Download as pdf or txt
Download as pdf or txt
You are on page 1of 20

A qualitative study of inter-

organizational knowledge
management in complex products
and systems development
Eric W.T. Ngai1, Chen Jin2 and Tong Liang2
1
Department of Management, The Hong Kong Polytechnic University, Hung Hom, Kowloon,
Hong Kong, PR China. mswtngai@inet.polyu.edu.hk
2
School of Management, Zhejiang University, Hangzhou, Zhejiang 310027, PR China.
cjhd@zju.edu.cn; cqtongliang@163.com

Much research has recognized that cross-disciplinary knowledge inputs are necessary to
successful product innovation, particularly in complex products and systems (CoPS) develop-
ment. This paper focuses on developing an exploratory framework for studying how an inter-
organizational knowledge management (KM) approach affects the development of CoPS, and
exploring how inter-organizational collaboration affects KM within a CoPS context. Based on
a review of the literature and in-depth case studies, this paper puts forward some propositions
and develops an integrated conceptual model showing the link between the dimensions of inter-
organizational KM and the performance of CoPS development. Finally, the implications of the
proposed model for inter-organizational KM in a CoPS environment are discussed.

1. Introduction involving a lot of integration of technologies and


knowledge, so most CoPS projects are developed

D uring recent years, the management litera-


ture has increasingly touched on studies of
different issues of inter-organizational collabora-
under an innovation network which involves the
integrators, contractors, buyers, suppliers, and in
most cases, the government (Hobday and Rush,
tion, knowledge management (KM) and complex 1999). In this kind of innovation network, lin-
products and systems (CoPS). The resource-based kages between firms and other institutions differ
view claims that under the emerging circumstances from those found in the traditional manufactur-
of a knowledge-based economy and globalization, ing approach, which focuses on individual firms
an organization’s networks of external relationships with clear boundaries, and on the transactions
have become a fundamental source of successful between firms and their operations. The literature
technological innovation and a major factor in a on KM and inter-organizational cooperation,
firm’s competitive advantage (Gulati et al., 2000). which is heavily based on evidence from a mass-
CoPS are defined as high cost technology and production context, has failed to help CoPS
software intensive products, systems and capital manufacturers cope with new challenges such as
goods, which are manufactured in small batches managing knowledge across project families or
or one-off projects (Hobday et al., 2000). CoPS organizational boundaries (Chen and Tong, 2004;
consist of various components and subassemblies, Oshri and Newell, 2005).

R&D Management 38, 4, 2008. r 2008 The Authors. Journal compilation r 2008 Blackwell Publishing Ltd, 421
9600 Garsington Road, Oxford, OX4 2DQ, UK and 350 Main St, Malden, MA, 02148, USA
Eric W.T. Ngai, Chen Jin and Tong Liang

Although the individual importance of inter- edge can be understood and applied in specific
organizational collaboration, KM and CoPS has contexts (Nonaka, 1994). Organizations should
been recognized, their synergic effects have thus try to positively devote themselves to inter-orga-
far been neglected and unexplored. Hence, nizational linkages, building up the mechanisms
this paper discusses different arguments and of long-term commitment and cooperation with
perspectives of the grounded theories of KM, each other to create more competitive advantages
inter-organizational collaboration and CoPS in- (Liu and Shou, 2004).
novation, and explores whether CoPS manufac- Knowledge work is fraught with tensions work-
turers can facilitate project success and attain ing together in a loosely coupled fashion (Dough-
technical and management knowledge via inter- erty et al., 2000): connecting market and
organizational collaboration within the CoPS technology knowledge into new product ideas
development network. This paper aims to address (product level), juxtaposing multiple ideas ac-
two key questions: cording to strategic standards (capability level),
and evolving from strategic settings to strategic
(1) How does inter-organizational collaboration
paths (strategic level), etc. The purpose of knowl-
impact on KM within a CoPS innovation
edge-related activities in a firm is to organize the
network? and
whole KM process to deliver the right informa-
(2) How is CoPS innovation performance related
tion to the right person at the right time, thus
to inter-organizational KM?
facilitating the right decision-making (Barthes
This paper begins with an overview of inter- and Tacla, 2002), to balance the tensions. As the
organizational collaboration, KM and CoPS in- technology and activities needed in innovation
novation. Next, follows a discussion about the become more complicated, firms are driven to
propositions of the inter-relationships among seek inter-organizational collaboration and mo-
inter-organizational collaboration, KM, and bilize knowledge resources. Furthermore, how to
CoPS innovation. In addition, experiences from assimilate and integrate knowledge from sources
two case studies will be delineated. Then, accord- across borders has become an emergent issue in
ing to the propositions, an integrated conceptual KM (Lynn et al., 2000).
model grounded in knowledge-based theory and Based on the research results from the CoPS
network theory is developed for inter-organiza- studies (e.g. Hobday and Rush, 1999), a CoPS
tional KM in a CoPS context. Finally, the im- innovation network is a closed network to which
plications and some suggestions for future entry is restricted by the core players, i.e. the system
research are discussed. integrators, core sub-contractors and users. Vaa-
land and Hakansson (2003) address the extreme
importance of inter-organizational collaboration
2. Literature review linkages between buying and selling parties of
CoPS projects, both at technology level and man-
The notion of inter-organizational networks and agement level. It is believed that the collaborative
collaboration are not new. Oliver (1990, p. 241) relationships between subcontractors, integrators,
defines inter-organizational collaboration as and buyers of CoPS projects affect the knowledge
‘enduring transactions, flows, and linkages that flow (Wuyts et al., 2004). Although a number of
occur among or between an organization and one studies have been conducted about KM, inter-
or more organizations in its environment’. Powell organizational collaboration and CoPS, there is a
et al. (1996) point out that innovations in the lack of studies on the relationships among them.
biotechnology industry are often generated within This paper builds on the issue, identifying the
a collaboration network, not by an individual linkages, based on a literature review and a case
firm. However, inter-organizational networks study of two typical CoPS projects.
and collaborations incur costs and risks to their
constituent organizations (Williams, 2005).
Galunic and Rodan (1998) contend that the
exchange of knowledge through inter-organiza-
3. Development of propositions and
conceptual framework
tional collaboration is problematic and costly,
especially for tacit know-how, one explanation
3.1. Research approach
of which is that the interchange of knowledge
requires some shared system of meaning among A case study is particularly suitable for problems
innovative participants so that transferred knowl- where research and theory are at early stages, as it

422 R&D Management 38, 4, 2008 r 2008 The Authors


Journal compilation r 2008 Blackwell Publishing Ltd
Qualitative study of inter-organizational knowledge management

is helpful in dealing with unstructured informa- A’s competitive strength lies in the design, man-
tion such as documents and interviews, and ufacturing and installation of DCS. In 2000, when
expanding the explanatory and predictive power Integrator A got the contract for SysProj A, the
of the studied theory (Benbasat et al., 1987). This number of its employees had reached 500, and the
study is exploratory in nature, because past re- registered capital was over 100 million RMB.
search has not addressed the issues of applying Similarly, Integrator B is also a specialized system
inter-organizational KM in the development of developer, providing R&D, manufacturing, and
CoPS projects. We believe that through in-depth engineering services for process automation con-
and systematic interviews, the case study metho- trol solutions in many sectors such as the chemi-
dology is the most valuable for exploring and cal, oil refining, electric power, construction and
understanding the impact of inter-organizational machinery industries. Integrator B employs
KM in the development of CoPS projects. nearly 200 R&D engineers, and invests a consid-
erable portion of annual revenue in R&D. As a
consequence, Integrator B possesses about 50
patents and had got CMM certification by 2002.
3.2. Background of case study Integrator B has mastered several core technolo-
In this section, a brief background of the case gies and developed a series of advanced, practical,
companies and CoPS projects are provided. Two reliable and economical systems platforms, devel-
cases were carefully selected from the system oping several types of advanced and mature
automation sector. The two projects were ‘Sy- products and completing over 200 projects, most
sProj A’ and ‘SysProj B’, developed by ‘Integrator of which have been highly tailored to client
A’ and ‘Integrator B’ respectively. SysProj A is a specifications.
control system developed for a Mass Transit In both system manufacturers, interviews were
Railway (MTR) urban transportation system in conducted with managers and employees at dif-
a city located in North China. SysProj B is also a ferent levels: senior project managers and com-
computerized control system installed in a power pany executives; supervisors and practitioners
plant in South China. Both projects have been involved in project and functional activities
active for more than 3 years. This kind of dis- (Table 1). Interviewees were selected based upon
tributed control system (DCS) is a typical exam- their participation in each project. The data we
ple of CoPS: high investment, capital goods, long tried to gather included project backgrounds,
development cycle, embedded complex hardware project management, KM practices, and relation-
and software components, and a high degree of ships with partners in innovation networks, etc.
customization. Moreover, for different interviewees, we prepared
Found in 1993, Integrator A is a High-Tech different topics relevant to the project manage-
enterprise specializing in the research & develop- ment and KM practices, which are summarized in
ment, manufacturing, marketing and engineering Table 1.
services of industrial automation products. In the The multi-level interviews helped us gain a
past l0 years, Integrator A has experienced a better understanding of the synthesis of project-
dramatic growth both in sales and profits. The level processes and firm-level processes through-
products and systems developed by Integrator A out the project duration, and allowed us to
have a good reputation in the industry. Integrator identify specific KM practices. The information

Table 1. List of interviewee roles and interviewed topics.


