Blackwell Science, LtdOxford, UKISJInformation Systems Journal1365-2575Blackwell Publishing Ltd, 200313275297Original ArticleWhy organizations adopt IS process innovationsE Mustonen-Ollila

& K Lyytinen

Info Systems J (2003) 13, 275297 275

Why organizations adopt information

system process innovations: a longitudinal
study using Diffusion of Innovation theory
Erja Mustonen-Ollila* & Kalle Lyytinen
*Department of Information Technology, Lappeenranta University of Technology, PO Box 20,
FIN-53851 Lappeenranta, Finland, email: Erja.Mustonen-Ollila@lut.fi, and Department of
Information Systems, The Weatherhead School of Management, Case Western Reserve
University, Cleveland, OH 441067235, USA, email: kalle@po.cwru.edu

Abstract. This paper identifies factors that affected over 200 information system
(IS) process innovation adoption decisions in three organizational environments
over a period that spanned four decades. The analysis is based on Rogerss
(1995) theory of Diffusion of Innovations (DOI). The results show that several DOI
factors strongly affect IS process innovation adoption. These include user need
recognition, availability of technological infrastructure, past experience, own trials,
autonomous work, ease of use, learning by doing and standards. Yet, a large num-
ber of IS process innovation adoptions followed no discernible pattern.

Keywords: Diffusion of Innovation, empirical research, IS development methods

and tools, adoption decisions, IS process innovations

INTRODUCTION

Despite huge investments in system development methods and tools, these innovations are
not necessarily widely used (Fitzgerald, 1997). Past studies (Sauer & Lau, 1997; Wynekoop &
Russo, 1997) indicate that several organizational, behavioural and contingent factors inhibit the
adoption of such innovations (Prescott & Conger, 1995). Yet, empirical research on information
system (IS) process innovations is largely lacking. Practitioners and academics alike still know
little about what mechanisms are effective and which factors influence IS process innovations.
Failure to address these factors leads to wasted investments and failed processes.
In this paper we represent results of a longitudinal study that investigated the factors that
explained IS process innovation adoptions in three environments over four decades. We apply
Rogerss (1995) theory of Diffusion of Innovations (DOI) in identifying potential factors that
influence the adoption of a method. We use DOI theory because it is well established and
widely used in information technology (IT) diffusion-related research (Prescott & Conger,
1995).

2003 Blackwell Publishing Ltd

 276 E Mustonen-Ollila & K Lyytinen

The reminder of the paper is organized as follows. We first introduce main concepts for our
study, and develop a conceptual framework to analyse IS process innovation behaviour.
Research goals and the research approach are briefly introduced. Then we develop a research
model based on the DOI theory to analyse IS process innovation adoptions (Rogers, 1995).
Thereafter, we analyse over 200 IS process innovation adoptions in the light of DOI.

THEORETICAL FRAMEWORK AND CONCEPTS

Information system process innovation defined

We shall define IS process innovation as any new way of developing, implementing and main-
taining IS in an organizational context (Swanson, 1994). Overall, IS process innovations are
expected to improve the quality and productivity of IS development where the products ISs
are defined as systems of hardware and software capable of digital information storage, pro-
cessing and communication that can serve some organizational functions or purposes.
Generally IS process innovations can be regarded as combinations of normative rules and
resources, which stand at hand, or are acquired into the environment before any development
activity starts (Giddens, 1984). They dictate or channel how the system development should,
or can, be done. IS process innovations are explicit in the sense that they are formalized, and
can thus be transmitted through organizational channels. Accordingly, innovations can be
either adopted from external sources or transferred from internal sources by learning from and
formalizing best practices.
IS process innovations cover not only changes in the technological core of the development
activity like the use of new programming languages or operating systems, but also organiza-
tional or administrative innovations like new project management methods, participative modes
of interactions, or new forms of contracting development work outside the organization. In
Swansons terminology IS process innovations thus cover technological process innovations
(type Ia) and administrative process innovation (type Ib) (Swanson, 1994). In many situations
these changes are inherently intertwined.
Fundamental drivers for these changes are technological innovation in the computing plat-
forms, and organizational learning and experimenting of how to effectively harness the new
computing potential (K. Lyytinen & G. Rose, submitted). We define after Rogers (1995, p. 389)
an adoption as a decision to use an IS process innovation (Sauer & Lau, 1997). Adoptions are
made by some decision-makers, who have resources and the decision rights to change behav-
iours, or control resources associated with development practices.
We can classify IS process innovations into four categories based on their scope, purpose
and content in how they align with technological and administrative innovations. These are
project management and control procedures (M), which are administrative innovations;
description methods (D), which are also administrative innovations; development tools (TO),
which are technological innovations; and baseline technology innovations here called tech-
nology innovations (T). The first category includes rules and administrative procedures that

2003 Blackwell Publishing Ltd, Information Systems Journal 13, 275297

 Why organizations adopt IS process innovations 277

help control, manage and co-ordinate development activities. Examples of type M innovations
are project management guidelines or organizational arrangements like chief programmer
teams (Swanson, 1994). Innovations of type D include notational systems and standards,
which help to describe the development product or process and/or its relationships to the envi-
ronment. Such innovations include well-known standardized modelling techniques like Data
Flow Diagrams, complete methodologies like Unified Modelling Method and the like and pro-
cess modelling approaches like Capability Maturity Model. The innovations of type TO include
all productivity tools for systems development covering application generators, CASE tools,
documentation tools, data dictionaries, or tools to configure or manage software components.
Innovations of type T consist of (externally) developed technical platforms like programming
languages, database management systems and middleware components.
Based on Friedman & Cornford (1989) we classify IS process innovations into several eras.
Friedman & Cornford (1989) point out based on an extensive empirical analysis of the his-
torical evolution of IS development that the four types of innovations discussed above are
often closely horizontally related, and they can be classified into a set of evolutionary gen-
erations accordingly. They point out that the changes in these generations are driven by: (1)
technological changes in hardware and software (type T innovation); (2) changes in types of
systems being developed; and (3) changes in types of users. The two latter are thus external
demand factors that drive the IS process innovation. They are outcomes of attempts to harness
the new computing capabilities into new organizational domains and tasks what Swanson
(1994) calls type II and type III innovations.
We will recognize accordingly four generations of IS process innovation. The first generation
(from the late 1940s to the mid-1960s) is largely hampered by hardware constraints, that is,
hardware costs and limitations in its capacity and reliability. The second generation (from the
mid-1960s to the early 1980s), in turn, is characterized by software constraints, that is, poor
productivity of systems developers and difficulties of delivering reliable systems on time and
within budget (lack of D and TO innovations). The third generation (from early 1980s to the
beginning of 1990s) was instead driven by the challenge to overcome user relationships con-
straints, that is, system quality problems arising from inadequate perception of user demand
and resulting in inadequate service (lack of M, D, and TO innovations). Finally, the fourth gen-
eration (from the beginning of 1990s) was affected by organizational constraints (lack of M
innovations). In the latter case the constraints arise from complex interactions between com-
puting systems and specific organizational agents including customers and clients, suppliers,
competitors, co-operators, representatives and public bodies (Friedman & Cornford, 1989).

