Chapter VIII

The Planned and Materialized Implementation of an Information

System
Pekka Reijonen, University of Turku/Laboris, Finland
Jukka Heikkilä, Helsinki School of Economics, Finland
EXECUTIVE SUMMARY
The object of this case study is a marketing and sales information system
in two local offices of a regional telephone company. A unified, advanced
client/server system was needed due to the merging of three companies
into a bigger regional company, keener competition, and the growing
complexity of the services provided. The system is tailormade to meet the
needs of the industry and it was developed by a software vendor in close
cooperation with the nationwide alliance of regional telephone
companies. This study illustrates the difficulties in simultaneously
aligning an organization and implementing a new information system.
Views on the skills and competence needed in using the system vary, and
lead to the negligence of education and training. The consequent lack of
skills and knowledge of some users, especially of those not using the
system regularly, create profound problems in the whole work process
and in productivity as the first, obvious work practices become the
dominant mode of operation bypassing the desired integrated workflow.
The findings are discussed and reflected to concepts of
institutionalization, positive reinforcement, and productivity paradox.
This case emphasizes the importance of the organizational
implementation and adaptation process which ought to begin after the
implementation of the technical system.

BACKGROUND
 It was only in the beginning of the 1990s when the
telecommunications sector was deregulated in Finland. Long distance
lines (i.e., crossing the local telecommunication areas), which were
previously operated by a state-owned company, opened to competition.
Deregulation took place simultaneously with the emergence of radio-
based telecommunications, most notably with the introduction of the
wireless analog NMT and later digital GSM and DCS networks. Radio-
based telecommunication is also open to competition. The mobile phone
has grasped a significant share of the phone traffic (there are more than
2.3 million mobile phone subscribers in a country of 5 million inhabitants
and 2.8 million subscriber lines). As a consequence, the rates of hard-
wired long distance calls have fallen by 80% and the rates of local calls
by 50% since deregulation. Deregulation has affected the regional
teleoperators the least, but as the former licensing of
telecommunication areas is deregulated, too, local operators have
started to merge.
Currently there are two main telecommunications operators in
Finland: Sonera plc (formerly known as the state-owned Telecom
Finland) and the Finnet consortium, which is owned by 46 regional
telephone companies. There is a third player, owned by Swedish Telia,
but as its market share is less than 2%, long-distance and radio-based
telecommunication markets have practically a duopoly. Our case
company belongs to the Finnet consortium, the market share of which is
about 50% of the total telecommunications turnover, 16 billion FIM
(about 3 billion USD), and the market is expected to grow to 22 billion
FIM by the year 2000.
The nationwide Finnet consortium was established by the regional
telephone operators in order to provide seamless long-distance calls for
their subscribers and to keep up with the pace of fast rapidly-developing
technology. The consortium developed a digital SDH-based backbone
network, primarily on optical cabling, for long-distance and mobile-voice
calls and data communication.
The regional telephone company, called here Areal Phone Ltd. (AP),
implemented a new information system (IS) about one year before this
case study. The implementation was not, actually, a free choice, rather
the company was driven to changes. The reason for the new IS was
simple: AP is the result of a merger of three mutual, local telephone
companies. The three companies each had their own tailor-made
operative information systems, so there was a need to unify the
operations and ISs.
The reasons for the merger stem from the previously mentioned
changes in the telephone operators’ business environment. On the one
hand, deregulation has driven small local companies to seek economies
of scale by merging. On the other hand, it is considered difficult for small
companies to keep up with the rapid development of digital telephone
technology, where different data services are becoming more and more
important (for example, the percentage of Internet users in Finland is
one of the highest in the world).
At present AP serves about 74 000 wire-based telephone
subscribers and 26,000 cable television (CATV) subscribers (see Table 2).
The company functions in 23 municipalities, where there are 152,000
inhabitants. The private customer base consists mainly of urban
households but a significant number of customers live in the countryside.
This means that maintaining the infrastructure is a demanding and
expensive responsibility of the telephone company. The digitalization of
the telephone network that started from switches has continued at the
subscriber end and the number of ISDN installations has been growing
constantly. For example, in 1997 the turnover of AP data services grew
over 30%. Data services are especially important in enterprises, which is
the other important customer sector of AP. The possibilities of AP
offering data and other add-on services are excellent because their
network is totally digitized.