Position Discussed contents Interviewee number

SysProj A SysProj B

Senior managers Project and company backgrounds, attitudes to 1 2


of the company inter-organizational collaboration and KM
Project managers Project management practices, issues and approaches of 1 1
inter-organizational relationships and interaction
Sales engineers Interactions with the users throughout the project duration 2 1
R&D engineers Some detailed practices of collaboration and KM 3 3
Staff form Technical and resource support for interaction with partners and 1 1
supportive users
departments

r 2008 The Authors R&D Management 38, 4, 2008 423


Journal compilation r 2008 Blackwell Publishing Ltd
Eric W.T. Ngai, Chen Jin and Tong Liang

about these case projects was supplemented with structural facilitator for inter-organizational
data from secondary sources such as the websites collaboration. Therefore, following prior re-
and internal publications of the integrators. search, we define network embeddability as the
tightness of linkages among actors within a CoPS
innovation network, both in the resource dimen-
sion and the relation dimension. According to
3.3. Case analysis Sporleder and Moss (2002), network embeddabil-
3.3.1. How does network embeddability impact on ity has potential explanatory power for the devel-
inter-organizational KM activities? opment and dynamics of knowledge-related
Following Nonaka (1994), and Nonaka and collaboration.
Takeuchi (1995), much KM literature has tried Based on these arguments, the following pro-
to find methods to capture, transfer and apply position is posited:
knowledge to create value (e.g. Davenport and
Proposition 1(P1): Network embeddability has
Prusak, 1998). It has been argued that organiza-
a significant influence on inter-organizational
tional learning occurs in four KM activities:
KM
knowledge acquisition, knowledge distribution,
knowledge interpretation and knowledge retrieval
(Huber, 1991). In CoPS development, the foun-
dational knowledge typically has to come to- 3.3.1.1. The interrelationship between network
gether in novel assemblages in order to solve embeddability and the degree of knowledge acqui-
specific problems (Marshall and Brady, 2001). sition. As partners with dense linkages are em-
That is functional capability in systems integra- bedded in the network, they tend to solve
tion: an application of system engineering is problems through negotiation, tolerate unfavor-
required to integrate the multi-disciplinary able actions on the part of each other as short-
knowledge inputs (Davies and Brady, 2000). term turbulence, and consider cooperation as a
Marshall and Brady (2001) also suggest that the long-term commitment building process (Lui
intention of KM in CoPS innovation is to cap- et al., 2005). Relational linkages such as ‘commit-
ture, depersonalize and disseminate expert knowl- ment’ and ‘trust’ are considered as aspects of
edge bases. Therefore, in this paper we consider embeddability (Sporleder and Moss, 2002).
inter-organizational KM as consisting of three Specifically, Hansen (2002) notes that established
processes: knowledge acquisition, knowledge im- dense ties between a focal team and other business
plementation, and knowledge transfer. Knowledge units are helpful when the team identifies knowl-
acquisition is concerned with the access and ab- edge that requires effort to be moved from
sorption of knowledge from the network. Knowl- the source and incorporated into the project.
edge implementation refers to exploiting and A high level of trust means a high level of
integrating the acquired knowledge when the embeddedness, which results in reciprocal ex-
integrator tries to develop technical solutions. change and joint problem solving (Granovetter,
Knowledge transfer means the integrator shares 1985).
the problem-solving experiences and generated In CoPS innovation networks, all actors which
knowledge with partners and users to facilitate specialize in some capabilities are engaged in the
their development of sub-systems. innovation process, investing complementary re-
The embeddability mechanisms can be identi- sources and skills in their cooperation, which
fied as relational linkages and structural linkages results in heavy dependence on each other for
(Dayasindhu, 2002). Lui et al. (2005) contend that successful CoPS development. Moreover, in a
the relational context is an important factor CoPS context, the existence of the innovation
having impacts on the transfer of strategic orga- network will last for a long project duration, so
nizational knowledge among partners. It has been the integrator, sub-contractors and users will
asserted relationship characteristics such as trust, think of their work together as a long-term
long-time orientation and commitment may affect commitment. In such a network, with tight lin-
knowledge flow across organizational boundaries kages among partners, effective knowledge acqui-
(e.g. Mavondo and Rodrigo, 2001). On the other sition from the network may be easier. Therefore,
hand, resource linkage between collaborative we assume embeddability among network mem-
partners, referring to the non-recoverable and bers helps the integrator acquire knowledge from
idiosyncratic investment that firms make in sub-contractors and users. The following sub-
a particular relationship (Parkhe, 1993), is a proposition is posited:

424 R&D Management 38, 4, 2008 r 2008 The Authors


Journal compilation r 2008 Blackwell Publishing Ltd
Qualitative study of inter-organizational knowledge management

P1a: The tightness of network embeddability social commitment that allows for the transfer of
has a significant effect on the degree of knowl- idiosyncratic and privileged knowledge, value
edge acquisition from the network may be created (Uzzi and Lancaster, 2003). In
terms of resource linkage, Ring and Van de Ven
(1992) contend the greater the risk resulting from
3.3.1.2. The interrelationship between network em- high investment, the more complex and tight the
beddability and knowledge implementation. Ac- linkage is likely to be, including reciprocal infor-
cording to Santoro and Gopalakrishnan (2000), mation and knowledge exchange, with the pre-
effective knowledge implementation is facilitated ference of organizations involved for high
by organic structures, i.e. the resource linkages embeddability. Therefore, when there is a high
and relational linkages embedded in the network. level of embeddability, a focal firm in a network
As a matter of fact, in a multi-company project has more confidence about the partners’ abilities
team, the team’s ability to implement what it has and expected behavior (Das and Teng, 1998), and
learned, i.e. by applying the information gathered is more willing to share its technical capabilities,
from partners and users, increases when the innovative ideas and visions with the partners
management of all the companies involved share (Santoro and Gopalakrishnan, 2000).
the vision of the project and are willing to invest In a CoPS context, every participant company
in relation specific assets (Lynn et al., 2000). On has its own technical competitive advantage, and
the basis of case studies in construction industry, multi-forms of coordination mechanism create a
Gann and Salter (2000) assert the linkages among network with fuzzy boundaries and transactions
the integrator and other actors in the network are among actors (Hobday et al., 2000). For the
far more complicated than standardized business system integrator to achieve successful develop-
contracts, and the embedded relationships pro- ment of a CoPS project, it must be willing to share
vide a lot of opportunities for both integrator and task-specific knowledge in the network with trust
sub-contractors to absorb tacit knowledge such as and a belief in resource interdependence.
know-how and know-who through joint work. Based on the above arguments, we thus state
The knowledge synergy framework developed by the following sub-proposition:
Nielsen (2005) also proposes that intensified em-
beddability serves to increase the effectiveness of, P1c: The tightness of network embeddability has
and relative capacity to internalize and integrate a significant effect on knowledge transfer
external knowledge.
Therefore, the following relationship can be 3.3.1.4. Network embeddability and KM: experi-
posited: ences from a case study. In the case of SysPorj
P1b: The tightness of network embeddability A, the system could be functionally divided into
has a significant effect on knowledge implemen- three major sub-systems: electric power supply
tation sub-system, environment control sub-system and
disaster alarm sub-system. Specifically, all sub-
systems could be further broken down into more
3.3.1.3. The interrelationship between network than 20 modules with nearly 40 interfaces, includ-
embeddability and knowledge transfer. In hyper- ing hardware platform, component software, in-
competitive business environments, enterprises terface program and field engineering, etc. Over
are making an effort to go beyond the transfer 40% of these sub-systems needed to be custo-
of day-to-day operational information with part- mized specially according to the operator’s re-
ners by establishing long-term and tight embedd- quirements and field characteristics of the MTR
ability and sharing knowledge such as market line. It took nearly 3 years to implement this
trends and future product roadmaps (Malhotra project from initial negotiation to final delivery.
et al., 2005). On the basis of a qualitative study on Integrator A was not specialized in every tech-
Indian software companies, Dayasindhu (2002) nological field needed in the project, so it had to
states that a tight level of embeddability is bene- seek appropriate sub-contactors and suppliers as
ficial to knowledge transfer. Malhotra et al. partners. For standardized modules/components,
(2005) further point out embeddability based on the criteria of outsourcing were cost and delivery
close and tight relationships may promote knowl- time. For unstandardized modules, which needed
edge exchange that enables firms to enhance their to be jointly developed, the reputation and
current competencies and processes. In other experiences of sub-suppliers became the most
words, when business processes are embedded in important consideration.

