in C can be regarded as ‘routine’ situations,
characterized by little task variety and a clear view of input-
output relations in task execution (Cell 1).
Task selection has been performed using the Withey
et al. (1983) 10-item instrument, in seven-point questions to
measure task analyzability and number of exceptions. The
main issues considered in such an instrument are, for
example: the existence (or not) of a clearly defined body of
knowledge which can guide the task, the existence (or not)
of established procedures and practices, the existence (or
not) of an understandable sequence of steps that can be
followed in carrying out the task.
Table 2 compares the features of the three companies,
while Fig. 4 summarises the characteristics of their
dispersed workers’ activities in terms of Perrow’s model.
3.1. Data collection and analysis
Data were collected through semi-structured interviews
with individual respondents and observations.
Case studies were conducted in late 2003 and early
2004. During the site visit, overall impressions and
informal observations were also recorded.
An interview guide and open-ended questions were used
to conduct 2 h semi-structured interviews. The interview
guide had four sections. The first part referred to the
general information about the firm; the second was focused
on DW profiles; the third investigated the KM tools and
the fourth the worker satisfaction about the KMS. As
regards worker satisfaction, it has been operationalized by
means of a five
TASK ANALYZABILITY
High Low
Low Routine
NUMBER OF 1 2
EXCEPTIONS 4 3
High Non Routine
Fig. 3. Perrow’s model of technology and structure.
Table 2
Case description
A B
Industry Consultancy
Core product E-solutions
Turnover (\000 V) 36,872
Number of employees 452
Number of DW 300
Investigated process Project management and carrying
out
point Likert scale, ranging from very low (Z1) to very high
(Z5).
As known, the triangulation of multi information
sources, designed to compensate for any single-source’s
weakness, is an effective way of increasing both reliability
and validity in qualitative research. Thus, data collection in
this study relied on interviews, observations and archival
documents. The study respected Kanter’s (1977) suggestion
that different sources of data can be used to validate one
another. Multiple informants were interviewed thereby
providing further opportunity for triangulation. These
informants come from all hierarchical levels.
Wherever possible, documents were consulted in order
to gather information about the KMS development to
discover any additional important elements that may have
been missed. This is also a way to further confirm or refuse
data obtained through other means and to guard against
bias in informants’ accounts (Leonard-Barton, 1990).
Data were analysed by first developing individual case
studies using the interviews and the record of impressions
and then comparing cross-cases to investigate the relation-
ship between DW activities and KM tools.
3.2. Empirical findings and test of the hypothesis
3.2.1. Case A
A is a leading Italian system integrator company. It
specializes in complete financial solutions, e-security,
intranet applications and web integration; everything from
feasibility studies and the drawing up of an e-Business
strategy, right up to the development and integration of
software solutions.
A started its internal KMS in 1998 and now it has
become a standard model for its customers. The solution
includes, in
addition to the standard intranet functionalities, specific
tools to support communities online.
A employs 300 consultants operating out of multiple
locations and frequently on customer sites. The primary
unit of work is a project that is carried out specifically for
one customer. Customers include private companies
(mainly banks and insurance companies), public services
and local governments in Italy. Common projects include
the design and development of Intranet/Extranet/Internet
portals, content and Knowledge Management solutions, e-
business
integration, e-security, e-ERM/PRM/CRM (electronic
C
Telecommunication Energy services
GSM communication system Heating systems management
9,021,805 400,000
10,660 2360
1000 450
Sales Maintenance
employee/partnership/customer relationship management),
e-learning and e-marketing applications. In order to
completely satisfy the customers and to guarantee the full
integration of the new solutions into existing systems, the
company has to work closely with its customers. An
important part of its business strategy is developing long-
lasting customer relationships. In terms of task complexity,
consultants in A have to face a high variety of situations
and exceptions that need customized and flexible solutions.
In 1998 the company started to design and develop a
new organizational model in order to manage the
knowledge of its highly dispersed consultants and to pursue
continuous improvement of internal processes and