Earlier information system process innovation research

Earlier research on IS process innovation shows that technological changes, internal learning
and communication channels to intermediating organizations influence adoptions. Declining
hardware costs and increases in its capacity and reliability (Friedman & Cornford, 1989),
changes in the IT functionality (Friedman & Cornford, 1989) and changes in communications
(McKenney et al., 1997) have been identified as key drivers of technological change.

2003 Blackwell Publishing Ltd, Information Systems Journal 13, 275297

 278 E Mustonen-Ollila & K Lyytinen

Studies on knowledge transfer mechanisms show that the scope and intensity of information
sources and communication channels affect adoptions (Nilakanta & Scamell, 1990). Moreover,
organizations develop their own IS process innovations in type D innovations by garnering best
design experience and formalizing it (Tolvanen, 1998). Such innovations seek to overcome lim-
itations in the existing description methods, their inadequate tool support and the paucity of
external method knowledge. Fitzgeralds (1997) findings show that internal learning is often
preferred because of the usefulness of such knowledge. Thus, designers typically do not object
new methodological guidelines and they use a new methodology if it is useful.
IS process innovation adoptions have been studied over the years by a number of IS schol-
ars though no systematic tradition has yet to emerge. Most of these studies are limited in terms
of number of decisions studied, types of innovations covered or types of factors included in the
study. Past studies include among others Huff & Munro (1985), Kozar (1989), Nilakanta & Sca-
mell (1990), Premkumar & Potter (1995) and Sauer & Lau (1997) (see Table 1).
We did not find studies that focused on IS process innovation adoptions involving a longi-
tudinal perspective with several adoption environments and factors. Previous studies have paid
attention to a specific decision category, but largely ignored the importance of sites where the
adoption decisions are made. Previous studies have also neglected task factors, and concen-
trated on specific groups or individuals as adoption-related decision-makers. Only Premkumar
& Potter (1995) covered multiple time periods that need to be observed during the adoption
process.

Diffusion of Innovation theory and information system process innovation adoption

We adopt the DOI theory (Rogers, 1995) as a theoretical basis to identify and analyse factors
that affect IS process innovation adoption. Based on DOI, we distinguish two broad sets
of activities in the innovation process: initiation and implementation. The adoption decision
separates initiation from implementation (Rogers, 1995), and it involves the primary activity
through which innovations are taken into use in the adopting units. Figure 2 summarizes Rog-
erss (1995) model of DOI. Overall, he identifies five sets of characteristics, called factors, that
affect innovation adoption. These are: (1) innovation factors; (2) individual factors; (3) task fac-
tors; (4) environmental factors; and (5) organizational factors. Because each factor is further
decomposed into multiple items (traits), Rogerss (1995) model incorporates a total of 28
attributes (see Table 2). Many of these items are perceptional measures. In DOI studies these
factors represent independent variables and the dependent variable is the likelihood or the pro-
pensity to adopt an innovation.
In the context of IS process innovations based on some of the earlier studies, we can for-
mulate the following research questions based on the DOI model (see Figure 1):

1 What DOI factors affect IS process innovation adoption, and to what extent do they change
overtime?
2 Do factors change overtime across different types of IS process innovation types and
generations?

2003 Blackwell Publishing Ltd, Information Systems Journal 13, 275297

 Table 1. Related research on information system (IS) process innovation adoption

Diffusion of Innovation
Source Focus factor/item Organization(s) Time period(s) Decision-maker(s) Main results

Sauer & Lau Methodology Organizational/ One One IS developers, IS Direct role of business
(1997) adoption (D, M) Managers role managers decision-makers
Environmental/ Business
pressures
Premkumar & CASE technology Organizational/ A product One Several IS managers All five factors were observed
Potter (1995) adoption (TO) champion, strong top to affect adoptions
management support,
lower IS expertise
Innovation/Relative
advantage, Cost
effectiveness
Huff & Munro Information Organizational/ Employee 10 One Manager of IS planning/ Organizational actions
(1985) technology and curiosity, repetition technical systems/ management commitments
adoption (TO, T) factor users, co-ordinator of were the main factors
special projects influencing the adoption
process
Environmental/ Other

2003 Blackwell Publishing Ltd, Information Systems Journal 13, 275297

organizations, vendors
Kozar (1989) IS description Individual/ Potential 11 One System analysts Only personal variables were
method adopters personal used as predictors
adoption (D) factors
Nilakanta & Database design Organizational/ Different 22 One Designers responsible At the individual level
Scamell (1990) tools and communication for database sophisticated innovations are
methods (TO) channels and sources, applications not influenced by
technical support communications. At the
organizational level the
Why organizations adopt IS process innovations

importance of communications
is demonstrated
279
 280 E Mustonen-Ollila & K Lyytinen

IS process IS development
innovation generations
types

Research question 2
Diffusion
factors

Research question 1

IS process
innovation
adoption

Figure 1. An information system (IS) process innovation adoption research model.