r 2008 The Authors R&D Management 38, 4, 2008 425


Journal compilation r 2008 Blackwell Publishing Ltd
Eric W.T. Ngai, Chen Jin and Tong Liang

At the beginning, explicit business contracts frame servers. The technological fields involved in
were the major coordination mechanism among the project could be divided into two categories:
Integrator A, MTR operator and partners. As software technology, including basic control pro-
SysProj A lasted for nearly 3 years, with the gram, software embedded in modules, design
progress of development, the actors in the net- blueprint, and control process solution; hardware
work invested significant resources and energy components consisting of computer, server, net-
into the project, which led to mutual understand- work infrastructure, signal collection device, ac-
ing in a comparatively efficient manner. As a tuators, solenoids, sensors, temperature switches,
result, trust was established, and the linkages terminal display, and integrated electro-circuit,
among actors (see Fig. 1) were not based only etc. As a consequence of system complexity, the
on contract. For example, the MTR operator project started in mid 1997 and ended in early
frequently generated new needs and requirements 2002. The development cycle lasted for over 4
were changed several times within the project years.
duration, resulting in two major modifications Similar to Integrator A, Integrator B did not
of the contract between Integrator A and the have all the capabilities and resources necessary
MTR operator. Every time, after contract mod- to implement the project, and had to search for
ification, Integrator A received a substantial partners and collaborators. Not only enterprises,
number of unsolicited suggestions about adjust- but also public research units such as universities
ments to the current system design from the and institutions were taken into consideration
design center and the equipment supplier. when designing the potential supplier pool. The
Although not every suggestion was adopted, power plant played a critical role in this phase,
they provided ideas and opportunities for Inte- having the authority to reject a supplier even if it
grator A to draw an improved design blueprint. was preferred by Integrator B. Finally, five core
In CoPS innovation, this situation is very com- collaborators, the power plant and Integrator
mon because a detailed description of the whole B composed the core innovation network (see
system cannot be developed at the bidding stage, Fig. 2).
and sometimes even the user does not have a clear The power plant not only played an active role
description about the final system. On the other in choosing partners, but also endeavored to
hand, when Integrator A got new insights it establish an information sharing culture in the
would pass them to its partners and the MTR network. Integrator B, the power plant and part-
operator, in order to post the latest work progress ners agreed with a documentation mechanism to
and help the other actors to maintain consistency ensure every actor within the network could have
in development. detailed records of problems occurring in devel-
In the case of SysProj B, the Input/Output opment and corresponding problem-solving solu-
points were nearly 12,000, including seven major tions, and make reasonable classification of
sub-systems, 12 control stations and two main- respective records to help each other get a better

Standard Core project


component supplier network

Tight linkage Significant


Trust relationship Equipment supplier project-specific
investment

Design
center MTR operator
Standard Standard
component Integrator component
supplier A supplier

Power Construction
bureau company
Knowledge flow

Explicit business Component sourcing


contract

Standard
component supplier

Figure 1. The development network for SysProj A.

426 R&D Management 38, 4, 2008 r 2008 The Authors


Journal compilation r 2008 Blackwell Publishing Ltd
Qualitative study of inter-organizational knowledge management

Peripheral standardized
product suppliers

University
X

Partner Partner
A D

Integrator B
Power plant Partner A: An integrated circuit manufacturer
Partner B: A nuclear energy research center
Partner Partner C: An electric power research center
Partner Collaborative C Partner D: An international specialist firm,
B innovation focusing on system integration

Tight linkage
Sourcing
Weak linkage
Peripheral standardized
product suppliers

Figure 2. The development network for SysProj B.

understanding of work progress. The sharing quency and breadth of information exchanged,
mechanism proved to be very important in coping which accelerates knowledge creation and pro-
with discontinuity resulting from major technical blem correction. As mentioned before, the com-
iterations or R&D engineer changes. plexities in CoPS innovation lead to frequent
Although the two case projects were alike to communication and joint decision-making during
some extent, there still existed some differences. system development, and the stability and flex-
In SysProj A the MTR operator did not put much ibility of the network are important for successful
effort into establishing direct access to suppliers. collaboration. Based on the above arguments, we
In contrast, in SysProj B the power plant played define inter-organizational interaction intensity as
an active role in creating direct access to suppli- the frequency and degree of joint problem sol-
ers. According to our interviews, both ‘direct ving, technical meetings, and communication.
access’ configuration and ‘indirect access’ struc- Furthermore, we propose inter-organizational in-
ture mobilized KM activities. Table 2 shows a teraction intensity is another important dimen-
summary of findings based on extensive personal sion in inter-organizational collaboration
interviews that indicate how evidence from the affecting KM within a CoPS development net-
two case projects supports the developed network work. We thus state the following proposition:
embeddability propositions.
Proposition 2 (P2): Inter-organizational inter-
action intensity has a significant influence on
3.3.2. How does interaction intensity influence
inter-organizational KM
inter-organizational KM?
Mavondo and Rodrigo (2001) have conceptua-
lized collaboration as a multi-dimensional con-
struct including joint interaction, harmony and 3.3.2.1. The relationship between inter-organiza-
flexibility. Some ‘interaction boosters’ have been tional interaction intensity and the degree of knowl-
identified (Inkpen and Dinur, 1998): formal meet- edge acquisition. When several companies with
ings (e.g. periodic visits and work meetings), different resources and complementary capacities
informal meetings (e.g. telephone and E-mail), work together, inter-organizational interaction
and collaboration (e.g. dialogue and teamwork). can be considered as a facilitator of knowledge
Uzzi (1997) asserts that appropriate joint problem acquisition (Cegarra-Navarro, 2005). In an inter-
solving and interaction arrangements not only organizational network, the interaction among
enable network actors to coordinate and work actors provides the channels for communication,
out problems, but also provide rapid explicit representing the richness of media. Consequently,
knowledge feedback. Specifically, Nielsen (2005) there exists much opportunity for the focal actor
proposes that the more the social interaction is to absorb external knowledge and learn from the
built up within networks, the greater the fre- network (Inkpen and Dinur, 1998). In CoPS

r 2008 The Authors R&D Management 38, 4, 2008 427


Journal compilation r 2008 Blackwell Publishing Ltd
Eric W.T. Ngai, Chen Jin and Tong Liang

Table 2. Network embeddability propositions in case studies.


Sub- Case Degree of Gathered information
proposition project support

P1a (references: SysProj A Supported As the project manager said, ‘The main difficulty in
Granovetter, 1985; cooperation was that the engineers from other firms might
Hansen, 2002) have been told to take chances on opportunism . . . To be
honest and take responsibility actively were the essential
qualities that we tried to forge in our relationships with
partners’. The partners were willing to provide their new
ideas about the MTR operator’s requirements other than
just work on allocated assignments, in order to help
Integrator A design and redesign the whole system
structure.
SysProj B Supported The power plant played an active role in building direct
relationships with sub-contractors, which further
contributed to trust and commitment establishment.
Moreover, a greater density of social linkages among
engineers was encouraged to maximize the potential
knowledge and experience to be captured. Scanning,
tracking and sourcing technical experiences among
partners became very easy. As the project manager said,
‘The user and we were heavily focused on this project,
choosing suppliers, and breaking down the functionality
requirements into specification parameters together. . .
valuable insights about holistic system design were
generated’.
P1b (references: Gann SysProj A N/A Limited information was gained in the interview about this
and Salter, 2000; proposition
Hansen, 2002; SysProj B Supported ‘Direct access’ configuration made the network more
Nielsen 2005) transparent and collaborative. As one engineer said, ‘To
facilitate integration of multi-disciplinary knowledge, we
tried to develop our own shared community with agreed
elements . . . such as shared team tags and team based
slogans’.
P1c (references: SysProj A Supported The actors were very willing to share with, and help each
Dayasindhu, 2002; other. As the project manager said, ‘Both the user and
Hansen, 2002; partners trusted us very much due to our honesty . . . They
Uzzi and were willing to cooperate based on our solution’
Lancaster, 2003) SysProj B Supported The power plant allocated staff to the development field,
and provided channels to connect core suppliers directly.
Knowledge could be passed to every actor efficiently. The
project manager told us ‘For example, some times the
integrated circuit manufacturer directly discussed with the
power plant about the special requirements for circuit
arrangements in different buildings. For this kind of
information less relevant to system integration, we didn’t
need to put much manpower or resources as an intermediary,
which saved a lot of our time and energy’.
NA, not accessible in this case.

development, all actors meet together regularly to P2a: The inter-organizational interaction inten-
report and review their progress in latest devel- sity has a significant effect on the degree of
opments. Wuyts et al. (2004) find that both the knowledge acquisition
vendor and buyer of complex projects value
strong ties and frequent interaction that run
from suppliers through the vendor to the buyer, 3.3.2.2. The relationship between inter-organiza-
because the interaction among actors provides the tional interaction intensity and knowledge imple-
system integrator with the direct access to com- mentation. Through interactions with others,
plex knowledge. Hence, the following relationship firms can combine external knowledge with exist-
can be proposed: ing knowledge. According to Yli-Renko et al.