performances. Such a model, supported by the intranet, is
based on the creation of improvement groups, the ‘cubes’,
aiming at suggesting and implementing improvement and
growth plans, in the company strategic areas, according to
EFQM (European Foundation for Quality Management)
that A joined in 1999. Each cube includes a limited number
of employees, selected by the top management and headed
by a chairman and a secretary. The chairman, directly
chosen by the CEO, the latter who sponsored the project
from the beginning, is responsible for the success of
the cube. He coordinates activities, controls the
newsgroup, approves the documen- tation and supervises
the contents. The group interacts through both face-to-face
and computer-mediated meetings. Each cube publishes the
results of its activity (white papers, frequently asked
questions) on the intranet and interacts with all the
employees through discussion newsgroups. Periodically,
the performances of each cube are evaluated measuring the
participation of the employees, the work in
TASK ANALYZABILITY
High Low
Low C
NUMBER OF 1 2
EXCEPTIONS 4 3
High
B A
Fig. 4. Case positioning in Perrow’s model.
progress and the achieved results. In case performances are
not satisfactory, the cube may be closed.
The newsgroup management requires high investments
both in terms of human resources (about one third of the
consultants is directly involved in the cubes) and of money
(the system is constantly improved by a new release every
2 months). The company’s top management has been very
satisfied with the results obtained till now by the cube
model. Besides promoting the ongoing improvements, the
cube model also changed progressively the organizational
relationships by intensifying horizontal communication and
making the vertical one easier. The cube model is now well
consolidated and A recently decided to launch it on the
market. As regards worker satisfaction, the users expressed
the following rates: 3.35 for the content management in the
corporate portal; 3.41 for the graphical interface and 2.95
for the speed in content retrieval. Overall user satisfaction
appears to be in the middle of the scale.
The KMS has a network structure that enables horizon-
tal/vertical communication among participants suiting the
employees’ communication needs. As an interviewee
underlines: ‘The newsgroups are open to everyone and
there are no barriers to communication: you can share
knowledge/thoughts even with people you have never met
and coming from all hierarchical levels’.
According to Perrow’s guidelines for non-routine tasks,
the KMS in A does not provide a centralized knowledge
storage-retrieval system. In fact, a company that sells
customized services-that is, most of its work goes toward
meeting customers’ unique needs-takes limited advantage
of codified knowledge.
However, some criticisms emerge concerning specific
weaknesses of the KMS, such as:
(i) Top-down development strategy. As emerged from the
case: the evolution of the system is planned centrally
by the Chief Knowledge Officer and the CEO that
decide the requirements of the new releases. In fact,
according to the worker satisfaction data, 40% of the
respondents complain about the usability of the KMS
because it does not care about the user needs, 29%
complain about interactivity features, 19% about the
content structure and 10% about graphics.
(ii) Rigid hierarchical structure. Tasks inside the cubes are
predefined and performance measurements are strictly
formalized. As emerged from the case: a chairman and
a secretary are nominated and coordinate a team of 8–9
people; the chairman is responsible for the success of
the cube; document templates and workflows are
predefined; performances about the participation of
the employees, the works in progress, and the achieved
results are periodically evaluated and in case they are
not satisfactory, the cube may be closed.
As an interviewee says, ‘The formal control causes
distortion in the use of cubes that tend to become auto
referential. The people that participate in the newsgroups
are the same people that manage them’.
(iii) High specialization of cubes. As regards this issue, an
interviewee notes that: ‘Some cubes are too
specialized and so they are scarcely useful to me’.
According to the worker satisfaction data, 54% of
participants identify the contents as the most urgent
issue to be improved and 27% ask for knowledge
about processes and daily activities, overcoming
functional boundaries. This point is connected to the
knowledge needs of consultants in A. In order to
manage the consulting process in a system integrator
company, a broad range of knowledge is required.
When specialized knowl- edge is needed, experts are
contacted on-demand.
The criticisms expressed by the users give some
indications that can be related to the appropriateness of
the system to the task type. In fact in case A, criticisms
regard the aspects of the system that do not seem to follow
Perrow’s suggestions. Regarding non-routine situations,