FIELD STUDY ON INFORMATION SYSTEM PROCESS INNOVATION ADOPTION

Research methodology

Because not much is known of IS process innovation adoptions overtime and across different
types of innovations a qualitative case study was deemed applicable (Johnson, 1975; Curtis
et al., 1988; Laudon, 1989). We followed a multisite case approach as our goal was to under-
stand the role of the organizational environment in adoption behaviours. Unfortunately, col-
lecting a representative data set by following a time-dependent vertical research design that
involves several organizations is difficult to carry out because of resource and access limita-
tions. Therefore, we limited our sample to three locales that were, or had been at some point,
part of the same company. Their origins were in the same company, though the company was
divided into two separate Finnish companies (where the other was later further divided into two
separate Finnish companies). A more detailed description of the evolution of the companies
can be found in Appendix 1. By doing so we had some control over some external factors that
could affect IS process innovation adoption. For example, economic fluctuations and market
changes affected all studied sites equally, as their resources were almost the same. Yet, these
three locales were independent in making decisions regarding their IS development, imple-
mented different IS and were physically located in three separate cities 150 km apart.
We followed a descriptive case study (Yin, 1993) approach in that the collected data set
embodied time, history and context of the sites. Being a longitudinal study it involved multiple
time periods (Pettigrew, 1985; 1989; 1990; Barley, 1990; Heiskanen, 1994). Because the bulk
of the gathered data was qualitative consisting of interviews and archival material, we followed
historical research methods when necessary (Copeland & McKenney, 1988; Mason et al.,
1997a,b). Our definition of an IS process innovation adoption formed the basis for interviews

2003 Blackwell Publishing Ltd, Information Systems Journal 13, 275297

 Why organizations adopt IS process innovations 281

and collecting data on adoption behaviours using archives. The data was mostly gathered
between February 1995 and May 1997.
The archival data covered the years 196097, and included interviewees private and public
documents about IS process innovation adoptions. They served as primary and secondary
sources of data (Jrvenp, 1991). Other empirical data contained tape-recorded semistruc-
tured interviews dealing with the experiences of adopting and using IS process innovations. The
interviews were carried out both with managers who were in charge of software development
in the companies and with some software developers who worked at those companies at that
time. The interviewees had been involved in the multiple IS process adoption situations and
events, and their working careers extended over periods of 1030 years in the studied orga-
nizations. We also gathered published news about changes in organizations environments and
examined documentation of developed systems, system development handbooks, etc.
All interviews were transcribed and the interviewees were allowed to check these transcripts
for mistakes. We used triangulation by cross-checking simultaneously several data sources to
improve the reliability of the data.
The data set was first arranged in a manuscript, which covered IS innovation events, decision-
makers, locales, organizational structures, technological development, changes in business
units, etc. All identified events were arranged in a chronological order. We sent in May 1997 the
manuscript to company As senior vice president of IT. He had worked in this organization in
different roles but mainly in systems development tasks for the whole period. The manuscript
was corrected based on his requests and observations. After this, more data were gathered until
November 1997, and a second version of the manuscript was written in December 1997. This
manuscript was divided into two different parts. The first part covered the years 195490, and
the second part included the years 198497. This division was warranted by the fact that the
senior vice president of IT in company A had previously held important positions both in company
A (196384) and in company B (198490). This gave him an overall view of all developments
within and outside of company A. The second set included data from 1984 to 1997 and it was
sent to the CEO of company C. Company Cs CEO was considered to be qualified to review
the latter part, because he had held several senior positions in companies B and C during his
tenure in the companies. This division was made to retain confidentiality of some of the data.
Using the information retrieved from the manuscript, we arranged into a table each observed
incidence of an IS process innovation, its locale and the year when the adoption decision was
made.1 At the data categorization stage, the IS process innovation adoptions were further
divided into four generations, four innovation types and three locales. Moreover, adoption deci-
sions were divided into eight adoption decision-maker categories. When possible, we used the
chi-square test to recognize significant differences between adoption cohorts.
For the DOI-based analysis we analysed each adoption decision in terms of how it would
match with the identified list of DOI items. Thus our analysis, because of the data collection
approach, analysed only those events where an adoption decision had been made already. It

1
When there were several IS process innovations included in the same initiative, these were separated into separate inno-
vation events using the type as a distinguishing criterion.

2003 Blackwell Publishing Ltd, Information Systems Journal 13, 275297

 282

Table 2. Diffusion of Innovation (DOI) factors and items

DOI factor Attribute Definition Source

Innovation Relative advantage The degree to which an innovation is perceived better than the Kwon et al. (1987); Kozar (1989);
idea it supersedes Chin et al. (1995); Prescott & Conger (1995);
Premkumar & Potter (1995); Rogers (1995)
Ease of use The degree to which an innovation is perceived difficult to Chin et al. (1995); Rogers (1995)
understand and use
Compatibility The degree to which an innovation is perceived consistent Kwon et al. (1987); Kozar (1989); Chin et al.
with the existing values, past experiences and the needs of (1995); Prescott & Conger (1995);
potential adopters, similar to suitability Rogers (1995)
Visibility To what extend the innovation is visible to others Kozar (1989); Rogers (1995)
Trialability The degree to which an innovation may be experimented with Kozar (1989); Rogers (1995)
E Mustonen-Ollila & K Lyytinen

a limited basis
Price The cost of an innovation Premkumar & Potter (1995); Rogers (1995)
Problem solver The desirability of adopting an innovation depends on the Rogers (1995)
problem the innovation promises to solve for the adopter
Standard Manufactures and clients begin to use a standard forcing the Rogers (1995)
user to follow
Technological edge Superiority to other innovations Rogers (1995)
Task Commercial advantage The internal or external vendor sells an innovation in a form Rogers (1995)
of a useful product. Later on this product is commercialized
User need recognition An innovation must match the user needs in the task Rogers (1995)
User resistance When tasks become difficult, users resist change Kwon et al. (1987); Prescott & Conger (1995);
Rogers (1995)
Individual Own testing Innovation is tried as experimental basis Rogers (1995)
Personal contact People relying on experiences of their peers, similar to Rogers (1995)
network interpersonal networks
Own rules and control The innovator has to conduct experimentation with the new Rogers (1995)
of own work idea in order to assure itself that innovation is advantageous
Learning by doing Learning to evaluate the innovations on the basis of Rogers (1995)
experience

2003 Blackwell Publishing Ltd, Information Systems Journal 13, 275297

 DOI factor Attribute Definition Source

Environmental Cultural values Cultural beliefs concerning change Rogers (1995)

Technological The maturity of the technological infrastructure Rogers (1995)
infrastructure
Community norms The obedience to norms Rogers (1995)
Funding Available resources to invest (slack of resources) Kwon et al. (1987); Prescott & Conger (1995);
Rogers (1995)
Organizational Interpersonal networks Evaluations of innovations are exchanged between individuals March et al. (1958); Kwon et al. (1987);
Nilakanta & Scamell (1990); Prescott &
Conger (1995); Rogers (1995)
Peer networks Social relationships Rogers (1995)
Informal communication Exchange of information is informal and unplanned Nilakanta & Scamell (1990); Rogers (1995)
Technological experience Technological experience over a long time period Parnisto (1995); Rogers (1995)
Working teams Team members keeping primary control over their Rogers (1995)
management