428 R&D Management 38, 4, 2008 r 2008 The Authors


Journal compilation r 2008 Blackwell Publishing Ltd
Qualitative study of inter-organizational knowledge management

(2001), the interaction creates a context in which et al., 2004). Hence, we propose the following
newly received knowledge can be applied and sub-proposition:
exploited. Malhotra et al. (2005) offer support
P2c: The inter-organizational interaction inten-
for the role of interaction in knowledge imple-
sity has a significant effect on knowledge transfer
mentation in their argument that cooperative
partners with high interaction may achieve high
performance by the development of inter-organi- 3.3.2.4. Interaction intensity and KM: experiences
zational integrative process mechanisms and joint from case study. In the case of SysProj A,
decision making to integrate privileged knowl- Integrator A endeavored to launch interaction
edge. In a CoPS context, intensive interactions initiatives including both an IT-based project
with clients and partners often result in tailoring reporting system and through a social approach,
unique solutions regarding clients’ needs (Oshri through which problem solving experiences could
and Newell, 2005). In addition, informal commu- be acquired, integrated, filed and disseminated
nication and interaction among engineers allo- among the innovation network members.
cated to a CoPS project from different firms IT-based tools were used to capture and translate
contribute to maintaining technical consistency the MTR operator’s requirements into design
and finding new combinations of resources or parameters. For instance, because the MTR line
activities for unexpected variation deviating was still under construction when the system was
from the planned trajectory (Gann and Salter, developed, Integrator A could not test the proto-
2000). Frequent and, typically, face-to-face com- type system in the actual MTR construction field.
munication and advanced information technolo- Integrator A had to build a special emulation
gies (IT) that can realize direct knowledge platform for simulating a field environment in
exchange are especially suited to the joint search which the system would be partially operated.
for solutions in CoPS environments, because they The platform could make an accurate simulation
permit sufficient density and iterability to knowl- and take measurements, and compare the system
edge integration (Marshall and Brady, 2001). parameters under different operation circum-
Drawing on the above material, we propose the stances, which facilitated Integrator A’s under-
following sub-proposition: standing of the MTR operator’s needs.
On the other hand, social interactions evolved
P2b: The inter-organizational interaction inten-
with the progress of development. Engineers
sity has a significant effect on knowledge
from all parties worked together to find solutions
implementation
for unplanned events. Their progress was re-
ported to their respective superiors regularly.
3.3.2.3. The relationship between inter-organiza- The reports were reviewed at monthly technical
tional interaction intensity and knowledge trans- meetings at which senior management from In-
fer. According to Dulaimi et al. (2003), tegrator A and the sub-contractors were present.
knowledge and technology transfer can be con- Certainly it is unrealistic to expect the total
sidered as a process of interaction. In the analo- elimination of task ambiguity via intensive inter-
gous procedures identified by Inkpen (1996) for action. In all of the sub-systems, only 60% were
knowledge transfer within joint ventures, interac- inter-connected successfully at the beginning,
tion between joint ventures and the parent com- which resulted in three major iterations of the
pany is a very important factor. Inkpen and integration process.
Dinur (1998) further conclude that technical in- In the case of SysProj B, Integrator B also
teraction is especially effective in knowledge adopted the codification approach and persona-
transfer at the group or organization level. Prior lization approach simultaneously to KM. The
empirical research on international business also intent of Integrator B was to build up a real-
suggests that interaction intensity among head- time monitoring and tracking system, which con-
quarters and different subsidiaries of multina- sisted of several levels of IT technologies from
tional corporations (MNCs) facilitates the knowledge transport to application (e.g. Local
frequency and thoroughness of knowledge trans- Area Network). However, the mere existence of
fer (Subramaniam and Venkatraman, 2001). In an IT system did not mean the action team would
terms of an innovation network for complex be willing to use it. There existed a big difference
products, direct interaction between buyer–inte- between the official operation procedures of the
grator–supplier is also considered as a way for the IT-based system and the actual actions of the
integrator to transfer complex knowledge (Wuyts engineers.

r 2008 The Authors R&D Management 38, 4, 2008 429


Journal compilation r 2008 Blackwell Publishing Ltd
Eric W.T. Ngai, Chen Jin and Tong Liang

Integrator B also implemented a social interac- activities are significantly related to NPD perfor-
tion approach to KM, which was more effective mance. In a study of 139 Spanish optical compa-
than the IT-based approach. At the pre-project nies, Cegarra-Navarro (2005) concludes that a
stage, Integrator B allocated several specialists positive relationship exists between KM activities
and sales engineers specializing in computerized and organizational learning within inter-organi-
control systems to the power plant to discuss the zational alliances. Hence, we posit the following
system structure and functionalities. In project proposition:
implementation, technical meetings were the for-
Proposition 3(P3): Inter-organizational KM
mal mechanism for coordinating system develop-
has a significant influence on CoPS project
ment. Meeting reports signed by all participants
performance
were seen as having the same status as a technical
contract. Besides the formal meetings, direct
3.3.3.1. The relationship between the degree of
commitment-based communication between
knowledge acquisition and project performance.
engineers in the project team overcame low
Cooper and Kleinschmidt (1987) stress the
efficiency of information transfer through the
knowledge acquired from customers and the
traditional hierarchical procedures, and brought
market is one of the critical success factors for
about most of the problem-solving. The project
NPD. Through an investigation of 120
manager from Integrator B said, ‘1/3 of our time
young technology-based firms in the United
was allocated to mutual communication and
Kingdom, Yli-Renko et al. (2001) contend the
coordination, including interaction with the
more knowledge a technology-based firm
customer’.
acquires about customer needs and ways of doing
In Table 3, a summary of the data is presented,
business, the more efficiently it will be able to
which shows how the evidence from the two case
provide its product and service. They further
projects supports the interaction intensity propo-
propose knowledge acquisition increases tangible
sitions.
NPD performance in two aspects: enhanced
speed-to-market and reduced costs. Simonin
3.3.3. How does inter-organizational KM have
and Helleloid (1993) also claim that the
effect on project performance?
know-how acquired from collaboration
Collaboration can lead to a number of tangible
contributes to tangible and intangible perfor-
and intangible benefits. Only looking at the tan-
mance. Besides the argument about the impact
gible benefits of collaboration may be misleading
on tangible NPD performance, Yli-Renko et al.
(Simonin and Helleloid, 1993). Tangible perfor-
(2001) extend the impact of knowledge acquisi-
mance includes generating profits, improving
tion on enhancing the breadth and depth of
market share, and sustaining competitive advan-
relation-specific knowledge, and claim that it
tages. Relative to the tangible performance,
increases the willingness to develop new products
much less attention is given to the intangible
for key customers, providing the potential for new
output of new product development (NPD). Or-
innovative combinations.
ganizational learning explanation provides an-
The number of subsystems incorporated in
other angle for exploring inter-organizational
CoPS is very high (Hobday and Rush, 1999;
KM and its influence on CoPS project perfor-
Hobday et al., 2000), and the product needs to
mance. As a result of specific collaboration,
be compatible with a broader technical platform
learning new skills and having new experiences,
for future system expansion. Logically, providing
and enhancing social capital can be considered as
CoPS requires a combination of many areas of
intangible performance (Simonin and Helleloid,
specialized knowledge. Knowledge acquisition
1993). In a study of the home appliance industry
has been described as one of such combinative
by Sobrero and Roberts (2002), the collaboration
capabilities (Kogut and Zander, 1992). Hence,
outcomes are decomposed into two categories:
in order to combine the users’ requirements,
efficiency (e.g. cost reduction) and learning (e.g.
implement development plans, and resolve pro-
development of new solutions). Drawing on the
duct development deficiencies, the CoPS integra-
above arguments, CoPS project performance can
tor needs to access relevant information and
be defined as the synthesis of tangible efficiency
knowledge. We thus state the following
(system success) and intangible learning (learning
sub-proposition:
performance).
Empirically, Lynn et al. (2000) examine the P3a: The degree of knowledge acquisition has a
KM processes in NPD and conclude that KM significant effect on project performance

430 R&D Management 38, 4, 2008 r 2008 The Authors


Journal compilation r 2008 Blackwell Publishing Ltd
Qualitative study of inter-organizational knowledge management

Table 3. Interaction intensity propositions in case studies.