The main business of AP is still to maintain the telephone network
and operating calls. The future trend is to increase the complexity of
value-added services both for the private customers (e.g., CATV-network
renewal for data communications and xDSL connections) and enterprises
(e.g., running outsourced switchboards and high-speed data
communications). However, as the hype for an information society
continues, telephone operators are seen as vital players; hence, the
interest for mergers continues: in 1997, AP acquired the telephone
company of a town situated nearby. In 1997, AP also introduced their
own radio-based DCS/GSM 1800 city phone network.
 SETTING THE STAGE
The new information system at AP was seen as a means to cope
with the competition and to improve marketing, sales, installation, and
maintenance of the increasingly complex value-added services. As we
browsed through the documentation, it became clear, however, that the
project had concentrated only on defining data structures and little
attention was paid to actual business concerns and work process
(re)organization.
The selection process of the new IS for AP had taken about one
year. The turn-key information system chosen had not been used earlier
in any of the three separate companies. The coalition worked in close
cooperation with a Finnish representative of a large international
software vendor. The new IS for operations control was to cover
practically all the functions of the company (see Figure 1).
The broad domain was not news to any of the companies, because
their former ISs covered also the majority of operations. As the new IS
was different from any of the three former ISs, the data conversion from
the old ISs was anticipated to become one of the most demanding tasks
of the implementation process.
The new IS was implemented on an Oracle database using Oracle
Forms and Report Writer development tools. The most computing
intensive routines were programmed in C++. The system runs on a UNIX-
server and the workstations (PCs) are connected to the server via an
ethernet-connection. The OC system is used with a terminal program, so
the architecture actually corresponds to a mainframe system. Office
programs, however, are used directly from the server.
The initial problems with the capacity of the server were due to the
greater-thanexpected number of concurrent connections. The number
of connections grew high as the IS did not support incomplete tasks (i.e.,
if a task was interrupted all the previous input was lost and the task had
to be restarted from the very beginning). Hence the users opened several
concurrent terminal sessions in order to manage their work fluently in a
customer service situation. At the time of the study, this capacity
bottleneck had been removed and the response times of the system
reached an acceptable level.
However, there arose additional, “fuzzy” problems that led the CIO
of AP to contact our research unit at the university to evaluate the
situation at the marketing and sales subsystem. The perception of the
problem by management was vague; it was based on casual user
complaints about the poor usability of the system. According to
management, the problems concerned mainly the private customer
sector and were caused primarily by deficient features of the IS: the IS
did not support business processes anymore because it had been
designed for selling hard-wired connections instead of addon services.
Therefore, management decided to set up a project for streamlining the
interface in order to improve the functionality of the IS by “reducing the
number of screens by half.” Our evaluation was planned to complement
the requirements analysis of the forthcoming renewal project. In
addition, management wanted to compare the IS usage and usability
problems to earlier evaluations made by our research unit.
The subsystem studied (see Figure 1) was labeled “Marketing and
Sales,” but actually “Customer and Product” would better describe the
system. This is because the marketing part of the subsystem was actually
abandoned — mostly for its lack of usability and functionality — after a
short trial period. The subsystem is linked to all other subsystems (see
Figure 1) because it is used for maintaining customer records (e.g.,
names and addresses, and long-term contracts).
 The goal of the evaluation was to describe end users’ work tasks,
work processes, and the use situation of the IS, so that more informed
decisions about further development of the organization and the IS could
be made. The study was carried out at all three regional offices of AP, but
we use two offices of comparable size and departments to illustrate our
case.

CASE DESCRIPTION
In both offices sales function of AP is organized according to their
clientele, into Private and Enterprise departments. There were 15 end
users in the Private Subscribers Department and 11 in the Enterprise
Department. Management named these two groups as the target groups
of the evaluation. It soon turned out, however, that about an equal
number of end users are using the same subsystems outside the
Marketing and Sales Department, and their IS usage is directly related to
the work processes of the personnel in marketing and sales.
The average age of the users is about 37 years in both departments.
In the Private Department, nearly all users are women, but in the
Enterprise Department the proportion of women is only about 20%. The
end users are competent long-term employees and are accustomed to
the organizational culture: mean employment time is nine years in the
Private Department and 14 years in the Enterprise Department.