2003 Blackwell Publishing Ltd, Information Systems Journal 13, 275297

Opinion leaders and An individual who influences clients innovation decisions Rogers (1995)
change agents
Interdependence from Each adopter increases the utility of the innovation for Rogers (1995)
others both future adopters
Adopter type The degree to which an individual is earlier in adopting Premkumar & Potter (1995); Rogers (1995);
new ideas than others Sauer & Lau (1997)
Management hierarchy An order is given to adopt an innovation Premkumar & Potter (1995); Rogers (1995)
Why organizations adopt IS process innovations
283
 284 E Mustonen-Ollila & K Lyytinen

Table 3. An analysis example of Diffusion of Innovation (DOI) item analysis for one process innovation adoption

DOI factors DOI item Evidence

Innovation Relative advantage Productivity was gained.

Ease of use COBOL Generator was easy and simple to use.
Compatibility The data searched from data dictionaries had to be read
with ASSEMBLER COBOL language. Integrated Database
Management System (IDMS)s Online Query Language did
not fit. COBOL Code Generator called Carelia was compatible.
Problem solver Automated tasks that took a lot of time and effort.
Individual Own trials Earlier used in massive IDMS information systems for controlling
the activities in Savonlinnas machine shop.
Environmental Technological infrastructure The tool was used in IDMS databases, which worked only in the
IBM environment.
Organizational Adopter roles Internal implementation of the tool (producer role), adopter role.
DOI, Diffusion of Innovation.

is also important to understand that we did not analyse responses to perceptual stimuli by
would-be decision-makers (like in survey methods using DOI) that were similar to all respon-
dents. Instead, we matched individual DOI items with actual decision-making behaviours (in
situ) by conducting content analysis of the adoption cases. Thus for each case we could iden-
tify a certain number of items that were likely to influence the decision at that point. Thus these
analyses of situations and associated documents acted as surrogates to actual decision-
making behaviours and intentions which are not any more accessible because of the nature of
our data set. For each item identified under a specific category, we also inferred that the spe-
cific factor had influenced the positive outcome of the adoption decision. Thus the external
validity of the data is higher than in many other DOI studies (because we analysed real deci-
sions), but our internal validity and reliability of the data and its analysis is lower (because we
do not have complete responses to all stimuli). Therefore some caution must be exercised
when interpreting the results.
An example of how one IS process innovation adoption was analysed with the DOI model is
presented in Table 3. The innovation adopted was COBOL Code Generator called Carelia. The
decision was made by the IS project group in 1983.

RESEARCH FINDINGS AND ANALYSIS

Adoption decisions explained by Diffusion of Innovation model

We show in Table 4 frequencies of the presence of DOI factor items over all adoption events
where each item had at least one observation. The data are organized into the four develop-
ment generations. By studying the innovation factors we can observe the following: (1) items
related to innovation factors are identified 200 times; (2) items related to task factors are iden-
tified 51 times; (3) items related to individual factors are identified 69 times; (4) items related

2003 Blackwell Publishing Ltd, Information Systems Journal 13, 275297

 Why organizations adopt IS process innovations 285

Table 4. Presence of adoption factors in four generations of information system process innovation

Generation 1 Generation 2 Generation 3 Generation 4

DOI factors (196066) (196782) (198390) (199197) Total

Innovation: sum total 5 92 38 65 200

Relative advantage 0 14 2 3 19
Compatibility 0 0 2 2 4
Ease of use 1 20 10 16 47
Visibility 0 7 2 4 13
Trialability 0 4 3 2 9
Price 1 4 1 4 10
Suitability 1 5 8 7 21
Problem solver 1 7 2 3 14
Standard 1 28 1 12 42
Technological edge 0 3 7 12 22

Task: sum total 0 20 13 18 51

Commercial advantage 0 0 1 2 3
User needs recognition 0 17 11 13 41
User resistance 0 3 1 3 7

Individual: sum total 1 34 15 19 69

Personal contact network 0 0 0 0 0
Own trials 0 13 10 7 30
Autonomous work 1 13 2 7 23
Learning by doing 0 8 3 5 16

Environmental: sum total 2 41 27 47 117

Cultural values 0 0 1 0 1
Technological infrastructure 2 29 13 41 85
Community norms 0 4 12 0 16
Funding 0 8 2 6 16

Organizational: sum total 4 35 29 17 85

Interpersonal networks and communication 0 0 2 0 2
channels
Near-peer networks 0 1 1 2 4
Informal communication 0 0 2 0 2
Past technological experience 2 12 13 11 38
Working teams 0 11 4 0 15
Opinion leaders, change agents 0 2 3 2 7
Interdependence from others 0 1 1 0 2
Adopter roles 1 4 3 2 9
Management and hierarchy 1 4 0 0 5

Total number of factors 12 222 122 166 522

to environmental factors are mentioned 117 times; and (5) items related to organizational fac-
tors are identified 85 times. The total number of items recognized in 208 adoptions was 522
items (in average 2.5 items per adoption decision). From the total number of items, each fac-
tors items relative frequencies were the following: innovation factor items 38.3%; task factor

2003 Blackwell Publishing Ltd, Information Systems Journal 13, 275297

 286 E Mustonen-Ollila & K Lyytinen

items 9.8%; individual factor items 13.2%; environmental factor items 22.4%; and organiza-
tional factor items 16.3%. Hence, overall the innovation and the task factor-related items played
the most significant role in adoptions.
We next investigated which items were the most important in affecting method adoptions.
This was carried out by analysing the frequencies of each item and identifying items in each
factor whose relative frequency was over 20%. The most common items over all IS process
innovation generations were for:

innovation factor: ease of use, standard;