Sub-proposition Case Degree of Gathered information
project support

P2a (references: Inkpen SysProj A Supported As a result of IT-based and social approaches to
and Dinur, 1998; interaction, information from the MTR operator and
Hobday and Brady, partners was easily accessed. As an IT-service
2000; Cegarra- department member of staff said, ‘We (the network
Navarro, 2005) members) achieved an agreement to upload and
download work progress with IT tools . . . something
like Lotus Notes and GroupWise . . . Detailed inputs
and outputs in every subsystem were documented . . .
We also recorded the changes of functionalities
generated by development iterations, whether for the
holistic system and subsystems...of course, the meeting
memos were stored in the central database too’.
SysProj B Supported The formal and informal approaches to dealing with
knowledge sources (partners) were seen as vehicles to
carry knowledge. As the project manager said, ‘At the
beginning, the IT-led interaction tools were not
designed very well and didn’t work as expected . . .
Joint work together was a supplement of IT tools,
which gave us a lot of ideas to shape the solutions as we
went forward’.
P2b (references: Lynn SysProj A Supported The interaction was largely project specific.
et al., 2000; Integrator A felt the partners were really technically
Yli-Renko specialized, so the problem of ‘not invented here’
et al., 2001; seldom occurred. As one engineer said, ‘There were a
Wu and lot of areas where we needed to rely on partners . . . We
Tsai, 2005) trusted the work they passed to us . . . 60% of
functionality was actualized at the first time we carried
through debugging’.
SysProj B Weakly Although Integrator B adopted initiatives to interact
Supported with partners and the power plant, system integration
didn’t go smoothly. As one of the engineers noted,
‘especially at the integration and debugging stages,
many sub-systems/modules didn’t work once they were
inter-connected together, though they functioned very
well separately’. Moreover, from system start-up to
final system installation, all actors gathered formally
18 times to hold technical meetings, 6 of which
directly led to modification of the prior contract.
P2c (references: SysProj A Supported In the monthly meetings, the participants discussed
Subramaniam and the context-specific difficulties in detail in face-to-face
Venkatraman, 2001; communication, which not only provided
Wuyts idiosyncratic and tacit insights about current
et al., 2004) development processes, but also enhanced the
atmosphere of knowledge sharing among all parties.
SysProj B Supported To achieve successful development and operation of
the system, Integrator B was willing to share its
technical know-how without worrying about
knowledge leakage. For example, after system
installation, Integrator B deployed 18 technicians and
coordinators at the control center in the power plant,
training the users in how to manipulate the system
appropriately, monitoring the system operation, and
capturing first-hand feedback information that was
important for system maintenance and update.

3.3.3.2. The relationship between knowledge areas of technology (Cohen and Levinthal, 1990),
implementation and project performance. NPD whether in intra-organizational or inter-organiza-
requires the integration and combination of tional contexts. Alavi and Leidner (2001) contend
specialized knowledge inputs from many different that performance often depends on an ability to

r 2008 The Authors R&D Management 38, 4, 2008 431


Journal compilation r 2008 Blackwell Publishing Ltd
Eric W.T. Ngai, Chen Jin and Tong Liang

turn knowledge into effective action. Further- To make sure the CoPS projects are successful,
more, some studies have pointed out that knowl- the CoPS manufacturers need to share knowledge
edge implementation is positively related to NPD with partners during joint development. As a
project performance, including market share, result, in inter-organizational CoPS innovation
sales, and profit, etc. In a study on complex networks, knowledge transfer and dissemination
software development by Tiwana (2004), knowl- may be intensified across firms’ boundaries.
edge implementation is highlighted to have had Malhotra et al. (2005) have developed a proposi-
impacts on various dimensions of software devel- tion indicating that enterprises engaged in ex-
opment performance, such as increasing software changing a range of high quality information
development effectiveness and reducing defect with partners will achieve high performance
density. In terms of learning performance, the through modular design of their interlinked
model of organizational learning process devel- processes. Therefore, we develop the following
oped by Crossan et al. (1999) suggests that knowl- proposition:
edge implementation is necessary for enterprises
P3c: Inter-organizational knowledge transfer
to update their knowledge and competences. In
has a significant effect on project performance
situations where the project team has an abun-
dance of records and documents, knowledge im-
plementation may be crucial for learning (Lynn 3.3.3.4. Inter-organizational KM and CoPS pro-
et al., 2000). Sobrero and Roberts (2002) ject performance: experiences from case study. In
conclude that the choice of coordination and the case of SysProj A, the system was put into
knowledge integration mechanisms must be actual operation in 2002 and experienced only
accurately balanced to have a positive impact on two minor field adjustments before final acknowl-
both long-term learning and short-term efficiency edgement by the MTR operator and the govern-
outcomes. ment. As the system was the first project
Because the system integrator has a lot of Integrator A had contracted in the urban public
opportunity to access knowledge from the transportation sector, much knowledge about
innovation network, the acquired knowledge MTR lines was new to it. The success of this
needs to be applied in the CoPS development project brought Integrator A into the urban
process, to make the inter-linked subsystems run public transportation sector and made itself
holistic functionalities and to update the existing accepted, responding to a call from the govern-
knowledge base. Following the above statements, ment for indigenous innovation in backbone in-
the relationship between knowledge implementa- dustries and winning the confidence of the
tion and CoPS project performance is expected: government. In the following few years, several
cities in China got approval from the central
P3b: Inter-organizational knowledge implemen-
government to construct MTR lines. Through
tation has a significant effect on project perfor-
the accumulated technical and organizational
mance
experiences from SysProj A, Integrator A won a
considerable portion of the contracts for devel-
3.3.3.3. The relationship between knowledge trans- oping control systems for these MTR lines.
fer and project performance. Prior studies have In the case of SysProj B, Integrator B also
reported the relationship between knowledge delivered the system successfully, which not only
transfer and performance at both firm and project directly generated considerable financial benefits,
level. Hoopes and Postrel (1999) conclude firms but also strengthened the competitive status of
that encourage the free flow of knowledge within Integrator B. The power sector is an industry with
the organization may be more competitive in a high degree of government monopoly in China.
NPD. Cegarra-Navarro (2005) believes organiza- The successful installation and operation of the
tions that track customer preferences and share system also improved the relationship between
their knowledge with partners can perform at a Integrator B and local government, opening an
higher level, delivering more satisfaction to cus- opportunity window for potential contracts.
tomers with the help from partners. Wu and Tsai Moreover, through knowledge exchange with
(2005) see knowledge transfer as one of the other parties, some technical weaknesses and
knowledge-creating activities, and empirically re- organizational routines were improved. Table 4
port a significant relationship among knowledge- provides a summary of the data that shows how
creating activities, intellectual capital and KM the two case projects support the performance
effectiveness (i.e. learning performance). propositions.

432 R&D Management 38, 4, 2008 r 2008 The Authors


Journal compilation r 2008 Blackwell Publishing Ltd
Qualitative study of inter-organizational knowledge management

Table 4. Performance propositions in case studies.


Sub-proposition Case project Degree of Gathered information
support

P3a: (references: Cooper SysProj A Supported To gain experience and get into the sector were the
and Kleinschmidt, 1987; primary goals of this project, both of which were
Simonin and achieved. The executive manager said ‘As our first
Helleloid, 1993; project in the public traffic sector, we learned a lot of
Yli-Renko new things . . . Profit was a secondary consideration at
et al., 2001) that time . . . as long as it covered the cost . . . even a
small loss was acceptable’
SysProj B Supported Senior management was convinced of the
contribution of idiosyncratic know-how to the
project’s success. In other words, both the
architecture knowledge and distributed component
knowledge (e.g. digital electronics and signal
processing etc.) made the system successful. In
addition, the acquired component knowledge from
partners helped broaden Integrator B’s in-house
R&D technological capabilities.
P3b: (references: Lynn SysProj A Supported Both the employee from the IT service department
et al., 2000) and the project manager noted that their emulation
platform was very useful in simulating system
integration, which saved much effort when the actors
did it in the field and decreased the error occurrence
after the system had been put into operation.
Furthermore, successful integration of all subsystems
enhanced the firm’s reputation on DCS integration,
representing its high technical capability.
SysProj B Supported Integrator B got profit and acknowledgement from
both the user and local government through this
project. As one senior manager said, ‘If you have some
special knowledge . . . such as connecting all subsystems
with appropriate structures and interfaces, something
like that, while other competitors don’t, it puts you in a
privileged position . . . it means other power plants or
users in other sectors may put you on their priority list
as potential DCS suppliers . . . we were satisfied with
the financial benefits from this project’
P3c: (references: Crossan SysProj A Supported The project manager told us that about 60% of
et al., 1999; Hoopes and learning opportunities were related to the interaction
Postrel, 1999; Malhotra with other parties in the innovation network, instead
et al., 2005) of to existing lessons from former projects that
Integrator A had finished. ‘Sharing experiences had
been a kind of routine among our partners’ For
example, Integrator A filed a useful handbook as
shared guidelines to handling unexpected events
through joint development, especially in subway
environments.
SysProj B Supported In fact, the system delivery was delayed for about half
a year due to one supplier being unable to integrate
the fire and smoke alarm sub-system to the whole
system. Both Integrator B and the power plant
allocated engineers to that supplier, responding to the
supplier’s request for help, rather than just providing
relevant technical documents. The allocated
engineers provided assistant technical knowledge and
the interface information, facilitating the sub-
system’s interconnection to other modules. Owing to
this, the system was successfully developed by all
actors’ joint work.