The users are also experienced computer users (on average, 8-10
years of usage), even though most users have experience only in the
systems used in the current work place. In the Private Department, the
IS is the main tool of its end users — on average, 87% of the total
computer usage, which is equivalent to about 74% of the working hours
spent using the new information system. In the Enterprise Department,
the average usage time is 48% (about 25% of the working hours) and the
variance in usage time is considerably higher. The situation in which the
IS is used is also different: in the Private Department, most of the usage
(40% of the working hours) took place in customer service situations,
whereas computer usage in the Enterprise Department is more back-
officelike; in only 4% of the working hours was the computer used while
dealing with customers.
Before the IS implementation, the program vendor first trained and
educated a few principal users who in turn trained their peers. The
average training time was eight days in the Private Department and five
days in the Enterprise Department — both figures can be considered
impressive in light of our earlier research (Laboris, 1997) — despite the
fact that in the Enterprise Department the training is unevenly
distributed and over half of the users had received less than four days of
training. The reported reason for missing the training courses was urgent
duties but, on the basis of interviews, a lack of motivation seemed to be
the major reason.
Technological Concerns
 The new IS has a text-based interface managed with the keyboard.
This solution was decided on in the cooperative planning association long
before the implementation of the system. The decision was justified by
the requirement that “the appearance of the interface must be the same
in all kinds of workstations like PCs, X-terminals, and ASCIIterminals.”
Even though consistency and uniformity are important aspects in
interface design, the grounds for this decision are not relevant to AP
because all the workstations are PCs. Contrary to its initial idea, the
decision eventually limited the possible range of interface options (e.g.,
it eliminated the possibility to implement a graphical user interface) and
the development tools.
As a result, the system is somewhat difficult to interact with — or,
actually, there is little interaction. The system consists of 90 separate
screens, and the coupling of the individual screens to the process of
completing a specific task is weak. For the end users this means that they
have to memorize the screens and in which order they are used for each
specific work task. This makes the system hard to learn and much
practice and repetition is required in order to master the system fluently.
This is confirmed by the fact that users’ estimates of the time required
to learn to use the system to the extent that one can work efficiently vary
from two months to two years (the average was nine months). The
estimates of the average learning time differ in the two departments: the
average estimate of the learning time is four months in the Private and
13 months in the Enterprise Department. As most of the learning takes
place during system usage, it is obvious that the perceived learning time
is correlated with the amount of system usage (r = 0.59, p = 0.01, n = 18).
At the time of the study (one year after its implementation), users rated
their current ability at 7.7 (on a scale ranging from 4 to 10) in the Private
Department and at 6.8 in the Enterprise Department.
 Users are unanimous that the system has been, and is, difficult to
learn and to use. There are too many screens and there is little help built
into the system to find the right screens in the right order. Obviously, the
IS supports occasional users poorly, but with a sufficient amount of
practice, the user eventually can learn it by heart. One clerk who has
been using the system for five to six hours a day for one year expressed
this by saying:

But the system usage is obligatory for less often or irregularly

performed tasks, too. In these cases, it has become a common practice
to write the necessary information on a piece of paper and key the data
in the IS later (with help from co-workers or support personnel).
The overall evaluation of the system by an EUCS-measure
demonstrates that all features of the system are rated lower in the
Enterprise Department than in the Private Department (Figure 2).
 The user opinions give grounds for two possible conclusions: the
system really is not well suited to the work in the Enterprise Department,
and/or the computer selfefficacy of the users is low, i.e., the users feel
that they cannot manage their computer supported work tasks properly
(c.f. Compeau & Higgins, 1995).
Organizational Concerns
The consequence of the poor alignment of the system functions
and the actual work flow is that less-frequent users did not learn to
utilize the system properly. Because some user groups (e.g., supervisors
of the installation function, technicians and some of the sales managers
in the Enterprise Department) received little education in the system
usage and on the ideal planned work flows, they developed apparently
successful, but manual routines, to work around the system’s limitations.
Management launched a redesign project to streamline the
process in the Private Department, but actually the problems were more
severe in the Enterprise Department. In the private sector, the service
procedure was rather simple: customers were either faceto-face or on
the telephone with the clerk who used the IS to carry out the service (see
Figure 2). However, the system did not fully meet the requirements of
marketing, selling, installing, and maintaining more complex services
such as corporate ISDN-subscriptions. The willingness and skills to learn
and use the system were also lower in the Enterprise Department.