Perrow expects that formal bureaucratic systems will not be
effective. Perrow argues that ‘collegial’ structural arrange-
ments are required in these settings. With regard to the
contents, cubes are frequently perceived as too specialized
and consequently not useful. This content structure seems
to follow the bureaucratic model that establishes the high
specialization of tasks inside the organization. In designing
the paths to navigate the corporate intranet, such a model
also seems to have been adopted by A.
Thus, guidelines in case A could be firstly shifting the
architecture to a more informal one, secondly avoiding a
too strict top-down control in favour of a self and peer
group control process. Furthermore the company, by means
of the KMS, should promote socialization processes that
could lead to the emergence of spontaneous knowledge
commu- nities overcoming the boundaries between
organizational functions.
Based on both positive and negative findings about the
effectiveness of the KMS used by A, we can therefore
confirm hypothesis 1 and refine it as follows:
Hyp. 1.1. in non routine situations, KMS based on formal
and bureaucratic controls (procedure guides and operating
manuals) and/or the use of superior authority to monitor
individual actions are not expected to be effective.
3.2.2. Case B
Company B is the leading European GSM operator and
leader in the domestic market with about 24.6 million lines.
Due to its former monopolistic position, B played a trail-
blazing role in the mobile telecommunication industry, both
opening a mass market for the mobile business in Italy as
well as making Italy itself one of the most advanced
European markets in terms of penetration and supply of
advanced services.
 In order to keep the leadership in a market in which the dialogues with customers or
penetration level of the mobile voice services is already
very high and competitors are more and more numerous, B
competes on innovation both at the technological and the
service level. It consolidates its own leadership in the
traditional voice business and concurrently becomes one of
the fundamental players in the development of the whole
telecommunication industry.
In the last years, B tried to enlarge its presence also
outside Italy, particularly in Europe and in South America.
In Italy, B’s sales force is divided into three groups.
Large firms are directly followed by dedicated account
managers, while for small and medium enterprises (SMEs),
B adopts an indirect sales force, composed of
approximately
150 agencies (Business Promoters), employing almost 1000
agents. On the other hand, final customers are contacted by
means of shops and e-commerce. The case focuses on B’s
indirect sales force.
At the end of 2000 the company was faced with the
necessity to improve the weak relationship with its agents
and its poor knowledge of them. This was made clear by
their high turnover rate (greater than 50%), their unsatis-
factory degree of qualification and their low commitment to
the firm. In order to improve this situation, B introduced a
web-based platform (B_plat) aimed at creating and devel-
oping a virtual community among the agents. The challenge
was to make B_plat an environment where the agents could
train, exchange experiences and acquire information
regarding the company and its products. In other words,
the aim of B_plat was to give the agents the social
dimension that they risk to lose and to build direct
communication channels with them, therefore, avoiding
the interface of the agency owners.
The use of the B_plat was encouraged by means of a
game rewarding those users who proved to be more active
in information sharing. The possibility of having a prize
and, most importantly, the satisfaction the agents got in
reaching the highest rank and boosting colleague esteem
proved to be the most stimulating factors to game
participation.
Since its introduction, the platform received even more
success than expected, as was demonstrated by the rapid
growth in the users as well as usage during after work
hours. The forum, where the agents have the opportunity to
ask questions/help and tell their own experience,
immediately became the heart of the system. Soon the
agents became proactive in proposing and asking for
changes to the platform. Among the most important
changes, that today make B_plat a KMS, the following
should be remembered: (i) the introduction of a channel
with up-to-date information on B products and services;
(ii) the setting up of a ‘direct line’ where B marketing staff
answer to agent questions (such a ‘direct line’ was soon
revealed as an important source of information and
impressions regarding the market); (iii) the introduction of
a news channel for the publication of both information
regarding industry and papers for stimulating
colleagues; (iv) the enrichment of the training channel
with programmes for the development of sales skills.
Nowadays, the platform is composed of five sections:

(i) News, where the most important information is

classified and published.
(ii) Training, where courses (with asynchronous fruition)
are provided.
(iii) Services-Products, containing information regarding B
portfolio and the benchmark with that of competitors.
(iv) Community, containing moderated forums, where
agents can meet, discuss, ask questions and exchange
experiences, as well as information.
(v) Desktop, with a schedule of interesting events and a
selection of useful tools: presentations, ideas for
customers and a service offered by experts providing
agents with supports on commercial, fiscal and
technological matters.

According to the interviewees, worker satisfaction is

very high. The KMS has a network structure that allows
horizontal/vertical communication among participants
really fitting the agents’ communication needs. According
to the interviewees, the system improved the feeling of
belonging to a community as underlined in the words of the
CEO: ‘Since the introduction of the tool, in a 6 month time
period, the agents’ turnover rate reduced from over 50 to
29%’. The connection between the introduction of the tool
and the turnover reduction is also evident by the
enthusiastic feedback from agents. The turnover reduction
together with knowledge sharing and training opportunities
(e-learning) provided by the system, positively impacted on
the agents professional quality, as emerged from some
interviews. The improvement of agents’ skills, enabling
them to better cope with a high number of exceptions, is
coherent with implications of Perrow’s model. In fact, if
task variety is very high (number of exceptions) the number
of procedures required to handle all the exceptions would
be prohibitively costly to document. In order to achieve the
coordination needed for task execution, the organisation
may need to resort to ‘personnel’ controls (Perrow, 1970).
Another aspect that is agrees with the implications of
Perrow’s model, is the emergent feature of the KM solution.
Feedback and suggestions from the field were taken into
high consideration during the development of the system its
following releases. The agents appreciated to be involved in
the continuous improvement of the system. As one of them
reports: ‘Our contributions proved to be determinant for the
success of the tool. Now services and functionalities really
fit our needs’. In fact, where tasks are highly analyzable and
exceptions are many, ‘it is more difficult to predict which
configuration will prove most effective’ (Perrow, 1967).
The system is dynamic, and thus new evolutions are going
to be implemented. For example, one of the main concerns
is decreasing the control of newsgroups and forums. At the
moment, forums are supervised by a chairman that
guarantees the reliability and privacy of published items. As regards (a), control data are collected by means of a
In the near future this may no longer be necessary. WAP (Wireless Application Protocol) technology and
According to Perrow, moving towards a less bureaucratic thus
model fits with the characteristics of the task that is in the
middle of the routine–non routine continuum.
Based on these findings we can, therefore, confirm
hypothesis 1 and refine it as follows:
Hyp. 1.2. in situations characterized by high task analyz-
ability and high number of exceptions, KMS based on non
pre-determined procedures, and person-to-person communi-
cation are expected to be effective.

3.2.3. Case C
C is the Italian subsidiary of the leading European
provider of facility management services to local
authorities and businesses. Its service portfolio includes
the management of heating and air conditioning facilities,
the operation of district heating systems, facility
management for public and private customers, and
industrial energy services. C has locations throughout Italy
and the head office is located in Milan. There are 700
employees, 450 of whom are involved in the maintenance
process. The case focuses on C maintenance staff, most of
whom are mobile except for a small group permanently
located in big plants that require daily control and
maintenance. There are about 5.500 sites managed by C
staff but, despite the high number of customers, task
execution does not involve high variety as C is focused
only on a specific type of facilities (heating and air
conditioning).
In order to improve efficiency and effectiveness of its
operations, the company started a fully in-house developed
Intranet project, sponsored by the Information System
Department, in 2001. Soon after, opportunities towards
information and Knowledge Management emerged, even
though a clear involvement of top management in improv-
ing the first applications is still to come.
There are two main activities supported by the Intranet:
Quality Management and Facilities Remote Control and
Management. In the present paper, only the latter will be
described, as directly related to DWs.
The Facilities Remote Control and Management allows
C:

(i) To collect customer calls through a Call Centre in

Milan and make them available to all the peripheral
offices. A scheduler assigns the call to a regional
office. After the validation of the person responsible
for local scheduling, the call is handed to a
maintenance technician or team.
(ii) To collect real time information from customer
facilities in order to:
(a) better perform heating and air conditioning
control data collection and
(b) better monitor both ordinary and extraordinary
maintenance interventions.
directly transmitted to a database, avoiding a second data forms to be filled), it fully meets
entry, previously made by the headquarter. The new ICT-
enabled procedure reduces the workload for maintenance
staff. It brings an end to collecting data on paper, going to
the regional office in order to fax them and the time-
consuming re-entering data with a high error rate.
Furthermore, an automatic error checking can control the
data entry online. It is worth noting that the control
frequency is very high: some facilities require control
checks on a weekly basis and some others on a monthly
basis. Statistical analysis on these data were usually
performed once a month (in order to send information
to the gas suppliers) but the new system could allow it to
be done more frequently thus enhancing the information
management process of C.
As regards (b), ordinary interventions are driven by a
technical hand-book and a predefined form to fill in, while
for extraordinary interventions technical guidelines are not
provided. The ordinary maintenance plan (in a 2 week time
window) is also available to every technician, together with
the list of all the interventions already done.
The application is essentially perceived as a customer
service improvement. During the maintenance intervention,
in fact, data about technical aspects are collected thus
permitting the customer to control them in real time
through an internet service. Previously, information about
work in progress and work closure were communicated to
customers by means of telephone calls, fax and file
transmissions. The real time customer involvement avoids
misunderstandings and complaints. The above described
applications, while strongly satisfying top management,
were not agreeable to maintenance staff at the beginning as
they perceived them as having an increasing control on
their daily lives.
After an effort among maintenance team leaders to
promote the application, they eventually had to resort to
imposing the adoption. It is worth noting that, after 1 year,
the maintenance staff recognized the usefulness of the
system in their recurrent practices. In order to make the
adoption easier and less traumatic, as an interviewee said,
very simple technical equipment was required: a mobile
phone equipped with WAP technology.
At the moment a formal performance measurement
system has not been developed, but the positive effects on
efficiency and effectiveness have been widely recognized
by all the interviewees.
According to the interviewees, worker satisfaction is
high. The KMS has a structure that, allowing vertical one-
way communication between DWs and headquarter, really
fits their communication needs. According to the
interviewees, the system improved the feedback process
from customers’ sites, making data entry easier and
reducing errors. In fact, the system operates in ‘real-time’.
As an interviewee underlines:
‘In my area, Sardinia, there is only one peripheral office,
which serves a large region. I immediately appreciated the
tool as it helped me in performing my daily tasks’.
Although the tool provides limited and simple
functionalities (one-way com- munication, pre-defined
the user information needs. In fact, where the task is highly
analyzable and with few exceptions (routine), Perrow
argues that organization should rely on procedure guides,
operating manuals, job codification and rigid lines of
reporting. The investments in the system have been low, till
now, but they satisfy the limited needs of the DWs. The
integration with the accounting system could be the next
step in order to improve the KMS.
Also in this case, the relationship among the character-
istics of the activity according to Perrow’s model and the
KMS has emerged quite clearly. It allows us to confirm
hypothesis 1 and refine it as follows.
Hyp. 1.3. in routine situations, KMS based on procedure
guides, operation manuals, task codification and rigid lines
of reporting for controlling employee behaviour are
expected to be effective.
4. Implications for management action and
further research
This paper aimed at exploring the relations between
dispersed worker activities, KM tools and worker
satisfaction. The DWs were described in terms of the work
they do, using Perrow’s model of technology and structure
and user satisfaction was considered a proxy of the
system’s appropriateness.
Empirical findings where derived from three case
studies representative of the three main types of knowledge
workers
Table 3
Comparative analysis
that we find as relevant from a survey, namely
consultants, sales personnel and field force technicians.
Tasks performed by these workers were analysed using
the dimensions of analyzability and variability suggested
by Perrow’s model.
Results from the three case studies, summarised and
compared in Table 3, confirmed that there is a connection
between the characteristics of the tasks and the
characteristics that a KMS should have in order to
contribute to business performance and worker satisfaction.
In particular, we can elaborate on our starting hypothesis
as follows:
Hyp. 1.1. in non routine situations, KMS based on formal
and bureaucratic controls (procedure guides and operating
manuals) and/or the use of superior authority to monitor
individual actions are not expected to be effective.
Hyp. 1.2. in situations characterized by high task analyz-
ability and a high number of exceptions, KMS based on
non pre-determined procedures, and person-to-person
communi- cation are expected to be effective.
Hyp. 1.3. in routine situations, KMS based on procedure
guides, operation manuals, task codification and rigid lines
of reporting for controlling employee behaviour are
expected to be effective.
In this study, we tried to reduce many of the reliability
problems that are inherent in qualitative research using:
multiple informants from different hierarchical levels,
triangulation of different types of data sources and a
systematic data analysis.