task factor: user need recognition;
individual factor: own trials, autonomous work, ease of use, learning by doing;
environmental factor: technological infrastructure; and
organizational factor: past technological experience.
Thus, the most IS process innovation adoption decisions could be explained by the following
logic. They have a standardized nature that simplifies their understanding and adopters share
a perception that the innovation would be easy to use. The process innovation fits with the
identified need of some developers (this explains a large portion of internal innovation we
observed), and there is a possibility for developers to conduct prior trials. Possibility to auton-
omously decide on adoption influenced positively the propensity to adopt. Moreover, it is impor-
tant that the process innovation was smoothly integrated into the current work practice
(learning by doing). Finally, because of the large number of technology- and tool-related pro-
cess innovations, the importance of past technological experience and learning and compat-
ibility with the existing technological infrastructure played an important role (path dependency).
It is also interesting to analyse which items did not influence adoption decision in each factor.
In the innovation factor the compatibility and trialability factors had low scores consistently. This
is somewhat surprising but can be explained by the fact that a majority of the innovations
resulted from internal learning and experimentation. At the time of decision such features
therefore did not matter very much. Regarding the task factor it appears that decision-makers
paid little attention to user resistance or the commercial value of the innovation (no cost/benefit
calculations). In the individual factor the impact of personal contact networks was insignificant.
This finding surprised us. This can be partly explained by the fact that these companies held
normally semiformal relationships and networks to a number of consulting houses which was
one of the main items affecting of IS process innovations in the organizational factor.
In the environment factor because of the type of innovations there was very little clash
with the cultural values. Regarding the organizational factor it looks that the decisions were not
much influenced by informal networks and impact of opinion leaders. This does not necessarily
mean that such issues cannot play an important role in the earlier steps of the innovation
process.
We analysed also the frequencies of each item for each factor. These analyses are sum-
marized in Figures 26. In each figure each item is represented as a percentage of the total
number observations for that factor. Our analysis shows that in all other factors than the inno-
vation factor the distribution of frequencies over items was skewed and normally only one item

2003 Blackwell Publishing Ltd, Information Systems Journal 13, 275297

 Why organizations adopt IS process innovations 287

25%
Relative advantage
20% Compatibility
Ease of use
15% Visibility

10% Trialability
Price
5% Suitability
Problem solver
0%
Standard
Figure 2. Innovation factor items. Technological edge

90%
80%
70%
60%
Commerce advantage
50%
User need recognition
40%
User resistance
30%
20%
10%
0%

Figure 3. Task factor items.

50%
45%
40% Own trials
35%
30% Autonomous work
25%
20% Learning by doing
15%
10%
5%
0%

Figure 4. Individual factor items.

2003 Blackwell Publishing Ltd, Information Systems Journal 13, 275297

 288 E Mustonen-Ollila & K Lyytinen

80%
70%
Cultural values
60%
50% Technological
infrastructure
40%
Community norms
30%
20% Funding
10%
0%

Figure 5. Environmental factor items.

50% Interpersonal networks

45% Peer networks
40% Informal communication
35% Technological experience
30% Working teams
25% Opinion leaders
20% Interdependence
15% Adopter roles
10% Management
5%
0%

Figure 6. Organizational factor items.

was established as strong predictor for an adoption decision. This suggests that more specific
instruments to measure IS process innovation adoption may be needed.

Factors affecting information system process innovation adoption

Using data in Table 4 we investigated in more detail the influence of DOI factors on IS process
innovation adoptions overtime. We used the standardized chi-square test (Vasama & Vartia,
1979) to detect if there were significant differences between how different factors affected
adoption decisions across different generations.
Unfortunately, general conditions to use the chi-square test were not fulfilled (there were
some entries with zero observations) for the whole data set. We had to drop consequently the
first generation data from our analysis, which covered about 2% of the observations. The anal-
ysis results of the remaining generations (c2 = 13.857, a = 0.05) showed that there were no sig-
nificant differences and the DOI factors affecting adoption decisions have thus remained the
same over different generations.
We analysed also the percentage distribution of the presence of factors across the four gen-
erations (Figure 7).

2003 Blackwell Publishing Ltd, Information Systems Journal 13, 275297

 Why organizations adopt IS process innovations 289

45%
40%
35% Innovation
30%
Task
25%
Individual
20%
15% Environmental
10% Organizational
5%
0%
Gen 1 Gen 2 Gen 3 Gen 4

Figure 7. Frequency distribution of the Diffusion of Innovation factors across the four generations.Gen, Generation.

Our analysis shows that the relative frequency of the innovation factor remained high
throughout all four generations. It was the most prominent during the second generation. Also
task factor remained relatively stable across three later generations. The only significant vari-
ation occurred in the organizational factor: it declined in importance overtime especially during
the third and the fourth generation adoptions. Based on our factor data alone we cannot pro-
vide any reasonable explanation for this. Rather this finding can be explained by some specific
changes in the environment of software development during that period. In the late 80s the
companies faced a decline in the number of projects on one hand, and outsourced develop-
ment activities on the other hand, which decreased the importance of organizational environ-
ments. Not surprisingly, environmental factors peaked during the fourth generation. This
reflects the growing infrastructural dependency of the IS development.
We can interpret the above findings as follows. When computing was in its infancy during the
first generation, the innovation and organizational factors (past technological experience) were
prominent. After that the rapid increase in the number of projects and in-house solutions during
the second generation resulted in a growing importance of recognizing better user needs for
the innovation (task), and, obviously, making them also environmentally more fit (infrastruc-
ture). Also new features of innovation like the ease of use and its standardized features
became more important. This resulted in a more encompassing and rounded approach to IS
process innovation. As one interviewee pointed this out:

In generation two the purpose of the IS department was to develop its own working methods
and methodologies; to improve the control and management of the IS projects; improve the
IS documentation; reduce the system complexity; improve the turnaround time; improve the
efficiency of development work; and to make procedures coherent and standard based.

The goals driving adoptions decisions also included a need to reach a mutual agreement
between the IS project groups and the IS department of the importance of the innovation. In
company A the clientvendor relationship between the IS department and customer depart-

2003 Blackwell Publishing Ltd, Information Systems Journal 13, 275297

 290 E Mustonen-Ollila & K Lyytinen

ments began in the 1970s. After the 1984 outsourcing, the business strategy of the client
departments became critical to the IS vendor. Several projects lacked appropriate tools
because of the growing diversity of technological infrastructures. Therefore company B and
later on company C were obliged to innovate internally especially tools and some baseline
technologies.
In the fourth generation the innovation factor again assumed more prominence as the num-
ber of projects increased. This made it imperative to seek and apply new IS process innova-
tions. It is notable that task and individual factor also remained high in importance reflecting
fundamental changes in the clientvendor relationships. Outsourcing, during the mid-80s,
turned the internal IS department in company A into a separate profit centre (independent
company) which necessitated that greater emphasis had to be placed on the clients com-
pany As needs and infrastructure.