In the Private Department, the implementation was carried out
according to the plan: at the time of the study, most end users have
reached such a skill level that the speed of the work process was
acceptable in both offices. The work is interrupted by breakdowns only
occasionally (about half of the users have problems weekly, and 57% of
the problems are due to the lack of skills and knowledge, mostly a
problem for newly enrolled clerks). In other words, there are few severe
problems in the use of the IS. Also the work processes closely follow the
planned workflow. Although the system is clumsy for some individual
tasks, and some operations are not institutionalized yet, the users have
learned to manage their work with the IS and to resolve the few
problems with assistance from their fellow workers. In other words,
private customer service either in direct contact face-to-face or by phone
proceeds rather fluently. Direct customer contact also increases the
clerks’ motivation for developing their skills in system usage because, by
being able to use the IS fluently, they could avoid embarrassing social
situations which might arise if they had to say to a customer: “Well,
actually, I cannot perform my work tasks.”
In the Enterprise Department, there has emerged an unplanned
supplemental work role (substitute IS user in Figure 3) because the sales
managers were unwilling to learn and use the IS. Instead of inputting the
order information into the IS, they deliver orders as handwritten notes
to “the substitute IS user” who then inputs the orders into the IS. The
sales managers’ notes rarely include product numbers, only the name of
the product or service. It is not unusual that while inputting data, extra
contacts between the sales manager and “the substitute IS user” are
needed because the product name the sales manager uses is not found
in the IS.
In the other office, technicians and their supervisors also ceased to
use the IS after a few weeks of initial use. Actually, they ceased to use
the IS as they were told to stop using it. This is because they had received
neither enough training nor enough practice to master the IS as
occasional users. For example, when the technicians were to record in
the IS the working hours spent and materials consumed for a certain
order, they often could not find the right order nor the right customer
number. Instead, they keyed in the data in the first found occurrence of
this specific customer — or created a new customer number. This led to
a situation where some orders were not invoiced, the incorrect
department of an enterprise was invoiced, and the database of the IS
was eroding because of the duplicates. Except for extra work, this also
caused customer complaints and in the present competitive
environment, flawless customer service, quick response, and a good
reputation are critical.

In order to correct the situation, the IT usage of these user groups

was replaced with pencil and paper: after inputting the work order, the
substitute IS user printed out the order, and data of the later phases of
the work process were documented on it. The notes made on the paper
document were then inputted into the IS by the substitute IS user as the
final step of the process after the installation was completed. This meant
new procedures and a lot of extra work in finding missing data, since the
data on the paperbased documents were only occasionally complete. As
a consequence, the invoicing process was delayed dramatically, in some
cases up to several months, despite the constant overtime work of the
substitute IS user.
A closer look at the episodes and encounters of the process that
led to the ban of usage in the Enterprise Department revealed that the
key problem was the end users’ inability to cope with the IS. The
incapability of technicians and supervisors at the other office resulted in
chaos, temporarily solved by the use of paper instead of the IS. However,
this arrangement became permanent. For the sales managers, it was also
a question of their power in the organization: a substitute user was
assigned to the “less valuable” task of learning to use the IS and keying
in the data.
Current Challenges
Even though the situation might have been avoided by managing the
organizational implementation process “better,” there are two
interrelated facts, which have contributed to the evolvement of the
problems.
First, the IS is not user-friendly with its character-based windowed
user interface, and it does not have any support for learning or
remembering the flow of operations. Its functionality for some tasks is
questionable as it was originally developed for a different business
environment. The learning time is long (but not extraordinary) and as a
consequence, some user groups are not willing to adopt the system. And
we are afraid that there will be more problems of this kind. This is
because the way the IS was built seems to have become an industry
standard: a shared relational database is normalized to atomistic tables,
which are updated by using screens and procedures (half) automatically
created from the columns of the table with an application generator. This
easily leads to a great number of separate screens. But, as the starting
point for development is a semantic data model, the design is seldom
compatible with the work procedures of the organization. It is also
difficult to learn to navigate in a bunch of separate, seemingly unrelated
screens.
The second problem is the inertia caused by the IS in a rapidly
changing environment. For example, the IS has been designed so that
there can only be one subscriber telephone number per customer ID.