r 2008 The Authors R&D Management 38, 4, 2008 433


Journal compilation r 2008 Blackwell Publishing Ltd
Eric W.T. Ngai, Chen Jin and Tong Liang

Based on a review of the literature and the In traditional innovation frameworks, relatively
findings from our case studies, we propose pro- clear boundaries make enterprises prefer in-
positions 1, 2, 3 and their respective sub-proposi- house development and manufacturing. Even
tions (see Fig. 3a). The inter-organizational other relevant actors involved in product devel-
collaboration, KM and CoPS theories provide opment and manufacturing are actually excluded
the underlying variables. A multiple case study from the entity.
approach was adopted to gain an understanding KM in a CoPS context aims not only at
of the relationship between inter-organizational codifying, storing and distributing learnt experi-
KM and its impact on CoPS project performance. ences in project development, but also at creating
The postulated relationships are shown in Fig. 3a and providing the organizational productive link-
and b. The framework explains KM in a CoPS age between knowledge and practice (Marshall
context characterized by inter-organizational col- and Brady, 2001). Nevertheless, little is known
laboration and knowledge intensiveness. about the interaction mechanism that facilitates
knowledge exchange and learning in CoPS pro-
jects (Oshri and Newell, 2005). Both cases in this
article have demonstrated an IT-based approach
4. Discussion and implications and social interaction approach were adopted as
vehicles for inter-organizational KM. However,
4.1. Dimensions of inter-organizational use of an IT system may not be favored by project
collaboration and KM in CoPS teams in CoPS innovation, as the engineer in
Although much of literature has contributed to SysProj B points out in this study. There may
the body of knowledge of KM, this article points be a dissonance between the intent of emphasizing
out that prior studies have focused on KM at an IT-based approach by management and actual
intra-organization level, not much at inter-orga- preference for social interaction by engineers in
nizational network level. Furthermore, little the field (Marshall and Brady, 2001).
research has been conducted on KM and inter- Although prior studies seldom discuss KM in
organizational networks within a CoPS context. CoPS, their arguments did offer a good starting

a Network P1a + The degree of


embeddability knowledge acquisition
P1b + P3a +

P1c +

Project
Inter-organizational P3b +
knowledge implementation Performance
P2a +

P2b +
Inter-organizational P3c +
Inter-organizational
interaction knowledge transfer
P2c +

b Knowledge Knowledge Knowledge


acquisition implementation transfer

Network
embeddability P1

Inter-organizational P3 Project
KM processes Performance

Inter-organizational
P2
interaction intensity

Figure 3. (a) Conceptual model of inter-organizational KM and its impact on CoPS project performance with the sub-
propositions. (b) Conceptual model of inter-organizational KM and its impact on CoPS project performance with main
propositions.

434 R&D Management 38, 4, 2008 r 2008 The Authors


Journal compilation r 2008 Blackwell Publishing Ltd
Qualitative study of inter-organizational knowledge management

point and references for our research. Overall, by action and communication provide a more suita-
a review of the relevant arguments from existing ble platform for supporting knowledge practice
literature and experiences from two cases of when facing unplanned complexity, whether by
computerized control systems, this paper devel- joint action, through technical meetings or
ops some propositions and builds up a conceptual through periodic visits (Subramaniam and Ven-
model, not only combining the three cross-fields katraman, 2001). Both integrators in our cases
discussion of inter-organization collaboration, adopt IT-based tools and personalization ap-
KM, and CoPS development, but also exploring proaches to contact network members, and en-
some important issues such as how the character- courage frequent formal and informal meetings
istics of collaboration in a CoPS innovation net- for the project duration. Therefore, the elements
work has impacted on inter-organizational KM related to joint decision-making, joint design,
and in turn on the final project performance. frequent technical meetings and frequent techni-
Table 5 presents a brief summary of implications, cal reviews could be considered in future studies
lesson learnt and constructs along their proposed for measuring the interaction intensity.
measurements for the inter-organizational KM in
 Knowledge acquisition
CoPS context.
Many KM researchers have noted that knowl-
edge acquisition is vital for learning. Knowledge
4.2. Measurement of inter-organizational acquisition in this study is defined as concern with
collaboration and KM in CoPS the access and absorption of knowledge from a
innovation network. According to our experience gained
from case analyses, CoPS consist of a number
In this study, the connections between inter- of sub-systems and sub-components. The system
organizational KM, network embeddability and integrators focus on absorbing technical data and
interaction intensity have been explored in a information from network members during joint
holistic way. However, the difficulty of better development, such as sub-system functionalities,
understanding the cognitive aspect still remained, system design, interface connection and system
and as this may not be realistically possible, one integration. Hence, we propose, in a CoPS con-
could explore it more through the form of mea- text, that the measurement of knowledge acquisi-
surement. Further studies could take on the task tion in any future study should consist of the
of investigating how to measure and quantify the elements of functionality information acquisition,
constructs. The measurement of individual con- design information acquisition, interface informa-
structs of the proposed conceptual model is dis- tion acquisition and integration information ac-
cussed below. quisition.
 Network embeddability  Knowledge implementation
Previous sections have shown that embeddabil- Crossan et al. (1999) argue that implementing
ity is closely related to the resource and relational acquired knowledge into business processes may
linkages between collaborators. Many reports enhance organizational competitiveness. Similar
have developed some well-defined relational vari- conclusions have been drawn by researching KM
ables, such as trust and commitment (e.g. Lui in the NPD process (Lynn et al., 2000). According
et al., 2005). Our case analyses also indicate that to our case experience, successful CoPS develop-
good relationships, including non-market lin- ment is inevitably achieved by several iteration
kages and implicit contracts, are built upon processes, including continuous problem identifi-
mutual resource dependence and trust. We sug- cation and correction, and applying acquired
gest, in a further study, that elements such as knowledge to development. Integrator A did a
trust, significant investment into the project, asset lot to remove the problem of ‘not invented here’,
complementarity and collaboration vision could making engineers from among network members
be used to generate the original statements on the put their hearts into co-operative development,
instrument of network embeddability. and so its system integration went more smoothly
 Interaction intensity than that of Integrator B. Hence, the elements
related to process improvement, and identifying
In KM literature, the importance of commu- and correcting problems could be considered as
nication and interaction has been widely studied. useful in future studies for measuring knowledge
In other words, close relationships, direct inter- implementation.

r 2008 The Authors R&D Management 38, 4, 2008 435


Journal compilation r 2008 Blackwell Publishing Ltd
Eric W.T. Ngai, Chen Jin and Tong Liang

Table 5. Brief summary of inter-organizational KM issues in a CoPS context.