A—Consulting B—Sales C—Maintenance

Main expected result Improve efficiency and effectiveness Enhance DW social Improve efficiency and
dimension effectiveness
Organizational impact Impact on internal communication Impact on power map Impact on workload
distribution
Main sponsor CEO Sales executive Information System
executive
ICTchoices ICT tools Intranet (Plumtree, Webcubee, Documentum, Internet (UNITIM, Intranet (also via
Cruisenete, Digital Think, Applix, Eyenete) SimToolkit, WAP, WAP)
GPRS)
ICT tool development In-house Outsourced In-house
Top-down Bottom-up Top-down
High investment High investment Low investment
Dynamic (a new release every 2 months) Dynamic (emergent) Static
Organizational choices Information flow Network Network One-direction (periph-
ery to headquarter)
Integration with other No Marketing IS No
IS
Managerial choices KMS adoption All the DWs were forced to adopt the system Adoption on voluntary All the DWs were
basis, fostered by forced to adopt the
managerial levers system
(awards)
Control system On frequency of use and contents No On frequency of use
Formal performance Yes Yes No
measurement
Worker satisfaction Medium Very high High
 Generalizability remains more of an issue. Further Bass, San Francisco.
research is needed to prove whether this finding can be Furst, S.R., Blackburn, D., Rosen, B., 1999. Virtual teams: a proposed
research agenda. Information Systems Journal 9 (4), 249–269.
verified in similar firms. However, although they must be Gupta, A.K., Govindarajan, V., 2000. Knowledge management’s
treated with caution, the findings of this exploratory study social dimension: lessons from Nucor steel. Sloan Management
are nonetheless suggestive. Review 42 (1),
Implications for management actions are telling: 71–80.
Perrow’s
model can be used to understand the characteristics of the
task performed by dispersed workers and provide clear
guidance for the design of the KMS in terms of degree of
control and codification. The study of the three cases
shows that worker satisfaction (in terms of usefulness,
appropriate- ness to the business and frequency of use) is
low with certain aspects of KMS that do not match with the
organizational suggestions derived from Perrow’s model.

Acknowledgements

The authors are especially grateful to Dr George

Hayward and to the anonymous reviewers for their useful
comments and suggestions on the paper.

References

Becker, M.C., 2001. Managing dispersed knowledge: organizational

problems, managerial strategies, and their effectiveness. Journal of
Management Studies 38 (7), 1037–1051.
Chang, R., Chang, Y., Paper, D., 2003. The effect of task
uncertainty,
decentralization, and AIS characteristics on the performance of AIS:
an empirical case in Taiwan. Information and Management 40, 691–
703.
Choe, J., 1996. The relationship among performance of
accounting
information systems, influence factors, and evolution level
of
Information Systems. Journal of Management Information
Systems
12 (4), 215–239.
Cohen, W.M., Levinthal, D.A., 1990. Absorptive capacity: a new
perspective on learning and innovation. Administrative Science Quar-
terly 35, 128–152.
Collins, H., Kusch, M. (Eds.), 1998. The Shape of Actions—What
Humans and Machines Can Do. MIT Press, Cambridge, MA.
Corso, M., Martini, A., Paolucci, E., Pellegrini, L., 2004a.
Knowledge management systems in continuous product innovation, in:
Leondes, C.T. (Ed.), Business and Technology of the New
Millennium. Kluwer Academic Press, Dordecht, pp. 36–66.
Corso, M., De Cantis G., Martini A., Massa S., Paolucci E., Pellegrini L.,
Testa S., 2004b. Managing dispersed workers: first results of a survey.
In Proceedings of the Fourth EURAM Annual Conference Governance
in Managerial Life, St Andrews, Scotland.
Cramton, C.D., 1997. Information problems in dispersed teams. Academy
of
Management Proceedings 1997;, 298–302.
Cramton, C.D., 2001. The mutual knowledge problem and its
consequences for dispersed collaboration. Organization Science 12,
346–371.
Cramton, C.D., 2002. Finding common ground in dispersed
collaboration.
Organizational Dynamics 30 (4), 356–367.
Davenport, T., Prusak, L. (Eds.), 1998. Working Knowledge:
How
Organizations Manage What They Know. HBS Press, Boston.
Duarte, D.L., Snyder, N.T. (Eds.), 1999. Mastering Virtual Teams. Jossey-
Haanes, K., Lowendhal, B. (Eds.), 1997. The Unit of Activity: Towards an
Alternative to the Theories of the Firm. Strategy, Culture and Style.
Wiley, London.
Kanter, R.M. (Ed.), 1977. Men and Women of the Corporation. Basic
Books, New York.
Leonard-Barton, D., 1990. A dual methodology for case studies:
synergistic use of a longitudinal single site with replicated multiple
sites. Organization Science 1 (3), 248–266.
Lipnack, J., Stamps, J. (Eds.), 1997. Virtual Teams: Working Across
Space, Time, and Organizations. Wiley, New York.
Maskell, P., 2001. Knowledge creation and diffusion in geographic
clusters. International Journal of Innovation Management 5
(2),
213–237.
Minkler, A.P., 1993. The Problem with Dispersed Knowledge: Firms in
Theory and Practice. Kyklos 46 (4), 569–587.
Nahapiet, J., Ghoshal, S., 1998. Social capital, intellectual capital and the
organizational advantage. The Academy of Management Review 23
(2),
242–266.
Nonaka, I., 1990. Redundant, overlapping organization: a Japanese
approach to managing the innovation process. California Management
Review 32 (3), 27–38.
Nonaka, I., Takeuchi, H. (Eds.), 1995. The Knowledge Creating Company.
Oxford University Press, Oxford.
Pe´rez, M.P., Sa´nchez, A.M., de Luis Carnicer, M.P., 2002. Benefits
and barriers of telework: perception differences of human resources
managers according to company’s operations strategy. Technovation 22,
775–783. Perrow, C., 1967. A framework for the comparative
analysis of
organizations. American Sociological Review 32 (2), 194–208.
Perrow, C., 1970. Departmental power and perspectives in industrial
firms, in: Zald, M.N. (Ed.), Power in Organizations, 59–89. Vanderbilt
University Press, Nashville, TN.
Petrash, G., 1996. Dow’s journey to a knowledge value management
culture.
European Management Journal 14, 365–373.
Roos, J., Roos, G., Dragonetti, N., Edvinsson, L. (Eds.), 1997. Intellectual
Capital: Navigating in the New Business Landscape. Macmillan,
London.
Schiuma, G., Marr, B., 2001. Measuring and managing intellectual
capital and knowledge assets in new economy organizations, in:
Bourne, M. (Ed.), Performance Measurement Handbook. GEE
Publishing Ltd, London, UK.
Sveiby, K. (Ed.), 1997. The New Organizational Wealth: Managing and
Measuring Knowledge Based Assets. Barret-Kohler Publishers, San
Francisco.
Tsoukas, H., 1996. The firm as a distributed knowledge system: a
constructionist approach. Strategic Management Journal 17, 11–15.
Wiig, K., 1997. Integrating intellectual capital and knowledge management.
Long Range Planning 30, 399–405.
Withey, M., Daft, R., Cooper, W., 1983. Measures of Perrow’s work unit
technology: an empirical assessment of a new scale. Academy of
Management Journal 1983;, 45–63.