This means problems with the rest of the possible numbers of an ISDN
line. Or, when a customer enterprise wants to regroup invoices
according to its new organizational structure, there are no mechanisms
for handling the update. It is inevitable that information systems freeze
the current mode of operation and all changes in operations almost
always call for reprogramming of ISs. And in the case of the IS, the
advantage of sharing the costs of development in the first place has
become an obstacle for further development. The software is owned by
the consortium, the members of which have to be convinced and agree
upon the need for changing the software.
To us the case clearly illustrates the common dilemma in IS
implementation: an information system that is somewhat awkward to
use, slightly incompatible in work processes, and difficult to change (but
otherwise technically functional) can be implemented in one
organization whereas in another, almost similar organization, the IS is
abandoned.
DISCUSSION
The design, production, and implementation phases of a computer-
based information system are commonly considered the key activities in
its life cycle — in money, time and other resources. Skillful software
development is about the technical maneuvers in conducting the phases.
However, the value of the IS, how technologically progressive it might
be, can not be realized until it is effectively implemented and deployed.
As pointed out by Kling and Allen (1996), effective use of information
technology does not depend on technology alone: organizational issues
often play an important role. In order to emphasize the importance of
non-technical factors, they use the term “organizational
implementation”: “Organizational implementation means making a
computer system accessible to those who could or should use it, and
integrating its use into the routine work practices” (ibid, p. 269). In the
world of outsourced development, this is more important than ever.
Organizational implementation necessitates, for example, that
users are willing to adopt the IS and that they have the necessary skills
and knowledge (both domain and technology/tools) needed in the
deployment of the CBIS. In research, the need for education and training
in computer-related skills is commonly recognized and its importance
emphasized (Nelson et al., 1995; Clement, 1994). In practice, however,
the training issues are often neglected and it has even been argued that
one of the reasons for less successful CBIS implementations is the lack of
user training (Mitev, 1996).
One reason for the relative ignorance of training and education
issues is that the deficiency of competence usually becomes apparent
only in the beginning of the usage phase of a new CBIS. If the usage of
the CBIS is obligatory, users learn — sooner or later — to deploy the
system while doing their daily work. When users have learned to manage
their work to the extent that there are no apparent problems, the
problems disappear. And, as there are no problems, there is no need to
evaluate the efficiency and effectiveness of the actual implementation
and work practices! Or, only the technical functionality, such as response
times, is checked.
 The time required to learn to use a new CBIS depends on many
factors, for example, the structure and appearance of the CBIS, the
amount of use (i.e., the practice obtained), and the individual learning
capacity. Learning times of operative ISs, such as the OCsystem, are
rather long with huge variation; in most cases the learning period is
several months. The learning process can be characterized as a trial-and-
error diffusion process where a single user or a group of users solve the
problems encountered when integrating their work practices into the
features of the CBIS. It is not a uniform process, rather there are multiple
simultaneous processes (Heikkilä, 1995). In this kind of learning
environment people tend to accept the first functional solution and stick
to it as long as it works. Its effectiveness is never questioned, nor are
alternative solutions actively sought: the functional solution will do. A
typical feature in this learning period is that users search for help mainly
from their peers (in 80-90% of the problems) whereas manuals, help-
desks and other types of support arrangements play only a secondary
role (Heikkilä, 1995; Laboris, 1997). It is worth noting that managers are
seldom in advising roles because they rarely can give support in CBIS-
related problems. As a consequence, managers often only have a vague
idea of the real CBIS deployment situation. Therefore, the effectiveness
of the applied procedures depends mainly on the initially invented
workable solution which becomes the best practice in that environment.
Implementation of a new CBIS generally also changes the division
of labor and incentive structure which cause changes in the work tasks
and motivation of individuals. After implementation users must learn
this new division of labor and potentially some new work tasks as well as
the functions of the new CBIS. The division of labor is planned and
institutionalized rather carefully, therefore, some kind of “role book” is
thus already available when a CBIS is designed. The process of dealing
out the roles to the actors is also part of the management of the
implementation. However, management, control, and evaluation after
implementation are often shallow and based more on beliefs, attitudes,
and opinions than technological or economical data (Kumar, 1990; Kling
& Iacono, 1984). Iivari (1986) has concluded the following about the
intra-organizational implementation of multi-user ISs: there are actually
two coexisting sets of activities in implementing an IS, or innovation in
general; the technically-oriented rational-constructive tasks and a
political process of resolving conflicts of interests and changing the
organization.