Construct Lesson Learnt Implication Proposed Measurement

Sysproj A Sysproj B

Network Co-operation Both the user and Asset specificity, trust, 1.Trust
embeddability mechanisms including Integrator B and commitment are 2. Significant
explicit contract and endeavoured to foster needed to strengthen investment into project
non-market good relationships in embeddability 3. Asset
relationship were built the network, based on complementarity
up in the innovation mutual resource 4.Collaboration vision
network dependence and trust
Interaction IT-based reporting Codification and Engineers from all 1.Joint decision-
intensity system and social personalization parties jointly acted as making
approach were approach to development 2. Joint design
launched and were interaction, and formal facilitators. As vehicles 3. Frequent technical
effective. Frequent and and informal to carry knowledge, meetings
regular meetings were communication various interaction 4. Frequent technical
held methods were mechanisms were reviews
introduced needed to expand
coordination and
communication
Knowledge Unsolicited suggestions Documentation Tight linkages and 1. Functionality
acquisition came from sub- mechanisms, social intensive interaction information
contractors. Reporting interactions among created channels for acquisition
system and technical engineers and ‘direct knowledge flow from 2. Design information
meetings also provided access’ configuration partners and users to acquisition
opportunities to helped the innovation system integrators 3. Interface
acquire knowledge network be more information
transparent, which acquisition
promoted knowledge 4. Integration
tracking and sourcing information
acquisition
Knowledge The problem of ‘not Although a shared Nearly all inputs were 1. Process
implementa- invented here’ seldom sense of community centralized in system improvement
tion occurred, which made with agreed elements integrators to assemble 2. Revealing problems
integration and such team tags was sub-systems, creating 3. Correcting problems
corresponding iteration established, system new knowledge 4. Knowledge
easier integration didn’t go incorporation
smoothly
Knowledge All kinds of technical Right after delivery, 18 To achieve system 1.Sharing culture
transfer meetings are set as engineers were consistency and 2. Mutual document
formal platforms for deployed to the user, successful delivery, retrieval
knowledge sharing providing service for system integrators 3. Technical training
with partners training and system needed to share 4.Staff deployment
maintenance insights and ideas with 5. Using KM systems
partners and users

 Knowledge transfer even after delivery. Therefore, we believe that


knowledge sharing in CoPS development may be
Methods of knowledge sharing have been measured by incorporating the elements of a shar-
widely discussed (Nonaka and Takeuchi, 1995). ing culture, mutual document retrieval, technical
In our case studies, many effective methods such training, staff deployment and using KM systems.
as staff relocation and technical meetings were Progress can be made in further research by: (1)
adopted by both integrators, which helped all adopting a more theoretical conceptual definition
collaborators to keep technical consistency. of the individual construct that will be derived
Furthermore, after delivering the system to the from a content analysis of the literature; (2)
power plant, Integrator B still seconded engineers generating a set of items and designing, evaluat-
to the power plant to transfer the necessary ing and refining a measurement instrument
technical expertise to the users. In other words, through several iterative processes; and 3) sum-
in a CoPS context, knowledge sharing still exists marising data from administration of the instru-

436 R&D Management 38, 4, 2008 r 2008 The Authors


Journal compilation r 2008 Blackwell Publishing Ltd
Qualitative study of inter-organizational knowledge management

ment to provide a statistical profile of the extent knowledge officers in charge of liaison and knowl-
to which the characteristics of collaboration in a edge filtering should be appointed, and appropri-
CoPS innovation network will impact on inter- ate IT systems need to be set up. The knowledge
organizational KM, and in turn on the final application mechanism inside the organization
project performance. should be adjusted to allow knowledge flow
across borders. However, reciprocal knowledge
exchange will not be sustainable if the integrators
have all of their knowledge under rigorous pro-
4.3. Managerial implications tection. The integrators should also post their
By focusing on the crucial dimensions of knowl- own development progress and relevant data, so
edge-related activities and inter-organizational that all members inside the innovation network
collaboration in CoPS innovation, we find some can update their understanding of the intercon-
underlying configurations of inter-organizational nection of sub-systems.
KM in a CoPS context that can be captured,
which are meaningful to both the academic and
the business executive. From an academic 5. Conclusion
perspective, our efforts provide important
groundwork for research into the combination The complexity and richness of KM through
of cross-school theories of KM, inter-organiza- inter-organizational collaboration seems to be as
tional collaboration and CoPS development. great as its potential to contribute to both the
From a practitioner’s viewpoint, our work offers system success and learning performance that
CoPS manufacturers a preliminary framework CoPS projects may achieve. We suggest that by
and some experiences from case studies, which looking at two dimensions of inter-organizational
they can refer to when they try to structure and collaboration (i.e. network embeddability and
formulate their inter-organizational KM strate- interaction intensity), KM at the inter-organiza-
gies. More broadly, our study also answers con- tional level can be implemented so that it has an
cerns about the cognitive processes of CoPS impact on the final CoPS project performance.
innovation (Hobday et al., 2000). The conceptual framework and testable proposi-
This study has two managerial implications. tions developed in this study can also highlight
First, the integrators need to devote themselves to and strengthen the knowledge-based dimensions
forging a collaborative atmosphere amongst net- that help to explain the unprecedented economic
work members. Our case experience tells us that, success and unique competitive advantages gener-
at first, collaborators may not be willing to actively ated by CoPS. A summary of the findings has been
provide knowledge about sub-system development. presented in the form of three main propositions
As a focal actor, the integrator should not only and nine sub-propositions that are worth explor-
improve the network structure through configuring ing further. We suggest that further empirical
complementary resource inputs with collaborators, research should develop our 12 propositions into
but should also establish mutual trust and commit- hypotheses that can be tested by means of data
ment inside the network as soon as possible. The from a survey of the CoPS firms. The focus of
integrator may also call more technical meetings further research should be on the measurement of
and increase the frequency of site visits, facilitating the constructs in the proposed conceptual model.
mutual communication and interaction, which Such research could provide confirmation of our
may significantly reduce the shifting of responsi- argument about the impacts of network embedd-
bilities onto others when problems occur. Connect- ability and interaction intensity on inter-organiza-
ing key personnel from each collaborator should tional KM processes and project performance.
be encouraged, especially for those on-site R&D
engineers, because systematic consistency will only
be achieved through frequent communication. 6. Limitations
These interactions and linkages should not be
considered just as a business cost or methodology, The major limitation of this study is its sample
but also as an investment in learning. size of two companies with a similar type of
Second, the integrators need to develop a series organization. Specifically, both of them are sys-
of KM capabilities to effectively manage inter- tems integrator firms. Although both Integrators
organizational knowledge flow, which in turn A and B are the project holders, one might have
will positively affect CoPS performance. Special expected to hear the collaborators’ points of view,

r 2008 The Authors R&D Management 38, 4, 2008 437


Journal compilation r 2008 Blackwell Publishing Ltd
Eric W.T. Ngai, Chen Jin and Tong Liang

rather than having it all filtered through the Cooper, R.G. and Kleinschmidt, E.J. (1987) What
integrator. There are always limitations to the makes a new product a Winner: success factors at
generalizability of case study research, particu- the project level. R&D Management, 17, 3, 175–189.
larly with a relatively small number of cases. We Crossan, M.M., Lane, H.W. and White, R.E. (1999)
therefore believe that further research is needed to An organizational learning framework: from intui-
confirm and extend our findings. Unsuccessful or tion to institution. Academy of Management Review,
less-than-successful cases that may also be in- 24, 3, 522–537.
Das, T.K. and Teng, B-S. (1998) Between trust and
structive in helping to test the propositions could
control: developing confidence in partner coopera-
be examined as well. One could also check for any
tion in alliances. Academy of Management Review,
additional antecedents that are important in in-
23, 491–512.
fluencing inter-organizational KM processes, as
Davenport, T.H. and Prusak, L. (1998) Working
well as checking and refining those we have Knowledge. Boston, MA: Harvard Business School
identified. We believe that a survey approach to Press.
this topic will become increasingly feasible as case Davies, A. and Brady, T. (2000) Organizational cap-
study work continues, and as the relevant con- abilities and learning in complex product systems:
structs are identified, and could provide a more towards repeatable solutions. Research Policy, 29,
complete descriptive picture of the current prac- 931–953.
tice of inter-organizational KM in CoPS devel- Dayasindhu, N. (2002) Embeddedness, knowledge
opment. We hope that this study can represent a transfer, industry clusters and global competitive-
small step along the path to the improved under- ness: a case study of the Indian software industry.
standing of KM in CoPS development. Technovation, 22, 551–560.
Dougherty, D., Borrelli, L., Munir, K. and O’Sullivan,
A. (2000) Systems of organizational sensemaking for
sustained product innovation. Journal of Engineering
Acknowledgements and Technology Management, 17, 321–355.
Dulaimi, M.F., Ling, F.Y.Y. and Bajracharya, A.
This research was supported by NSFC under (2003) Organizational motivation and inter-organi-
grant number 7022004 and The Hong Kong zational interaction in construction innovation in
Polytechnic University under grant number Singapore. Construction Management and Econom-
G-YF02. The authors are grateful for the ics, 21, 307–318.
constructive comments of the two referees on an Galunic, D.C. and Rodan, S. (1998) Resource recom-
earlier version of this article. binations in the firm- knowledge structures and the
potential for schumpeterian innovation. Strategic
Management Journal, 19, 1193–1201.
Gann, D.M. and Salter, A J. (2000) Innovation in
References project-based, service-enhanced firms: the construc-
tion of complex products and systems. Research
Alavi, M. and Leidner, D.E. (2001) Review: knowledge Policy, 29, 955–972.
management and knowledge management systems: Granovetter, M. (1985) Economic action and social
conceptual foundations and research issues. MIS
structure: the problem of embeddedness. American
Quarterly, 25, 1, 107–136.
Journal of Sociology, 91, 3, 481–510.
Barthes, J.A. and Tacla, C.A. (2002) Agent-supported
Gulati, R., Nohria, N. and Zaheer, A. (2000) Strategic
portals and knowledge management in complex
networks. Strategic Management Journal, 3, 21,
R&D projects. Computers in Industry, 48, 3–6.
Benbasat, I., Goldstein, D.K. and Mead, M. (1987) The 203–215.
case research strategy in studies of information Hansen, M.T. (2002) Knowledge networks: explaining
systems. MIS Quarterly, 11, 3, 369–386. effective knowledge sharing in multiunit companies.
Cegarra-Navarro, J.G. (2005) An empirical investiga- Organization Science, 13, 3, 232–248.
tion of organizational learning through strategic Hobday, M. and Brady, T. (2000) A fast method for
alliance between SMEs. Journal of Strategic Market- analyzing and improving complex software pro-
ing, 13, 3–16. cesses. R&D Management, 30, 1, 1–21.
Chen, J. and Tong, L. (2004) The knowledge manage- Hobday, M. and Rush, H. (1999) Technology manage-
ment mechanism in CoPS innovation. Proceedings of ment in complex product systems (CoPS): ten ques-
International Engineering Management Conference, tions answered. International Journal of Technology
18–21, October Singapore, 651–655. Management, 17, 618–638.
Cohen, W. and Levinthal, D. (1990) Absorptive capa- Hobday, M., Rush, H. and Tidd, J. (2000) Innovation
city: a new perspective on learning and innovation. in complex products and system Research Policy, 29,
Administrative Science Quarterly, 35, 128–152. 793–804.