Variation in method categories affecting diffusion factors over time

Our second research question was whether the significance of factors changed across differ-
ent types of IS process innovation types.
We therefore classified the data in terms of the four innovation types as shown in Table 5.
At the same time we normalized the data by dividing the total number of factors with total num-
ber of adoption decisions in each of the four innovation categories.
Table 5 was analysed with the chi-square test to detect differences between how factors
influenced adoptions in four IS process innovation types. The test (c2 = 21.026, a = 0.05)
showed statistically significant differences. This suggests that IS process innovation types are
different, and consequently they are influenced by different DOI factors.
A single factor affected adoptions in project management and control procedures category
(M). Moreover, these factors varied from one case to another. In technology (T) and description
techniques category (D), the adoption decision was influenced normally by two factors. In the
tools (TO) category the adoptions were influenced by three factors (see Table 5). This shows
that in different innovation categories the adoption decision may depend on or be influenced
by a varying set of factors. This variance may also reflect the relative complexity of associated

Table 5. The frequency of factors affecting information system process innovation adoption in four innovation types

Factor M T TO D Total number

Innovation 25 84 65 26 200
Task 6 16 19 10 51
Individual 20 18 22 9 69
Environmental 15 41 55 6 117
Organizational 11 24 30 20 85
Total number of observations 77 183 191 71 522
Total number of adoption decisions/factor 72 97 59 35 263
Total number of factors affecting an adoption decision 1 2 3 2 2 (in average)
M, management innovation; T, technology innovation; TO, tool innovation; D description innovation.

2003 Blackwell Publishing Ltd, Information Systems Journal 13, 275297

 Why organizations adopt IS process innovations 291

innovations. The types of factors influencing the adoptions also varied. In all adoptions the
innovation factor played the most important role. In management innovations (M) the second
important factor was individual factor, while in technology (T) and tool (TO) innovations it was
the environmental factor (infrastructure). In the description techniques innovation (D) the sec-
ond factor in significance was the organizational factor.
We sorted the adoption data into the four innovation types and into three locales to inves-
tigate possible variation of the factor influence over locales. Adoptions were also sorted
according to the year of adoption decision to map adoption decisions into different generations.
Likewise the frequencies of items were tallied. The sum totals were counted for the items of
each of the five factors. The results of this analysis are presented in Figures 811. They depict
sum totals of items for each factor, for each generation and for each innovation type.
If we examine Figures 811 we can discover that in the first generation technology (T), tools
(TO) and description techniques category (D) adoptions were not influenced by any specific
factor because of a low number of observations. During the second generation the project
management (M) category was influenced by all DOI factors. Overall, the second largest num-
ber of adoptions were in M category innovations. Yet, nearly 90% of all the DOI items for this
category were found during the first and second generations.
After the second generation the DOI factors decreased in importance for M adoptions as the
whole area decreased in popularity. For example, task and organizational factors did not have
any significant influence. In fact, we observed during our analysis that other factors not rec-
ognized in the DOI theory like the time factor became critical. After second generation the
influence of the clientvendor relationships became more pronounced. Shorter projects
became critical for both parties for economic reasons.

50
45
40 Innovation
factors
35
Task factors
30
Individual factors
25
20 Environmental
factors
15
Organizational
10 factors
5
0
Gen 1 Gen 2 Gen 3 Gen 4

Figure 8. Diffusion of Innovation factor influences on the management (M) type innovation. The locales over generations
1, 2, 3 are different as locales B and C existed only during generations 3 and 4.Gen, Generation.

2003 Blackwell Publishing Ltd, Information Systems Journal 13, 275297

 292 E Mustonen-Ollila & K Lyytinen

50
45
40
Innovation factors
35
30 Task factors

25 Individual Factors
20 Environmental
15 factors

10 Organizational
factors
5
0
Gen 1 Gen 2 Gen 3 Gen 4

Figure 9. Diffusion of Innovation factor influences on technology (T) innovations.Gen, Generation.

50
45
Innovation factors
40

35 Task factors
30
Individual factors
25
20 Environmental
15 factors

10 Organisational
factors
5
0
Gen 1 Gen 2 Gen 3 Gen 4

Figure 10. Diffusion of Innovation factor influences on tools (TO) adoptions.Gen, Generation.

Description techniques (D) were used during project specifications with the client and as an
education tool for clients. This innovation became increasingly important when these problems
surfaced but its importance declined after these issues had been solved. Maybe because of
this, the smallest number of adoptions fell into category D. This is somewhat ironic in the sense
that a largest body of IS process innovation research reviewed above has analysed method-
based process innovations. DOI factors did not significantly affecting type D innovation
adoptions in the third and fourth generations. One reason for this is that the infused process

2003 Blackwell Publishing Ltd, Information Systems Journal 13, 275297

 Why organizations adopt IS process innovations 293

50
45
Innovation factors
40
Task factors
35
30 Individual factors
25
Environmental
20 factors
15
Organizational
10 factors
5
0
Gen 1 Gen 2 Gen 3 Gen 4

Figure 11. Diffusion of Innovation factor influences on description techniques (D) adoptions. Gen, Generation.

technologies were found adequate, and available methods were chosen when such needs
arose.
In tools category (TO) the importance of all innovation factors increased steadily. To our sur-
prise, through all four generations innovations in this category including application generators,
network management tools and the like were outcomes of in-house engineering. Often there
were no suitable tools in the market place because of diverse computer platforms used in the
organization, which rendered the technological infrastructure difficult to manage.
In the technology category (T), innovation factor became the most important factor affecting
adoptions in the last three generations. For example, the database management and program-
ming tools belonged to this category. Normally they required no modifications and were taken
in use directly.

DISCUSSION AND CONCLUSIONS

This study examined a DOI-based model of IS process innovation adoption using a longitudinal
data set of IS process innovation adoptions. The study shows that several factors recognized
in DOI theory affect IS process innovation adoptions. The most important items in these factors
observed to influence adoptions were: user need recognition, technological infrastructure, past
technological experience, own trials, autonomous work, ease of use, learning by doing and
standards.
Adoption decisions within innovation categories were influenced by different sets of factors.
Moreover, the significance of a given set of factors varied by IS process innovation generations.