In practice, many of the problems, which are caused by the changes
in the division of labor and encountered when deploying a CBIS, are
solved with the same kind of adhoc principle as other work-related
problems. As management’s concerns are in restructuring the industry,
the most important thing is to get the work done by the doers, no matter
how it is done. And, in the same way as an individual repeats an action
after a positive reward, an organization gives legitimate status rather
quickly to an arrangement which seems to be functional (see positive
reinforcement trap; Argyris, 1985). Finally, the division of labor becomes
institutionalized and legitimated (Berger & Luckman, 1966). At this stage,
the way of acting feels to be the right one and the actor becomes
confident: “This is how things are done.” In his book on organizational
culture, Schein (1992) has described the process as follows: “When a
solution to a problem works repeatedly, it comes to be taken for granted.
What was once a hypothesis, supported only by a hunch or a value,
comes gradually to be treated as a reality” (ibid, p. 21).
The insufficient training and the consequent emerging ad hoc work
practices that get institutionalized with the IS can easily lead to
inefficient and unchallenged operations, as our case study clearly
demonstrates. There is growing evidence of the high hidden costs of
using computers in the workplace: about two-thirds of the total costs
seem to come from the invisible opportunity costs in the usage phase of
the computer technology (Heikkilä, 1995; Gartner Group, 1995; van
Hillegersberg & Korthals Altes, 1998). In other words, the out-of-pocket
costs from the IS Department budget represent only one-third of the
total costs. Two-thirds of the costs arise from the users’ working hours
spent in tackling computer-related problems, advising other users,
learning new computer skills, and taking part in various CBIS planning,
development, and coordinating taskforces. And the number of the hours
of education, training and peer-hands-onlearning on applying ISs in work
settings has not kept up with the increase in IS-based activities.
How much of this hassle can be removed by training is still unclear,
but the fact is that it is one of the key actions to diminish the time
wasted. And what has been found so far, is that the traditional teaching
courses are deficient in meeting the needs of users applying ISs in their
work practices. More innovative ways of learning are needed, such as
peer tutoring, support activities, and workshops for the users of ISs
(Reijonen, 1998).
The problems highlighted in this case are not typical in just one kind
of business sector or type of information system, but a more general
trend which we have noticed in several organizations. Hence, the
emphasis is on the level of the general findings and the case is used to
illustrate how problems can be manifested in a particular situation.
ACKNOWLEDGMENTS
This study was financed partially by the grants from The Finnish
Work Environment Fund and The Academy of Finland.
REFERENCES
 Argyris, C. (1985). Strategy, change and defensive routines. Boston:
Pitman. Berger, P. L., & Luckman, T. (1966). The social construction of
reality. Middlesex: Penguin Books.
Brynjolfsson, E., & Yang, S. (1997). Information technology and
productivity: A review of the literature. Advances in Computers, 43, 179-
214.
Clement, A. (1994). Computing at work: Empowering action by ‘low
level users’. Communications of the ACM, 37(1), 53-63.
Compeau, D. R., & Higgins, C. A. (1995). Computer self-efficacy:
Development of a measure and initial test. MIS Quarterly, 19(2), 189-
211.
Doll, W. J., & Torkzadeh, G. (1988). The measurement of end-user
computing satisfaction. MIS Quarterly, 12(2), 259-274.
Gartner Group. (1995). Desktop computing: Management
strategies to control the rapidly escalating cost of ownership.
Gartner Group. Heikkilä, J. (1995). Diffusion of a learning intensive
technology into organisations: The case of PC-technology. Doctoral
Thesis, Helsinki School of Economics and Business Administration A-104.
Iivari, J. (1986). An innovation research perspective on information
system implementation. International Journal of Information
Management, 6, 123-144.
Kling, R., & Allen, J. P. (1996). Can computer science solve
organizational problems? The case for organizational informatics. In R.
Kling (Ed.), Computerization and controversy (2nd ed., pp. 261-276). New
York: Academic Press.
 Kling, R., & Iacono, S. (1984). The control of information systems
developments after implementation. Communications of the ACM,
27(12), 1218-1226.
Kumar, K. (1990). Post implementation evaluation of computer-
based information systems: Current practices. Communications of the
ACM, 33(2), 203-212.