Mariano Corso, PhD is Associate Professor of Business

Economics and Organisation and Management at the faculty of
System Engineering, Politecnico di Milano where he chairs the course
of Management Engineering at the Cremona site. At the Business
School of Politecnico di Milano he is director of the Master in
‘Management and Organisation Development’ and of the
Observatory on Intranet applications. He promoted and co-ordinated
national and international researches on Knowledge Management. He
is board member of the Continuous Improvement Network (CINet)
and co-ordinator of the PhD network. He is author of more than 70
publications at the international level.
Antonella Martini, PhD is Assistant Professor at the faculty of
Engineering, University of Pisa where she teaches ‘Innovation
Manage- ment’ and ‘Business Economics and Organisation’. Her
main research interests concern Knowledge Management and
Continuous Innovation: she is actively involved in national and
international researches on the fields and member of the Continuous
Improvement Network (CINet) and EUROMA. She is author of more
than 25 publications at the international level.
Luisa Pellegrini, PhD is Assistant Professor of Management
Engineer- ing at the faculty of Engineering, University of Pisa where
she teaches
‘Innovation Management’ and ‘Business Economics and
Organisation’. She is actively involved in national and international
research projects on Knowledge Management and Continuous
Innovation. She is member of the Continuous Improvement Network
(CINet) and EUROMA and author of more than 25 publications at
the international level.
Silvia Massa, PhD is Assistant Professor of Management
Engineering at the Department of Communication, Computer and
System Sciences, University of Genoa (Italy), where she received
her PhD. Her main research interest is knowledge management in
organizations and she is actively involved in national and
international researches on the field.
Stefania Testa, PhD is Assistant Professor of Management
Engineering at the Department of Communication, Computer and
System Sciences, University of Genoa (Italy), where she received her
PhD. Prior to joining the University as a faculty member, she was a
consultant in a leading American consultancy firm. Her main research
interest is information systems adoption in small and large
enterprises.
View publication stats