2003 Blackwell Publishing Ltd, Information Systems Journal 13, 275297

 294 E Mustonen-Ollila & K Lyytinen

An interesting finding is that over all IS process innovation adoptions the influence of DOI fac-
tors between 1967 and 1997 have remained largely the same.
In the study we confirmed the validity of DOI theory in explaining IS process innovation adop-
tions. Despite this finding, a large number of DOI factors followed no discernible pattern across
adoption events. Business pressures demanded that new IS had to be developed, and con-
sequently new IS process innovations had to be adopted. This is in line with Sauer & Laus
(1997) findings. Kozars (1989) personal factors were not found to be important. No evidence
for his age importance factor was found either. Instead, our findings confirm the importance
of innovation and environmental factors in IS process innovation. In Huff & Munros (1985) nor-
mative model, the most important thing was the ability to identify the appropriate technology to
support each of the application areas. This finding is similar to our observation of the impor-
tance of user need recognition. Little evidence was found to support Premkumar & Potters
(1995) findings of the importance of a product champion, strong top management support,
lower IS expertise, relative advantage and cost effectiveness. Nilakanta & Scamells (1990)
findings about the importance of communication channels and sources, and technical support
were confirmed.
An interesting discovery of IS process innovation adoptions was that during periods of busi-
ness prosperity organizations started to concentrate on methods. This normally happened
when a new technology generation came to the market and the platform systems had to be
renewed. Therefore organizations invested in IS process innovations in cycles which is not
accounted for by the DOI theory.
A major difficulty we encountered during the study concerned the DOI model. It suggests a
large of number of items for each factor. However, many of them were not easily observable
with a recall method we followed. Another limitation concerns the limited number of organiza-
tions studied. The third limitation concerns our classification into innovation types. Sometimes
the classification had to be made roughly, as there were some complex adoption events to be
classified. The fourth limitation concerns analysis of cases where several factors had a bearing
on a decision. Weighting the relative importance of each factor against the rest of the factors
proved to be an impossible task. The fifth limitation concerns the comprehensiveness and thus
reliability of the data: despite our efforts to obtain, through in-depth interviews and extensive
use of archival material and all relevant facts that affected adoptions, we have to accept the
limitation of a historical method. Due to the data set and the poor measurement accuracy, sta-
tistical generality is difficult to gain because of weak statistical tests. Finally, the obtained
results may not be applicable to other organizations because the phenomena studied in this
study can be atypical.
The study has several theoretical implications. First, the DOI model should be extended to
incorporate resource restrictions and time as important factors. Second, we observed external
adoption mechanisms, which are not accounted for by the DOI theory.
We have here only dealt with IS process innovation adoptions. Whether these innovations
were taken into use and how the knowledge of these innovations was transferred to local prac-
tice will qualify as a further subject of study. According to our preliminary data, approximately
90% of the adopted innovations were also used which is surprisingly high figure.