Laboris. (1997). A collection of customer reports on the evaluation
of the deployment of information technology in organizations (1993-
1997). University of Turku, Laboris, Laboratory for Information Systems
Research, Finland (in Finnish).
Mitev, N. N. (1996). More than a failure? The computerized
reservation systems at French Railways. Information Technology &
People, 9(4), 8-19.
Nelson, R. R., Whitener, E. M., & Philcox, H. H. (1995). The
assessment of end-user training needs. Communications of the ACM,
38(7), 27-39.
Reijonen, P. (1998, June 4-6). End-user training and support: A
comparison of two approaches. In W. R. J. Baets (Ed.), Proceedings of the
6th European Conference on Information Systems, Aix-en-Provence,
France (pp. 660-672).
Schein, E. H. (1992). Organizational culture and leadership (2nd
ed.). San Francisco: Jossey-Bass.
Telecommunications Statistics. (1998). The Finnish Ministry of
Transport and Communications. Retrieved from
http://www.vn.fi/lm/telecom/stats/index.htm
van Hillegersberg, J., & Korthals Altes, P. (1998, June 4-6).
Managing IT-infrastructures: A search for hidden costs. In W. R. J. Baets
(Ed.), Proceedings of the 6th European Conference on Information
Systems, Aix-en-Provence, France (pp. 1655-1662)
ENDNOTES
1 At a first glance the mutual telephone companies were geographical
(natural) monopolies. But as the mutual companies were owned by the
subscribers, who also elected representatives to the board of directors,
the monopoly power was never actually exploited to its full extent.
Nevertheless, this led to a price discrimination between leased lines and
“owned” lines, which is currently banned by the European Union
regulation.
2 The subscribers of the conventional mobile phones (NMT and GSM) are
served via a separate, nation-wide company, Radiolinja, owned by the
Finnet consortium.
3 As Finnet operates GSM networks via Radiolinja, the regional
telephone companies are most interested in getting their share of the
radio-based telecommunications by providing geographically limited
mobile services for private and business customers at favorable prices.
4 Preliminary discussions about the development of the system had
actually started several years earlier at Finnet. One of the companies that
merged with AP took part in the requirements analysis. The first
installations took place only a couple of years before this study, and
nobody at AP was experienced with the system.
5 UNIX environment and client/server architecture were selected
because of its openness (it is possible to select between different
software vendors) and versatility (technically easy to extend).
6 The data were gathered with a survey questionnaire (26 end users), by
interviewing end users, managers and IS-personnel (16 interviews), from
documentation (e.g., documents of the CBIS development project, end-
user manuals), by observing the use situation, and by using the training
version of the CBIS.
7 EUCS is End User Computing Satisfaction (see Doll & Torkzadeh, 1988).
8 For example, it takes almost half an hour to register and schedule
installation for a new telephone subscriber.
9 For instance, some Private Department clerks preferred to call the
warehouse keeper for the price of a product instead of checking it
themselves from the database.
10 Both the product numbers and product names were changed when
the new IS was implemented. The new numbering/naming system has
its benefits because it is becoming a standard in the Finnet-group.
11 The design of the IS did not make the situation easier because there
were no practical means for managing enterprise level customer data,
so, for example, the departments of a large enterprise occurred as
separate customers with unique customer numbers in the database.
12 The use of paper-based documents also meant that the IS no longer
was a real time system, so answering customer inquiries of the state of
an order always required contacting supervisors or technicians.
Pekka Reijonen earned his MSocSc in psychology from the
University of Turku. He is a researcher at the Information Systems
Laboratory Laboris, University of Turku. His research interests are
usability and deployment issues of information systems in organizations
and he currently is working on a project defining end users’ skills and
knowledge requirements and their evaluation. His earlier work has
appeared in the British Journal of Psychology, Education & Computing,
Interacting with Computers, and information systems conferences.
 Jukka Heikkilä holds a PhD in information systems from the Helsinki
School of Economics. He is a junior research fellow of the Academy of
Finland at the Helsinki School of Economics. His current research focuses
on the problem of adopting, implementing, and integrating innovative
technologies to support business processes. He is also a member of the
Information Systems Laboratory Laboris, University of Turku specializing
in the evaluation of the utilization problems of implemented IS in
organizations. He has published in the Information & Management,
European Journal of Information Systems, Journal of Systems
Management, Journal of Global Information Management, and
Scandinavian Journal of Information Systems.