438 R&D Management 38, 4, 2008 r 2008 The Authors


Journal compilation r 2008 Blackwell Publishing Ltd
Qualitative study of inter-organizational knowledge management

Hoopes, D.G. and Postrel, S. (1999) Shared knowledge, Parkhe, A. (1993) Strategic alliance structuring: a game
‘glitches’, and product development performance. theoretic and transaction cost examination of inter-
Strategic Management Journal, 20, 837–865. firm cooperation. Academy of Management Journal,
Huber, G.P. (1991) Organizational learning: the con- 36, 794–829.
tributing process and the literatures. Organization Powell, W.W., Koput, K. and Smith-Doerr, L. (1996)
Science, 2, 1, 88–115. Inter-organizational collaboration and the locus of
Inkpen, A.C. (1996) Creating knowledge through innovation: networks of learning in biotechnology.
collaboration. California Management Review, 39, Administrative Science Quarterly, 41, 116–145.
1, 123–140. Ring, P.S. and Van de Ven, A.H. (1992) Structuring
Inkpen, A.C. and Dinur, A. (1998) Knowledge man- cooperative relationships between organizations.
agement processes and international joint venture. Strategic Management Journal, 13, 6, 483–98.
Organization Science, 9, 4, 454–468. Santoro, M.D. and Gopalakrishnan, S. (2000) The
Kogut, B. and Zander, U. (1992) Knowledge of the institutionalization of knowledge transfer activities
firm, combinative capabilities, and the replication of within industry-university collaborative ventures.
technology. Organization Science, 3, 383–397. Journal of Engineering and Technology Management,
Liu, T.L. and Shou, I. (2004) Enhancement of customer 17, 299–319.
network relationship via governance mechanism of Simonin, B.L. and Helleloid, D. (1993). Do organizations
inter-organizational core resource and core knowl- learn? An empirical test of organizational learning in
edge strategic alliance. The Journal of American international strategic alliances. In Moore, D. (ed),
Academy of Business, 5, 220–229, September. Academy of Management Best Papers Proceedings.
Lui, S.S., Ngo, H.Y. and Hon, A.H.Y. (2005) Coercive Madison, WI: Omnibus.
strategy in interfirm cooperation: mediating roles of Sobrero, M. and Roberts, E.B. (2002) Strategic man-
interpersonal and interorganizational trust. Journal agement of supplier-manufacturer relations in new
of Business Research, 59, 4, 466–474. product development. Research Policy, 31, 159–182.
Lynn, G.S., Reilly, R.R. and Akgun, A.E. (2000) Sporleder, T.L. and Moss, L.E. (2002) Knowledge
management in the global food system: network
Knowledge management in new product teams:
embeddedness and social capital. American Journal
practices and outcomes. IEEE Transactions on En-
of Agricultural Economics, 84, 5, 1345–1352.
gineering Management, 47, 2, 221–231.
Subramaniam, M. and Venkatraman, N. (2001) Deter-
Malhotra, A., Gosain, S. and EI Sawy, O.A. (2005)
minants of transnational new product development
Absorptive capacity configurations in supply chains:
capability: testing the influence of transferring and
gearing for partner-enabled market knowledge crea-
deploying tacit overseas knowledge. Strategic Man-
tion. MIS Quarterly, 29, 1, 145–187.
agement Journal, 22, 359–378.
Marshall, N. and Brady, T. (2001) Knowledge manage-
Tiwana, A. (2004) An empirical study of the effect of
ment and the politics of knowledge: illustrations
knowledge integration on software development per-
from complex products and systems. European Jour-
formance. Information and Software Technology, 46,
nal of Information Systems, 10, 99–112. 899–906.
Mavondo, F.T. and Rodrigo, E.M. (2001) The effect of
Uzzi, B. (1997) Social structure and competition in
relationship dimensions on interpersonal and inter- inter-firm networks: the paradox of embeddedness.
organizational commitment in organizations con- Administrative Science Quarterly, 42, 1, 35–67.
ducting business between Australia and China. Uzzi, B. and Lancaster, R. (2003) Relational embedded-
Journal of Business Research, 52, 2, 111–121. ness and learning: the case of bank loan managers and
Nielsen, B.B. (2005) The role of knowledge embedded- their clients. Management Science, 49, 4, 383–399.
ness in the creation of synergies in the strategies Vaaland, T. and Hakansson, H. (2003) Exploring
alliances. Journal of Business Research, 58, 9, interorganizational conflict in complex projects.
1194–1204. Industrial Marketing Management, 32, 127–138.
Nonaka, I. (1994) A dynamic theory of organizational Williams, T. (2005) Cooperation by design: structure
knowledge creation. Organization Science, 5, 1, 14–37. and cooperation in inter-organizational networks.
Nonaka, I. and Takeuchi, H. (1995) The Knowledge Journal of Business Research, 58, 2, 223–231.
Creation Company: How Japanese Companies Create Wu, W-Y. and Tsai, H-J. (2005) Impact of social capital
the Dynamics of Innovation. Oxford: Oxford Univer- and business operation mode on intellectual capital
sity Press. and knowledge management. International Journal of
Oliver, C. (1990) Determinants of interorganizational Technology Management, 30, 1/2, 147–171.
relationships: integration and future directions. Wuyts, S., Stremersch, S., van Den Bulte, C. and
Academy of Management Review, 15, 241–265. Franses, P.H. (2004) Vertical marketing systems for
Oshri, I. and Newell, S. (2005) Component sharing in complex products: a triadic perspective. Journal of
complex product and systems: challenges, solutions, Marketing Research, 41, November, 479–487.
and practical implications. IEEE Transaction on Yli-Renko, H., Autio, E. and Sapienza, H.J. (2001)
Engineering Management, 52, 4, 509–521. Social capital, knowledge acquisition, and knowl-

r 2008 The Authors R&D Management 38, 4, 2008 439


Journal compilation r 2008 Blackwell Publishing Ltd
Eric W.T. Ngai, Chen Jin and Tong Liang

edge exploitation in young technology-based firms. Jin Chen is a Full Professor at the College of
Strategic Management Journal, 22, 587–613. Public Administration, Zhejiang University. He
received his PhD from Zhejiang University and
was a Visiting Scholar at Sloan School of Man-
Eric W. T. Ngai is an Associate Professor in the agement, MIT as well as visiting fellow at SPRU,
Department of Management and Marketing at University of Sussex, UK. His main research
The Hong Kong Polytechnic University. His areas are managing technological innovation
current research interests are in the areas of and R&D management. He has published four
decision support systems and knowledge and books on management of technology and innova-
technology management. He has papers pub- tion and various papers which appeared in IEEE
lished in a number of international journals Transactions on Engineering Management and
including MIS Quarterly, Journal of Operations others. He is also a guest editor for International
Management, Production & Operations Manage- Journal of Manpower. E-mail: cjhd@zju.edu.cn
ment, IEEE Transactions on Systems, Man and
Cybernetics, Information & Management, Decision Liang Tong received his PhD from the School of
Support Systems, European Journal of Operational Management, Zhejiang University. His research
Research, Transportation Research and others. He interests are in managing R&D-based innovation
serves on the editorial board of six international and organizational change. E-mail: cqtongliang@
journals. E-mail: mswtngai@inet.polyu.edu.hk 163.com

440 R&D Management 38, 4, 2008 r 2008 The Authors


Journal compilation r 2008 Blackwell Publishing Ltd

You might also like