2003 Blackwell Publishing Ltd, Information Systems Journal 13, 275297

 Why organizations adopt IS process innovations
ACKNOWLEDGEMENTS
The authors wish to express their gratitude to Mr Mauri Mustonen, Mr Tapani Ollila, Dr Philip
Powell, Dr Nancy L. Russo, Dr Jukka Heikkonen and three anonymous reviewers. We also
thank interviewees for their time and support.
REFERENCES
Barley, S.R. (1990) Images of imaging: notes on doing lon-
gitudinal field work. Organization Science, 1 (3), 220
247.
Chin, W.W. & Gopal, A. (1995) Adoption intention in
GSS: relative importance of beliefs. The Data Base
for Advances in Information Systems, 26 (2&3), 42
64.
Copeland, D.G. & McKenney, J.L. (1988) Airline reserva-
tions systems: lessons from history. MIS Quarterly,
September Issue, 353370.
Curtis, B., Krasner, H. & Iscoe, N. (1988) A field study of
the software design process for large systems. Commu-
nications of the ACM, 31 (11), 12681287.
Fitzgerald, B. (1997) The use of systems development
methodologies in practice: a field study. Information Sys-
tems Journal, 7, 201212.
Friedman, A. & Cornford, D. (1989) Computer Systems
Development: History, Organization and Implementa-
tion. John Wiley & Sons, Information Systems Series,
New York.
Giddens, A. (1984) The Constitution of Society. Polity
Press, East Sussex, UK.
Heiskanen, A. (1994) Issues and factors affecting the suc-
cess and failure of a student record system development
process, a longitudinal investigation based on reflection-
in-action, PhD Dissertation. University of Helsinki,
Yliopistopaino.
Huff, S. & Munro, M.C. (1985) Information technology
assessment and adoption: a field study. MIS Quarterly,
December Issue, 327339.
Jrvenp, S. (1991) Panning for Gold in Information
Systems Research: Second-hand data. Proceedings
of the IFIP TC/WG 8.2, Working Conference on the
Information Systems Research Arena of the 90s
Challenges, Perceptions and Alternative Appro-
aches, pp. 6380. Copenhagen, Denmark, 1416
December.
Johnson, J.M. (1975) Doing Field Research. The Free
Press, New York.
2003 Blackwell Publishing Ltd, Information Systems Journal 13, 275297
295
Kozar, K.A. (1989) Adopting systems development meth-
ods: an exploratory study. Journal of Management Infor-
mation Systems, 5 (4), 7386.
Kwon, T.H. & Zmud, R.W. (1987) Unifying the fragmented
models information systems implementation. In: Critical
Issues in Information Systems Research, Boland, R.J. &
Hirchheim, R.J. (eds), pp. 227251. John Wiley & Sons,
New York.
Laudon, K.C. (1989) Design guidelines for choices involv-
ing time in qualitative research. The Information Sys-
tems Research Challenger: Qualitative Research
Methods, Vol. 1, pp. 112. Harvard Business School
Research Colloquium, Boston, MA, USA.
March, J.G. & Simon, H. (1958) Organizations. John Wiley
& Sons, Inc., New York.
Mason, R.O., McKenney, J.L. & Copeland, D.G. (1997b)
An Historical Method for MIS Research: Steps and
Assumptions, MIS Quarterly, 21 (3), 307320.
Mason, R.O., McKinney, J.L. & Copeland, D.C. (1997a)
Developing an historical tradition in MIS research. MIS
Quarterly, 21 (3), 257278.
McKenney, J.L., Mason, R.O. & Copeland, D. (1997) Bank
of America: the crest and trough of technological lead-
ership. MIS Quarterly, 21 (3), 321353.
Nilakanta, S. & Scamell, R.W. (1990) The effect of infor-
mation sources and communication channels on the
diffusion of innovation in a data base development envi-
ronment. Management Science, 36 (1), 2440.
Parnisto, J. (1995) Related absorptive capacity: a perspec-
tive on groupware adoption, managing information &
communications in a changing global Environment. Pro-
ceedings of the 1995 Information Resources Manage-
ment Association International Conference, 2124 May,
Atlanta, Georgia, USA, pp. 154161.
Pettigrew, A. (1985) The Wakening Giant, Continuity and
Change in ICI. Blackwell, Oxford and New York.
Pettigrew, A. (1989) Issues of Time and Site Selection in
Longitudinal Research on Change. The Information Sys-
tems Research Challenger: Qualitative Research Meth-
 296 E Mustonen-Ollila & K Lyytinen
ods, Vol. 1, pp. 1319. Harvard Business School
Research Colloquium, Boston, MA, USA.
Pettigrew, A.M. (1990) Longitudinal field research on
change: theory and practice Organization Science, 1 (3),
267292.
Premkumar, G. & Potter, M. (1995) Adoption of Computer
Aided Software Engineering (CASE) Technology: an
innovation adoption perspective. The Data Base for
Advances in Information Systems, 26 (2&3), 105124.
Prescott, M.B. & Conger, S.A. (1995) Information technol-
ogy innovations: a classification by IT locus of impact
and research approach. The Data Base for Advances in
Information Systems, 26 (2&3), 2041.
Rogers, E.M. (1995) Diffusion of Innovations, 4th edn. The
Free Press, New York.
Sauer, C. & Lau, C. (1997) Trying to adapt systems devel-
opment methodologies a case-based exploration of
business users interests. Information Systems Journal,
7, 255275.
Swanson, E.B. (1994) Information systems innovation
among organizations. Management Science, 40 (9),
10691088.
Tolvanen, J.-P. (1998) Incremental Method Engineering
with Modeling Tools: Theoretical Principles and Empiri-
cal Evidence, PhD Dissertation Thesis. Jyvskyl stud-
ies in Computer Science, Economics and Statistics 47,
University of Jyvskyl, Jyvskyl, Finland. Engineering Journal, and Information Systems Journal. He
Vasama, P.-M. & Vartia, Y. (1979) Johdatus tilastotietee-
seen, osa II, Oy Gaudeamus Ab. In: Satakunnan Kirja-
teo Ilisuus, Pori 1980, Finland.
Wynekoop, J.L. & Russo, N.L. (1997) Studying system
development methodologies: an examination of
research methods. Information Systems Journal, 7, 47
65.
Yin, R.Y. (1993) Applications of Case Study Research,
Applied Social Research Methods Series, Vol. 34. SAGE
Publications, International Educational and Professional
Publisher, Newbury Park, London, UK.
Biographies
Erja Mustonen-Ollila (MSc, Eng.) is working currently as
an information systems lecturer in the Lappeenranta
University of Technology (LUT, Finland), within the depart-
APPENDIX 1: SHORT DESCRIPTION OF CASE ORGANIZATIONS
Three Finnish organizations (units), here called companies A, B and C, respectively, were the data collection site. Com-
pany A is a big paper-producer, whereas company Bs business was to design, implement and maintain information sys-
2003 Blackwell Publishing Ltd, Information Systems Journal 13, 275297
ment of Information Technology. She graduated in 1984
from the LUT from the department of mechanical engineer-
ing and worked in several IS positions in the industry from
19841991, such as in analysing, designing, and imple-
menting information systems, and managing and control-
ling the IS projects. From 1992 she began to work in LUT
in the IT department, first as an assistant, then as a
research lecturer.
Her main research interests are information systems, IS
methodologies, project controlling and planning activities,
Diffusion of Innovation theory, organizational learning, and
IT outsourcing issues.
She is doing her research and her PhD project on Infor-
mation System Process Innovation Adoption, Adaptation,
Learning, and Evolution: A Longitudinal Case Study. The
goal of the project is to examine system development
IS process innovation (ISPI) choice, adoption, learning,
unlearning, and adaptation factors in three Finnish organi-
zations over a 43 period. The study is carried out as a lon-
gitudinal case study.
Kalle Lyytinen is a Professor in Information Systems at
Case Western Reserve University, and an adjunct profes-
sor at the University of Jyvskyl, Finland. He serves cur-
rently on the editorial boards of several leading IS journals
including, AIS journal, Information Systems Research,
EJIS, JSIS, Information & Organization, Requirements
has published over 150 scientific articles and conference
papers and edited or written eight books on topics related
to system design, method engineering, implementation,
software risk assessment, computer supported coopera-
tive work, standardization, and ubiquitous computing. He is
currently involved in research projects that look at the IT-
induced innovation in software development, architecture
and construction industry, and is developing a high-level
requirements model for large-scale systems. He is also
engaged in a project supported by NSF that focuses on the
institutional forces involved in the development of global
electronic commerce. His research interests include infor-
mation system theories, computer-aided system design
and method engineering, system failures and risk assess-
ment, computer-supported cooperative work, nomadic
computing, and the innovation and diffusion of complex
technologies and the role of institutions in such processes.
 Why organizations adopt IS process innovations 297

tems (IS). In 1984 company A transferred (Mustonen-Ollila, 2001) its IS function into a newly formed company and
company B was set up by the former employees of company A. In 1995 company B was divided into five separate com-
panies, one of them being company C. Until the end of 1997, company C formed a division within company B. Ever since
their start-up these companies have been in close co-operation with company A and have had intense customer
relationship.
The companies locales are situated in three Finnish cities. Company A located in Imatra and housed several IS func-
tions during the years 195469 in its different departments (accounting, etc.). In 1969 a separate formal IS department
was established and it was continued until 1984 when the department was made a separate profit centre and a separate
company. Then a company B was set by its former employees.
Company A had also in-house IS functions in Helsinki between 1961 and 1969. During 196984 these belonged to the
IS department of company A. Despite having separate locations, we chose to treat both Helsinki and Imatra as a single
locale, named as locale 1, because of the fact that the two were working intimately together and belonged to the same
IS department and followed the same methodology. After the 1984 outsourcing decision until 1995, Imatra was treated as
a separate locale. Between 1995 and 1997, Imatra site continued to be treated as locale 2. The Lappeenranta locale, the
third research subject, was established in 1989 within company B and it continued until 1995 under the authority of com-
pany B. We treat it as a separate locale because it had separate and independent IS development functions and different
technological platform.

2003 Blackwell Publishing Ltd, Information Systems Journal 13, 275297